{
"schemaVersion": 1,
"taskId": "yambda-discovery-ranking-two-tower-case-1",
"attemptId": "gemini-3-8-flash-high-case1-fixed-r04-3d6fca60",
"taskChecksum": "8404d794ce8e18d154cd225cfe6b12bf6b4ee02aaefba1d315aa215c1175a013",
"agentVersion": "0.61.0",
"model": "gemini-3.8-flash",
"sessionId": "333a7290-91e9-45e9-97c9-d83fb9219278",
"steps": [
{
"id": 1,
"role": "user",
"timestamp": "2026-09-29T05:36:41.212Z",
"blocks": [
{
"kind": "message",
"timestamp": "2026-09-29T05:36:41.212Z",
"text": "\nThis is the Gemini CLI. We are setting up the context for our chat.\nToday's date is Tuesday, September 29, 2026 (formatted according to the user's locale).\nMy operating system is: linux\nThe project's temporary directory is: /home/agent/.gemini/tmp/app\n- **Workspace Directories:**\n - /app\n - /task\n- **Directory Structure:**\n\nShowing up to 200 items (files + folders).\n\n/app/\n└───solution/\nShowing up to 200 items (files + folders).\n\n/task/\n├───.keep\n├───evaluate_public.py\n├───ranking_evaluation.py\n├───requirements.txt\n├───current_model/\n│ ├───train_current_model.py\n│ └───two_tower_model.py\n└───data/\n ├───candidate_catalog.parquet\n ├───target_users.parquet\n ├───test_eligible_candidates.npz\n ├───test_preference_state.parquet\n ├───train_dislikes.parquet\n ├───train_likes.parquet\n ├───train_listens.parquet\n ├───train_undislikes.parquet\n ├───train_unlikes.parquet\n ├───val_dislikes.parquet\n ├───val_likes.parquet\n ├───val_listens.parquet\n ├───val_undislikes.parquet\n ├───val_unlikes.parquet\n ├───validation_eligible_candidates.npz\n └───validation_preference_state.parquet\n\n\n"
}
]
},
{
"id": 2,
"role": "user",
"timestamp": "2026-09-29T05:36:41.257Z",
"blocks": [
{
"kind": "message",
"timestamp": "2026-09-29T05:36:41.257Z",
"text": "# Improve a production music-discovery two-tower\n\nYou have joined the recommendations team for a music streaming service. The\n\"Made for You\" shelf is served by a mature production two-tower model. Its\nchronological data protocol and indexed serving path are sound, but its offline\nretrieval quality no longer clears the team's next launch gate.\n\nYour job is to investigate the production model, improve its retrieval quality,\nand ship a better ranking artifact. Use the implementation and public validation\ndata to form and test your own diagnosis.\n\nThe source used to reproduce the production model is included under\n`/task/current_model/`. You may copy it into `/app/solution/`, edit it, add\ntests, and improve its training objective, but keep the data protocol,\ntwo-tower serving contract, and evaluation boundary intact.\n\n## Production Context\n\nThe model must satisfy a strict request-time latency budget. Item vectors are\nprecomputed offline. At request time, one user vector is computed from observed\ninteraction history and compared with candidate item vectors by a batchable\ntwo-tower score. Do not introduce a cross-encoder, per-candidate sequence model,\nor user-ID lookup table.\n\nThe current trainer is a mature baseline with chronological internal\nvalidation. Focus your investigation on how its training objective selects and\nuses negative examples. Diagnose limitations in the current negative-sampling\nstrategy and improve it under the fixed CPU budget. Validate changes with the\npublic backtest. The final approach must still train and rank within the supplied\nCPU environment while preserving the serving and data boundaries above.\n\n## What You Have\n\nAll public inputs are under `/task/data/`. A stage cutoff is the timestamp at\nwhich observed history and candidate eligibility are frozen for a ranking. The\npublic validation stage cutoff is `24786800`, when `train_*` ends. A 30-minute\ngap precedes the seven-day `val_*` window. For the final submission, that\nvalidation window becomes observed history, and another 30-minute gap precedes\nthe hidden seven-day outcome window.\n\nThe package includes:\n\n- `/task/current_model/train_current_model.py`: the production training and\n ranking script. It can reproduce the public validation ranking and produce a\n hidden-test submission after refitting.\n- `/task/current_model/two_tower_model.py`: model primitives used by the\n production script.\n- `/task/evaluate_public.py`: the public validation evaluator for development.\n- `/task/data/validation_eligible_candidates.npz` and\n `/task/data/test_eligible_candidates.npz`: platform-owned candidate sets for\n the validation and hidden-test cutoffs.\n- `/task/data/validation_preference_state.parquet` and\n `/task/data/test_preference_state.parquet`: active explicit preference state\n at the corresponding cutoffs.\n- `/task/requirements.txt`: the exact offline Python distribution inventory.\n\nYou can reproduce the production public backtest with:\n\n```sh\npython /task/current_model/train_current_model.py \\\n --stage validation \\\n --public-dir /task \\\n --output /app/solution/production_validation.parquet\npython /task/evaluate_public.py /app/solution/production_validation.parquet\n```\n\n## Data\n\nFor each event family there is a `train_*.parquet` and `val_*.parquet` file:\n\n| Family | Meaning | Columns |\n| --- | --- | --- |\n| `listens` | Playback events | `uid`, `timestamp`, `item_id`, `is_organic`, `played_ratio_pct`, `track_length_seconds` |\n| `likes` | Explicit positive actions | `uid`, `timestamp`, `item_id`, `is_organic` |\n| `dislikes` | Explicit negative actions | `uid`, `timestamp`, `item_id`, `is_organic` |\n| `unlikes` | Reversal of a prior like | `uid`, `timestamp`, `item_id`, `is_organic` |\n| `undislikes` | Reversal of a prior dislike | `uid`, `timestamp`, `item_id`, `is_organic` |\n\n`is_organic=1` means the listener sought out the track; `0` means the action came\nthrough a recommendation surface. A listen is a conventional positive at 50%\ncompletion. The discovery-quality objective below uses the stronger 80%\nthreshold.\n\nThe two public windows have different roles:\n\n1. For model development and selection, train only from `train_*` and score the\n subsequent `val_*` outcomes.\n2. After selecting an improvement, refit it on `train_* + val_*`. At hidden-test\n time the validation events are observed history, not recommendable outcomes.\n\nDo not use `val_*` events as features when measuring a validation ranking. They\nmay and should be used when refitting the final hidden-test ranker.\n\nAdditional files under `/task/data/`:\n\n- `candidate_catalog.parquet`: the 30,000 retrieval-catalog tracks, ordered by\n training positive-listen popularity.\n- `target_users.parquet`: users with at least 20 positive listens in the 90 days\n before the training cutoff, selected without future labels.\n- `validation_preference_state.parquet`: active explicit state at the training\n cutoff, for validation ranking.\n- `test_preference_state.parquet`: active explicit state after validation, for\n the final hidden-test ranking. State is `liked` or `disliked`; a pair may have\n both active channels.\n- `validation_eligible_candidates.npz` and `test_eligible_candidates.npz`:\n cutoff-correct candidate sets for every target user. Each contains aligned\n arrays `uids`, `item_ids`, and `packed_eligible`. For user row `r`, recover\n the Boolean item mask with `np.unpackbits(packed_eligible[r],\n count=len(item_ids), bitorder=\"little\")`.\n\nValidation candidates exclude all `train_*` encounters. Test candidates exclude\nall `train_*` and `val_*` encounters. Future test interactions are hidden and\nused only by the separate Harbor verifier.\n\n## Discovery Objective\n\nRank only the candidates returned for the relevant cutoff. Relevance is based\non organic behavior in the immediately following window:\n\n- grade 3: the user organically likes the track;\n- grade 1: the user organically listens to at least 80% of the track;\n- grade 0: otherwise.\n\nIf both positive events occur, grade 3 wins. Ranking quality is measured with\ngraded NDCG@10. DCG uses gain `2^grade - 1` and discount\n`1/log2(rank + 1)`, then averages per-user NDCG over users with at least one\nrelevant item.\n\nThe public evaluator reports this same objective for development, but its score\nis not a separate acceptance stage. Final acceptance is determined on the\nhidden test using the same objective. The Yambda logs contain interactions, not\nimpression propensities; treat the organic objective as a robust offline target,\nnot as a causal estimate of online lift.\n\n## Deliverable\n\nSubmit one final ranking as `/app/solution/submission.parquet` or\n`/app/solution/submission.csv`. It must be a regular file, not a symbolic link.\nIf both files exist, the verifier evaluates `submission.parquet` and ignores\n`submission.csv`.\n\nThe ranking must:\n\n- include the columns `uid`, `item_id`, and `rank`; only these columns are\n evaluated;\n- use finite, integer-valued entries in all three required columns;\n- cover exactly the users in `target_users.parquet`, with no missing or extra\n users;\n- contain exactly 100 rows and 100 distinct items for every target user;\n- contain every integer rank from 1 through 100 exactly once per user; and\n- use only items present in that user's `test_eligible_candidates.npz`\n candidate set.\n\nAlso leave the improved implementation used to build the artifact at\n`/app/solution/train_current_model.py`, along with any local modules it imports.\nA concise `/app/solution/README.md` with the diagnosis, before-and-after public\nscore, model change, and submission command is useful for the experiment record,\nbut it is not an automated acceptance gate.\n\n## Constraints\n\n- Write deliverables and intermediate artifacts only under `/app/solution/`.\n- Treat `/task/` and the installed Python environment as read-only.\n- Do not invoke nested agents or attempt to access verifier files, credentials,\n other trials, or host paths.\n- Do not install packages or use network access except for the model-provider\n connection managed by Harbor.\n- Do not alter cutoff definitions, target users, candidate eligibility, or\n relevance labels.\n- Improve the production pipeline in place. Do not substitute an unrelated\n model family or evade the indexed two-tower serving constraint.\n- The environment includes CPU PyTorch 2.8.0. No accelerator is available.\n- You have up to 60 minutes. Prioritize measured validation improvements, a\n correct final refit, and a valid submission over broad architecture search.\n- Keep the best valid submission saved throughout the attempt so it survives\n the deadline.\n"
}
]
},
{
"id": 3,
"role": "agent",
"timestamp": "2026-09-29T05:36:47.736Z",
"blocks": [
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:36:47.736Z",
"text": "Analyzing Model Requirements\n\nMy current focus is on dissecting the core mandates for enhancing the music-discovery tower model. I'm reviewing topic updates, identifying critical constraints, and understanding the operational environment to refine the task instructions."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:36:47.736Z",
"text": "Investigating Negative Sampling\n\nI'm now drilling into the training objective, specifically how negative examples are selected and utilized. My priority is to pinpoint limitations within the current negative-sampling strategy and devise improvements that fit within our strict CPU constraints. This diagnostic work aims to prepare for validation against the public backtest."
},
{
"kind": "tool",
"timestamp": "2026-09-29T05:36:47.736Z",
"callId": "update_topic__call_145386",
"name": "update_topic",
"input": {
"title": "Baseline Exploration and Benchmarking",
"strategic_intent": "Investigate the production two-tower baseline and evaluate its public validation performance.",
"summary": "Beginning the project by examining the production codebase in /task/current_model, reviewing the evaluation logic in evaluate_public.py and ranking_evaluation.py, and running the baseline validation pipeline to establish a performance benchmark."
},
"recorded": {
"id": "update_topic__call_145386",
"name": "update_topic",
"args": {
"title": "Baseline Exploration and Benchmarking",
"strategic_intent": "Investigate the production two-tower baseline and evaluate its public validation performance.",
"summary": "Beginning the project by examining the production codebase in /task/current_model, reviewing the evaluation logic in evaluate_public.py and ranking_evaluation.py, and running the baseline validation pipeline to establish a performance benchmark."
},
"result": [
{
"functionResponse": {
"id": "update_topic__call_145386",
"name": "update_topic",
"response": {
"output": "Current topic: \"Baseline Exploration and Benchmarking\"\nTopic summary: Beginning the project by examining the production codebase in /task/current_model, reviewing the evaluation logic in evaluate_public.py and ranking_evaluation.py, and running the baseline validation pipeline to establish a performance benchmark.\n\nStrategic Intent: Investigate the production two-tower baseline and evaluate its public validation performance."
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:36:47.754Z",
"resultDisplay": "## 📂 Topic: **Baseline Exploration and Benchmarking**\n\n**Summary:**\nBeginning the project by examining the production codebase in /task/current_model, reviewing the evaluation logic in evaluate_public.py and ranking_evaluation.py, and running the baseline validation pipeline to establish a performance benchmark.\n\n> [!STRATEGY]\n> **Intent:** Investigate the production two-tower baseline and evaluate its public validation performance.",
"description": "Update topic to: \"Baseline Exploration and Benchmarking\"",
"displayName": "Update Topic Context",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:36:47.755Z",
"content": "Current topic: \"Baseline Exploration and Benchmarking\"\nTopic summary: Beginning the project by examining the production codebase in /task/current_model, reviewing the evaluation logic in evaluate_public.py and ranking_evaluation.py, and running the baseline validation pipeline to establish a performance benchmark.\n\nStrategic Intent: Investigate the production two-tower baseline and evaluate its public validation performance.",
"isError": false
},
"responseMessageId": "ba52bf80-6e42-4f9e-b372-f2cd25dd7b45"
},
{
"kind": "tool",
"timestamp": "2026-09-29T05:36:47.736Z",
"callId": "read_file__call_145389",
"name": "read_file",
"input": {
"start_line": 1,
"file_path": "/task/evaluate_public.py",
"end_line": 150
},
"recorded": {
"id": "read_file__call_145389",
"name": "read_file",
"args": {
"start_line": 1,
"file_path": "/task/evaluate_public.py",
"end_line": 150
},
"result": [
{
"functionResponse": {
"id": "read_file__call_145389",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1-150 of 158 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 151.\n\n--- FILE CONTENT (truncated) ---\n#!/usr/bin/env python3\n\"\"\"Score a validation-stage ranking against public discovery labels.\n\n``build_task.py`` publishes this module with the solver-visible task files. It\nloads target users, validation eligibility, and validation-window events from\nthe adjacent ``data`` directory, then evaluates one ranking supplied on the\ncommand line. Stage-independent mechanics come from the adjacent public\n``ranking_evaluation.py`` module. The builder installs the same checked-in\nmodule in the private judge, so public and private scoring mechanics cannot\ndrift. A relevant candidate receives grade 3 for an organic like or grade 1 for\nan organic listen with at least 80 percent completion; a like takes precedence\nwhen both events exist.\n\nBefore scoring, the ranking must contain exactly 100 distinct eligible items\nfor every target user, with integer-valued ``uid``, ``item_id``, and ``rank``\ncolumns and ranks 1 through 100 exactly once per user. Extra columns are\nignored. Quality is graded NDCG@10, averaged only over users who have at least\none eligible relevant item in the validation window.\n\nThe evaluator prints a JSON object containing the model path and scores. It\ndoes not apply the task's pass threshold; its purpose is model development and\nselection. The private judge applies the threshold to the final submission.\n\nUsage::\n\n python evaluate_public.py \n\"\"\"\nfrom __future__ import annotations\n\nimport json\nimport sys\nfrom pathlib import Path\n\nimport pandas as pd\n\nfrom ranking_evaluation import (\n CandidateSet,\n RankingValidationError,\n ranked_items,\n read_ranking,\n score_rankings,\n validate_ranking,\n)\n\n\nHERE = Path(__file__).resolve().parent\nDATA = HERE / \"data\"\nLISTEN_STRONG = 80\n\n\ndef validation_truth(\n eligibility: CandidateSet, targets: set[int]\n) -> dict[int, dict[int, int]]:\n \"\"\"Build cutoff-correct graded relevance from public validation events.\n\n Only target-user events for eligible candidates are retained. Organic\n listens completed to at least 80 percent receive grade 1, and organic likes\n receive grade 3. Repeated user-item events are collapsed, and processing\n likes second ensures that grade 3 wins over grade 1.\n\n Args:\n eligibility: Validation-cutoff candidate membership service.\n targets: Complete set of target user IDs.\n\n Returns:\n A nested ``uid -> item_id -> grade`` mapping. Users without an eligible\n relevant event are omitted.\n \"\"\"\n listens = pd.read_parquet(DATA / \"val_listens.parquet\")\n likes = pd.read_parquet(DATA / \"val_likes.parquet\")\n strong = listens[\n listens.uid.isin(targets)\n & (listens.is_organic == 1)\n & (listens.played_ratio_pct >= LISTEN_STRONG)\n ][[\"uid\", \"item_id\"]].drop_duplicates()\n organic_likes = likes[\n likes.uid.isin(targets) & (likes.is_organic == 1)\n ][[\"uid\", \"item_id\"]].drop_duplicates()\n\n graded: dict[int, dict[int, int]] = {}\n for raw_uid, raw_item in strong.itertuples(index=False, name=None):\n uid, item = int(raw_uid), int(raw_item)\n if eligibility.is_eligible(uid, item):\n graded.setdefault(uid, {})[item] = 1\n for raw_uid, raw_item in organic_likes.itertuples(index=False, name=None):\n uid, item = int(raw_uid), int(raw_item)\n if eligibility.is_eligible(uid, item):\n graded.setdefault(uid, {})[item] = 3\n return graded\n\n\ndef load_submission(\n path: Path, eligibility: CandidateSet, targets: set[int]\n) -> dict[int, list[int]]:\n \"\"\"Load and validate a public validation ranking.\n\n Files ending in ``.parquet`` are read as parquet; every other suffix is\n read as CSV. The required columns must be integer-valued, cover exactly the\n target-user set, contain 100 rows and ranks 1 through 100 per user, avoid\n duplicate user-item pairs, and contain only cutoff-eligible items.\n\n Args:\n path: Ranking file to read.\n eligibility: Validation-cutoff candidate membership service.\n targets: Exact set of users the ranking must cover.\n\n Returns:\n A ``uid -> ranked item IDs`` mapping ordered by ascending rank.\n\n Raises:\n ValueError: If the ranking violates the required schema or invariants.\n \"\"\"\n frame = validate_ranking(read_ranking(path), targets)\n if not all(\n eligibility.is_eligible(uid, item)\n for uid, item in frame[[\"uid\", \"item_id\"]].itertuples(index=False, name=None)\n ):\n raise RankingValidationError(\n \"submission contains an item rejected by eligibility service\"\n )\n return ranked_items(frame)\n\n\ndef score(\n ranked: dict[int, list[int]],\n graded: dict[int, dict[int, int]],\n) -> dict[str, float | int]:\n \"\"\"Aggregate per-user NDCG@10 over users with public relevance labels.\n\n Args:\n ranked: Ranked item IDs for every target user.\n graded: Eligible validation relevance keyed by user and item.\n\n Returns:\n The six-decimal ``quality_ndcg@10`` value and number of scored users.\n \"\"\"\n return score_rankings(ranked, graded)\n\n\ndef main() -> None:\n \"\"\"Run the public evaluator CLI and print its JSON score report.\"\"\"\n targets_frame = pd.read_parquet(DATA / \"target_users.parquet\")\n targets = set(targets_frame.uid.astype(int))\n eligibility = CandidateSet(DATA / \"validation_eligible_candidates.npz\")\n graded = validation_truth(eligibility, targets)\n\n if len(sys.argv) != 2:\n raise SystemExit(\n \"usage: evaluate_public.py \"\n )"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:36:47.754Z",
"resultDisplay": "Read lines 1-150 of 158 from ../task/evaluate_public.py",
"description": "../task/evaluate_public.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:36:47.755Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1-150 of 158 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 151.\n\n--- FILE CONTENT (truncated) ---\n#!/usr/bin/env python3\n\"\"\"Score a validation-stage ranking against public discovery labels.\n\n``build_task.py`` publishes this module with the solver-visible task files. It\nloads target users, validation eligibility, and validation-window events from\nthe adjacent ``data`` directory, then evaluates one ranking supplied on the\ncommand line. Stage-independent mechanics come from the adjacent public\n``ranking_evaluation.py`` module. The builder installs the same checked-in\nmodule in the private judge, so public and private scoring mechanics cannot\ndrift. A relevant candidate receives grade 3 for an organic like or grade 1 for\nan organic listen with at least 80 percent completion; a like takes precedence\nwhen both events exist.\n\nBefore scoring, the ranking must contain exactly 100 distinct eligible items\nfor every target user, with integer-valued ``uid``, ``item_id``, and ``rank``\ncolumns and ranks 1 through 100 exactly once per user. Extra columns are\nignored. Quality is graded NDCG@10, averaged only over users who have at least\none eligible relevant item in the validation window.\n\nThe evaluator prints a JSON object containing the model path and scores. It\ndoes not apply the task's pass threshold; its purpose is model development and\nselection. The private judge applies the threshold to the final submission.\n\nUsage::\n\n python evaluate_public.py \n\"\"\"\nfrom __future__ import annotations\n\nimport json\nimport sys\nfrom pathlib import Path\n\nimport pandas as pd\n\nfrom ranking_evaluation import (\n CandidateSet,\n RankingValidationError,\n ranked_items,\n read_ranking,\n score_rankings,\n validate_ranking,\n)\n\n\nHERE = Path(__file__).resolve().parent\nDATA = HERE / \"data\"\nLISTEN_STRONG = 80\n\n\ndef validation_truth(\n eligibility: CandidateSet, targets: set[int]\n) -> dict[int, dict[int, int]]:\n \"\"\"Build cutoff-correct graded relevance from public validation events.\n\n Only target-user events for eligible candidates are retained. Organic\n listens completed to at least 80 percent receive grade 1, and organic likes\n receive grade 3. Repeated user-item events are collapsed, and processing\n likes second ensures that grade 3 wins over grade 1.\n\n Args:\n eligibility: Validation-cutoff candidate membership service.\n targets: Complete set of target user IDs.\n\n Returns:\n A nested ``uid -> item_id -> grade`` mapping. Users without an eligible\n relevant event are omitted.\n \"\"\"\n listens = pd.read_parquet(DATA / \"val_listens.parquet\")\n likes = pd.read_parquet(DATA / \"val_likes.parquet\")\n strong = listens[\n listens.uid.isin(targets)\n & (listens.is_organic == 1)\n & (listens.played_ratio_pct >= LISTEN_STRONG)\n ][[\"uid\", \"item_id\"]].drop_duplicates()\n organic_likes = likes[\n likes.uid.isin(targets) & (likes.is_organic == 1)\n ][[\"uid\", \"item_id\"]].drop_duplicates()\n\n graded: dict[int, dict[int, int]] = {}\n for raw_uid, raw_item in strong.itertuples(index=False, name=None):\n uid, item = int(raw_uid), int(raw_item)\n if eligibility.is_eligible(uid, item):\n graded.setdefault(uid, {})[item] = 1\n for raw_uid, raw_item in organic_likes.itertuples(index=False, name=None):\n uid, item = int(raw_uid), int(raw_item)\n if eligibility.is_eligible(uid, item):\n graded.setdefault(uid, {})[item] = 3\n return graded\n\n\ndef load_submission(\n path: Path, eligibility: CandidateSet, targets: set[int]\n) -> dict[int, list[int]]:\n \"\"\"Load and validate a public validation ranking.\n\n Files ending in ``.parquet`` are read as parquet; every other suffix is\n read as CSV. The required columns must be integer-valued, cover exactly the\n target-user set, contain 100 rows and ranks 1 through 100 per user, avoid\n duplicate user-item pairs, and contain only cutoff-eligible items.\n\n Args:\n path: Ranking file to read.\n eligibility: Validation-cutoff candidate membership service.\n targets: Exact set of users the ranking must cover.\n\n Returns:\n A ``uid -> ranked item IDs`` mapping ordered by ascending rank.\n\n Raises:\n ValueError: If the ranking violates the required schema or invariants.\n \"\"\"\n frame = validate_ranking(read_ranking(path), targets)\n if not all(\n eligibility.is_eligible(uid, item)\n for uid, item in frame[[\"uid\", \"item_id\"]].itertuples(index=False, name=None)\n ):\n raise RankingValidationError(\n \"submission contains an item rejected by eligibility service\"\n )\n return ranked_items(frame)\n\n\ndef score(\n ranked: dict[int, list[int]],\n graded: dict[int, dict[int, int]],\n) -> dict[str, float | int]:\n \"\"\"Aggregate per-user NDCG@10 over users with public relevance labels.\n\n Args:\n ranked: Ranked item IDs for every target user.\n graded: Eligible validation relevance keyed by user and item.\n\n Returns:\n The six-decimal ``quality_ndcg@10`` value and number of scored users.\n \"\"\"\n return score_rankings(ranked, graded)\n\n\ndef main() -> None:\n \"\"\"Run the public evaluator CLI and print its JSON score report.\"\"\"\n targets_frame = pd.read_parquet(DATA / \"target_users.parquet\")\n targets = set(targets_frame.uid.astype(int))\n eligibility = CandidateSet(DATA / \"validation_eligible_candidates.npz\")\n graded = validation_truth(eligibility, targets)\n\n if len(sys.argv) != 2:\n raise SystemExit(\n \"usage: evaluate_public.py \"\n )",
"isError": false
},
"responseMessageId": "ba52bf80-6e42-4f9e-b372-f2cd25dd7b45"
},
{
"kind": "tool",
"timestamp": "2026-09-29T05:36:47.736Z",
"callId": "read_file__call_145392",
"name": "read_file",
"input": {
"file_path": "/task/ranking_evaluation.py",
"end_line": 150,
"start_line": 1
},
"recorded": {
"id": "read_file__call_145392",
"name": "read_file",
"args": {
"file_path": "/task/ranking_evaluation.py",
"end_line": 150,
"start_line": 1
},
"result": [
{
"functionResponse": {
"id": "read_file__call_145392",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1-150 of 232 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 151.\n\n--- FILE CONTENT (truncated) ---\n\"\"\"Shared public primitives for Yambda discovery-ranking evaluation.\n\nBoth the solver-visible validation evaluator and the private submission judge\nimport this module from the generated task's ``public`` directory. It owns the\nstage-independent contract: reading a ranking, validating its shape, converting\nit to per-user rank order, looking up packed candidate eligibility, and\ncomputing graded NDCG@10.\n\nThe module deliberately contains no event-derived labels, private artifact\npaths, or pass thresholds. Each evaluator supplies its own relevance data and\nper-cutoff eligibility checks around these common operations.\n\"\"\"\nfrom __future__ import annotations\n\nfrom pathlib import Path\n\nimport numpy as np\nimport pandas as pd\n\n\n__all__ = [\n \"CandidateSet\",\n \"QUALITY_K\",\n \"RankingValidationError\",\n \"REQUIRED_COLUMNS\",\n \"TOP_N\",\n \"ndcg_at_k\",\n \"ranked_items\",\n \"read_ranking\",\n \"score_rankings\",\n \"validate_ranking\",\n]\n\nTOP_N = 100\nQUALITY_K = 10\nREQUIRED_COLUMNS = (\"uid\", \"item_id\", \"rank\")\n\n\nclass RankingValidationError(ValueError):\n \"\"\"Report a ranking-schema or structural-contract violation.\"\"\"\n\n\nclass CandidateSet:\n \"\"\"Provide membership lookups over one packed candidate-set archive.\n\n The archive contains aligned ``uids``, ``item_ids``, and\n ``packed_eligible`` arrays. Candidate column ``c`` is represented by bit\n ``c % 8`` of byte ``c // 8`` using little-endian bit order. Arrays are\n copied into memory so they remain valid after the archive is closed.\n \"\"\"\n\n def __init__(self, path: Path):\n \"\"\"Load an eligibility archive and build ID-to-position indexes.\n\n Args:\n path: Path to a ``.npz`` archive with the candidate-set schema.\n \"\"\"\n with np.load(path, allow_pickle=False) as artifact:\n self.uids = artifact[\"uids\"].astype(np.int64, copy=True)\n self.item_ids = artifact[\"item_ids\"].astype(np.int64, copy=True)\n self.packed = artifact[\"packed_eligible\"].astype(np.uint8, copy=True)\n self.uid_row = {int(uid): row for row, uid in enumerate(self.uids)}\n self.item_column = {\n int(item): column for column, item in enumerate(self.item_ids)\n }\n\n def is_eligible(self, uid: int, item_id: int) -> bool:\n \"\"\"Return whether ``item_id`` is eligible for ``uid``.\n\n IDs absent from the target-user or candidate-item arrays are treated as\n ineligible rather than raising an exception.\n\n Args:\n uid: Target user ID to query.\n item_id: Candidate item ID to query.\n\n Returns:\n ``True`` only when the corresponding packed eligibility bit is set.\n \"\"\"\n row = self.uid_row.get(int(uid))\n column = self.item_column.get(int(item_id))\n if row is None or column is None:\n return False\n byte, bit = divmod(column, 8)\n return bool((int(self.packed[row, byte]) >> bit) & 1)\n\n\ndef read_ranking(path: Path) -> pd.DataFrame:\n \"\"\"Read a parquet or CSV ranking into a pandas frame.\n\n Args:\n path: Input path. A ``.parquet`` suffix selects parquet; every other\n suffix is interpreted as CSV.\n\n Returns:\n The ranking frame as loaded by pandas.\n \"\"\"\n return pd.read_parquet(path) if path.suffix == \".parquet\" else pd.read_csv(path)\n\n\ndef validate_ranking(\n frame: pd.DataFrame,\n targets: set[int],\n) -> pd.DataFrame:\n \"\"\"Normalize and validate the stage-independent ranking contract.\n\n Required values must be finite integers. The user set must exactly match\n ``targets``; every user must have 100 rows, distinct items, and each rank\n from 1 through 100 exactly once. Cutoff-specific candidate eligibility is\n intentionally left to the calling evaluator.\n\n Args:\n frame: Raw ranking frame. Extra columns are discarded from an internal\n copy; the caller's frame is not modified.\n targets: Exact set of target user IDs the ranking must cover.\n\n Returns:\n A normalized frame containing only the required ``int64`` columns.\n\n Raises:\n RankingValidationError: If the ranking violates the shared contract.\n \"\"\"\n required = list(REQUIRED_COLUMNS)\n if not all(column in frame for column in required):\n raise RankingValidationError(f\"submission requires columns {required}\")\n\n normalized = frame[required].copy()\n for column in required:\n numeric = pd.to_numeric(normalized[column], errors=\"coerce\")\n values = numeric.to_numpy(dtype=float)\n if numeric.isna().any() or not np.isfinite(values).all():\n raise RankingValidationError(\n f\"submission {column} must be finite numeric values\"\n )\n if not np.equal(numeric, np.floor(numeric)).all():\n raise RankingValidationError(\n f\"submission {column} must contain integers\"\n )\n normalized[column] = numeric.astype(np.int64)\n\n submitted_users = set(map(int, normalized.uid.unique()))\n if submitted_users != targets:\n raise RankingValidationError(\n f\"submission user set mismatch: missing={len(targets - submitted_users)} \"\n f\"extra={len(submitted_users - targets)}\"\n )\n counts = normalized.groupby(\"uid\").size()\n if not (counts == TOP_N).all():\n raise RankingValidationError(\n f\"submission must contain exactly {TOP_N} rows for every target user\""
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:36:47.754Z",
"resultDisplay": "Read lines 1-150 of 232 from ../task/ranking_evaluation.py",
"description": "../task/ranking_evaluation.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:36:47.755Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1-150 of 232 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 151.\n\n--- FILE CONTENT (truncated) ---\n\"\"\"Shared public primitives for Yambda discovery-ranking evaluation.\n\nBoth the solver-visible validation evaluator and the private submission judge\nimport this module from the generated task's ``public`` directory. It owns the\nstage-independent contract: reading a ranking, validating its shape, converting\nit to per-user rank order, looking up packed candidate eligibility, and\ncomputing graded NDCG@10.\n\nThe module deliberately contains no event-derived labels, private artifact\npaths, or pass thresholds. Each evaluator supplies its own relevance data and\nper-cutoff eligibility checks around these common operations.\n\"\"\"\nfrom __future__ import annotations\n\nfrom pathlib import Path\n\nimport numpy as np\nimport pandas as pd\n\n\n__all__ = [\n \"CandidateSet\",\n \"QUALITY_K\",\n \"RankingValidationError\",\n \"REQUIRED_COLUMNS\",\n \"TOP_N\",\n \"ndcg_at_k\",\n \"ranked_items\",\n \"read_ranking\",\n \"score_rankings\",\n \"validate_ranking\",\n]\n\nTOP_N = 100\nQUALITY_K = 10\nREQUIRED_COLUMNS = (\"uid\", \"item_id\", \"rank\")\n\n\nclass RankingValidationError(ValueError):\n \"\"\"Report a ranking-schema or structural-contract violation.\"\"\"\n\n\nclass CandidateSet:\n \"\"\"Provide membership lookups over one packed candidate-set archive.\n\n The archive contains aligned ``uids``, ``item_ids``, and\n ``packed_eligible`` arrays. Candidate column ``c`` is represented by bit\n ``c % 8`` of byte ``c // 8`` using little-endian bit order. Arrays are\n copied into memory so they remain valid after the archive is closed.\n \"\"\"\n\n def __init__(self, path: Path):\n \"\"\"Load an eligibility archive and build ID-to-position indexes.\n\n Args:\n path: Path to a ``.npz`` archive with the candidate-set schema.\n \"\"\"\n with np.load(path, allow_pickle=False) as artifact:\n self.uids = artifact[\"uids\"].astype(np.int64, copy=True)\n self.item_ids = artifact[\"item_ids\"].astype(np.int64, copy=True)\n self.packed = artifact[\"packed_eligible\"].astype(np.uint8, copy=True)\n self.uid_row = {int(uid): row for row, uid in enumerate(self.uids)}\n self.item_column = {\n int(item): column for column, item in enumerate(self.item_ids)\n }\n\n def is_eligible(self, uid: int, item_id: int) -> bool:\n \"\"\"Return whether ``item_id`` is eligible for ``uid``.\n\n IDs absent from the target-user or candidate-item arrays are treated as\n ineligible rather than raising an exception.\n\n Args:\n uid: Target user ID to query.\n item_id: Candidate item ID to query.\n\n Returns:\n ``True`` only when the corresponding packed eligibility bit is set.\n \"\"\"\n row = self.uid_row.get(int(uid))\n column = self.item_column.get(int(item_id))\n if row is None or column is None:\n return False\n byte, bit = divmod(column, 8)\n return bool((int(self.packed[row, byte]) >> bit) & 1)\n\n\ndef read_ranking(path: Path) -> pd.DataFrame:\n \"\"\"Read a parquet or CSV ranking into a pandas frame.\n\n Args:\n path: Input path. A ``.parquet`` suffix selects parquet; every other\n suffix is interpreted as CSV.\n\n Returns:\n The ranking frame as loaded by pandas.\n \"\"\"\n return pd.read_parquet(path) if path.suffix == \".parquet\" else pd.read_csv(path)\n\n\ndef validate_ranking(\n frame: pd.DataFrame,\n targets: set[int],\n) -> pd.DataFrame:\n \"\"\"Normalize and validate the stage-independent ranking contract.\n\n Required values must be finite integers. The user set must exactly match\n ``targets``; every user must have 100 rows, distinct items, and each rank\n from 1 through 100 exactly once. Cutoff-specific candidate eligibility is\n intentionally left to the calling evaluator.\n\n Args:\n frame: Raw ranking frame. Extra columns are discarded from an internal\n copy; the caller's frame is not modified.\n targets: Exact set of target user IDs the ranking must cover.\n\n Returns:\n A normalized frame containing only the required ``int64`` columns.\n\n Raises:\n RankingValidationError: If the ranking violates the shared contract.\n \"\"\"\n required = list(REQUIRED_COLUMNS)\n if not all(column in frame for column in required):\n raise RankingValidationError(f\"submission requires columns {required}\")\n\n normalized = frame[required].copy()\n for column in required:\n numeric = pd.to_numeric(normalized[column], errors=\"coerce\")\n values = numeric.to_numpy(dtype=float)\n if numeric.isna().any() or not np.isfinite(values).all():\n raise RankingValidationError(\n f\"submission {column} must be finite numeric values\"\n )\n if not np.equal(numeric, np.floor(numeric)).all():\n raise RankingValidationError(\n f\"submission {column} must contain integers\"\n )\n normalized[column] = numeric.astype(np.int64)\n\n submitted_users = set(map(int, normalized.uid.unique()))\n if submitted_users != targets:\n raise RankingValidationError(\n f\"submission user set mismatch: missing={len(targets - submitted_users)} \"\n f\"extra={len(submitted_users - targets)}\"\n )\n counts = normalized.groupby(\"uid\").size()\n if not (counts == TOP_N).all():\n raise RankingValidationError(\n f\"submission must contain exactly {TOP_N} rows for every target user\"",
"isError": false
},
"responseMessageId": "ba52bf80-6e42-4f9e-b372-f2cd25dd7b45"
},
{
"kind": "tool",
"timestamp": "2026-09-29T05:36:47.736Z",
"callId": "read_file__call_145398",
"name": "read_file",
"input": {
"start_line": 1,
"file_path": "/task/current_model/train_current_model.py",
"end_line": 150
},
"recorded": {
"id": "read_file__call_145398",
"name": "read_file",
"args": {
"start_line": 1,
"file_path": "/task/current_model/train_current_model.py",
"end_line": 150
},
"result": [
{
"functionResponse": {
"id": "read_file__call_145398",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1-150 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 151.\n\n--- FILE CONTENT (truncated) ---\n#!/usr/bin/env python3\n\"\"\"Train the Yambda windowed two-tower baseline.\n\nEvery catalog listen with at least 50 percent completion is a sequence token.\nWithin each selected chronological window, every adjacent target contributes a\nloss. The fast default retains each user's most recent 128 training targets and\n16 held-out validation targets; passing zero for either cap restores the\nuncapped behavior for that side of the split. A 100-step truncated-BPTT block\ncomputes all of its timestep losses and carries the detached GRU hidden state\ninto the next block.\n\nThe model is a pure two-tower retriever: no user-ID embedding, no output\nnormalization, raw dot-product retrieval, sample-weighted binary cross entropy,\nand 24 uniform plus eight propensity-corrected unique in-batch negatives. Every\nselected adjacent target retains a nonzero weight. Only files under the supplied\npublic directory are read.\n\"\"\"\nfrom __future__ import annotations\n\nimport argparse\nimport json\nimport random\nimport sys\nimport time\nfrom dataclasses import asdict, dataclass\nfrom pathlib import Path\n\nimport numpy as np\nimport pandas as pd\nimport torch\nfrom torch import nn\nfrom torch.nn import functional as F\n\nimport two_tower_model as base\n\n\nCACHE_VERSION = 2\n\n\n@dataclass(frozen=True)\nclass Config:\n bptt_steps: int = 100\n user_batch_size: int = 64\n embedding_dim: int = 192\n hidden_dim: int = 192\n user_hidden_1: int = 512\n user_hidden_2: int = 256\n item_hidden_1: int = 512\n item_hidden_2: int = 256\n dropout: float = 0.05\n epochs: int = 1\n learning_rate: float = 1.0e-3\n weight_decay: float = 1.0e-5\n gradient_clip: float = 2.0\n loss_microbatch_size: int = 65_536\n uniform_negatives: int = 24\n inbatch_negatives: int = 8\n inbatch_correction_power: float = 0.5\n inbatch_correction_min: float = 0.25\n inbatch_correction_max: float = 4.0\n negative_filter_scope: str = \"full-training-history\"\n target_weighting: str = \"organic-strong\"\n ordinary_target_weight: float = 0.25\n organic_strong_target_weight: float = 1.0\n organic_like_target_weight: float = 3.0\n discovery_horizon_days: int = 7\n validation_fraction: float = 0.10\n max_train_targets_per_user: int = 128\n max_validation_targets_per_user: int = 16\n early_stopping_patience: int = 1\n early_stopping_min_delta: float = 1.0e-4\n seed: int = 29\n\n\n@dataclass(frozen=True)\nclass CandidateSet:\n uids: np.ndarray\n item_ids: np.ndarray\n packed: np.ndarray\n\n @classmethod\n def read(cls, path: Path) -> \"CandidateSet\":\n with np.load(path, allow_pickle=False) as artifact:\n result = cls(\n uids=artifact[\"uids\"].astype(np.int64, copy=True),\n item_ids=artifact[\"item_ids\"].astype(np.int64, copy=True),\n packed=artifact[\"packed_eligible\"].astype(np.uint8, copy=True),\n )\n expected = (len(result.uids), (len(result.item_ids) + 7) // 8)\n if result.packed.shape != expected:\n raise ValueError(\n f\"invalid candidate artifact shape {result.packed.shape}; \"\n f\"expected {expected}\"\n )\n return result\n\n\n@dataclass(frozen=True)\nclass SequenceData:\n item_ids: np.ndarray\n item_features: np.ndarray\n user_ids: np.ndarray\n offsets: np.ndarray\n sequence_timestamps: np.ndarray\n sequence_items: np.ndarray\n event_features: np.ndarray\n activity_features: np.ndarray\n state_ids: np.ndarray\n liked_states: np.ndarray\n disliked_states: np.ndarray\n state_counts: np.ndarray\n inference_user_rows: np.ndarray\n inference_liked_items: np.ndarray\n inference_disliked_items: np.ndarray\n inference_user_features: np.ndarray\n\n @property\n def lengths(self) -> np.ndarray:\n return np.diff(self.offsets)\n\n @property\n def target_count(self) -> int:\n return int(np.maximum(self.lengths - 1, 0).sum())\n\n\n@dataclass(frozen=True)\nclass TargetPartition:\n \"\"\"A contiguous recent target window and its chronological BCE split.\"\"\"\n\n window_starts: np.ndarray\n training_counts: np.ndarray\n validation_counts: np.ndarray\n history_training_counts: np.ndarray\n\n @property\n def selected_count(self) -> int:\n return int(self.training_counts.sum() + self.validation_counts.sum())\n\n\ndef log(message: str) -> None:\n print(message, file=sys.stderr, flush=True)\n\n\ndef seed_everything(seed: int) -> None:\n random.seed(seed)\n np.random.seed(seed)\n torch.manual_seed(seed)\n\n\ndef model_config(config: Config) -> base.TowerConfig:"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:36:47.754Z",
"resultDisplay": "Read lines 1-150 of 2165 from ../task/current_model/train_current_model.py",
"description": "../.../train_current_model.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:36:47.755Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1-150 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 151.\n\n--- FILE CONTENT (truncated) ---\n#!/usr/bin/env python3\n\"\"\"Train the Yambda windowed two-tower baseline.\n\nEvery catalog listen with at least 50 percent completion is a sequence token.\nWithin each selected chronological window, every adjacent target contributes a\nloss. The fast default retains each user's most recent 128 training targets and\n16 held-out validation targets; passing zero for either cap restores the\nuncapped behavior for that side of the split. A 100-step truncated-BPTT block\ncomputes all of its timestep losses and carries the detached GRU hidden state\ninto the next block.\n\nThe model is a pure two-tower retriever: no user-ID embedding, no output\nnormalization, raw dot-product retrieval, sample-weighted binary cross entropy,\nand 24 uniform plus eight propensity-corrected unique in-batch negatives. Every\nselected adjacent target retains a nonzero weight. Only files under the supplied\npublic directory are read.\n\"\"\"\nfrom __future__ import annotations\n\nimport argparse\nimport json\nimport random\nimport sys\nimport time\nfrom dataclasses import asdict, dataclass\nfrom pathlib import Path\n\nimport numpy as np\nimport pandas as pd\nimport torch\nfrom torch import nn\nfrom torch.nn import functional as F\n\nimport two_tower_model as base\n\n\nCACHE_VERSION = 2\n\n\n@dataclass(frozen=True)\nclass Config:\n bptt_steps: int = 100\n user_batch_size: int = 64\n embedding_dim: int = 192\n hidden_dim: int = 192\n user_hidden_1: int = 512\n user_hidden_2: int = 256\n item_hidden_1: int = 512\n item_hidden_2: int = 256\n dropout: float = 0.05\n epochs: int = 1\n learning_rate: float = 1.0e-3\n weight_decay: float = 1.0e-5\n gradient_clip: float = 2.0\n loss_microbatch_size: int = 65_536\n uniform_negatives: int = 24\n inbatch_negatives: int = 8\n inbatch_correction_power: float = 0.5\n inbatch_correction_min: float = 0.25\n inbatch_correction_max: float = 4.0\n negative_filter_scope: str = \"full-training-history\"\n target_weighting: str = \"organic-strong\"\n ordinary_target_weight: float = 0.25\n organic_strong_target_weight: float = 1.0\n organic_like_target_weight: float = 3.0\n discovery_horizon_days: int = 7\n validation_fraction: float = 0.10\n max_train_targets_per_user: int = 128\n max_validation_targets_per_user: int = 16\n early_stopping_patience: int = 1\n early_stopping_min_delta: float = 1.0e-4\n seed: int = 29\n\n\n@dataclass(frozen=True)\nclass CandidateSet:\n uids: np.ndarray\n item_ids: np.ndarray\n packed: np.ndarray\n\n @classmethod\n def read(cls, path: Path) -> \"CandidateSet\":\n with np.load(path, allow_pickle=False) as artifact:\n result = cls(\n uids=artifact[\"uids\"].astype(np.int64, copy=True),\n item_ids=artifact[\"item_ids\"].astype(np.int64, copy=True),\n packed=artifact[\"packed_eligible\"].astype(np.uint8, copy=True),\n )\n expected = (len(result.uids), (len(result.item_ids) + 7) // 8)\n if result.packed.shape != expected:\n raise ValueError(\n f\"invalid candidate artifact shape {result.packed.shape}; \"\n f\"expected {expected}\"\n )\n return result\n\n\n@dataclass(frozen=True)\nclass SequenceData:\n item_ids: np.ndarray\n item_features: np.ndarray\n user_ids: np.ndarray\n offsets: np.ndarray\n sequence_timestamps: np.ndarray\n sequence_items: np.ndarray\n event_features: np.ndarray\n activity_features: np.ndarray\n state_ids: np.ndarray\n liked_states: np.ndarray\n disliked_states: np.ndarray\n state_counts: np.ndarray\n inference_user_rows: np.ndarray\n inference_liked_items: np.ndarray\n inference_disliked_items: np.ndarray\n inference_user_features: np.ndarray\n\n @property\n def lengths(self) -> np.ndarray:\n return np.diff(self.offsets)\n\n @property\n def target_count(self) -> int:\n return int(np.maximum(self.lengths - 1, 0).sum())\n\n\n@dataclass(frozen=True)\nclass TargetPartition:\n \"\"\"A contiguous recent target window and its chronological BCE split.\"\"\"\n\n window_starts: np.ndarray\n training_counts: np.ndarray\n validation_counts: np.ndarray\n history_training_counts: np.ndarray\n\n @property\n def selected_count(self) -> int:\n return int(self.training_counts.sum() + self.validation_counts.sum())\n\n\ndef log(message: str) -> None:\n print(message, file=sys.stderr, flush=True)\n\n\ndef seed_everything(seed: int) -> None:\n random.seed(seed)\n np.random.seed(seed)\n torch.manual_seed(seed)\n\n\ndef model_config(config: Config) -> base.TowerConfig:",
"isError": false
},
"responseMessageId": "ba52bf80-6e42-4f9e-b372-f2cd25dd7b45"
},
{
"kind": "tool",
"timestamp": "2026-09-29T05:36:47.736Z",
"callId": "read_file__call_145407",
"name": "read_file",
"input": {
"file_path": "/task/current_model/two_tower_model.py",
"start_line": 1,
"end_line": 150
},
"recorded": {
"id": "read_file__call_145407",
"name": "read_file",
"args": {
"file_path": "/task/current_model/two_tower_model.py",
"start_line": 1,
"end_line": 150
},
"result": [
{
"functionResponse": {
"id": "read_file__call_145407",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1-150 of 434 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 151.\n\n--- FILE CONTENT (truncated) ---\n\"\"\"Shared public-data primitives for the Yambda two-tower baseline.\"\"\"\nfrom __future__ import annotations\n\nimport sys\nfrom dataclasses import dataclass\nfrom pathlib import Path\nfrom typing import Iterator\n\nimport numpy as np\nimport pandas as pd\nimport pyarrow.parquet as pq\nimport torch\nfrom torch import nn\n\n\nDAY = 86_400\nTRAIN_END = 24_786_800\nVALIDATION_END = 25_393_400\nLISTEN_POSITIVE_THRESHOLD = 50\nPROFILE_LOOKBACK_DAYS = 120\nLIKE_ITEMS = 48\nDISLIKE_ITEMS = 24\nTOP_N = 100\n\n\n@dataclass(frozen=True)\nclass TowerConfig:\n sequence_length: int = 100\n embedding_dim: int = 192\n hidden_dim: int = 192\n user_hidden_1: int = 512\n user_hidden_2: int = 256\n item_hidden_1: int = 512\n item_hidden_2: int = 256\n dropout: float = 0.05\n epochs: int = 1\n learning_rate: float = 1.0e-3\n weight_decay: float = 1.0e-5\n gradient_clip: float = 2.0\n uniform_negatives: int = 24\n inbatch_negatives: int = 8\n inbatch_correction_power: float = 0.5\n inbatch_correction_min: float = 0.25\n inbatch_correction_max: float = 4.0\n seed: int = 29\n\n\n@dataclass(frozen=True)\nclass ListenEvents:\n timestamps: np.ndarray\n items: np.ndarray\n organic: np.ndarray\n completion_pct: np.ndarray\n\n\n@dataclass(frozen=True)\nclass ExplicitEvents:\n timestamps: np.ndarray\n items: np.ndarray\n channel: np.ndarray\n active: np.ndarray\n organic_like: np.ndarray\n\n\ndef log(message: str) -> None:\n print(message, file=sys.stderr, flush=True)\n\n\ndef build_item_features(catalog: pd.DataFrame) -> np.ndarray:\n \"\"\"Return four standardized features built only from the public catalog.\"\"\"\n positives = catalog.positive_listens.to_numpy(dtype=np.float64, copy=False)\n organic = catalog.organic_positive_listens.to_numpy(dtype=np.float64, copy=False)\n popularity_rank = catalog.popularity_rank.to_numpy(dtype=np.float64, copy=False)\n features = np.column_stack(\n [\n np.log1p(positives),\n np.log1p(organic),\n organic / np.maximum(positives, 1.0),\n 1.0 - (popularity_rank - 1.0) / max(len(catalog) - 1, 1),\n ]\n )\n means = features.mean(axis=0, keepdims=True)\n scales = np.maximum(features.std(axis=0, keepdims=True), 1.0e-6)\n standardized = ((features - means) / scales).astype(np.float32)\n if not np.isfinite(standardized).all():\n raise RuntimeError(\"catalog features contain non-finite values\")\n return standardized\n\n\ndef _combine_listen_chunks(parts: list[ListenEvents]) -> ListenEvents:\n if len(parts) == 1:\n return parts[0]\n timestamps = np.concatenate([part.timestamps for part in parts])\n items = np.concatenate([part.items for part in parts])\n organic = np.concatenate([part.organic for part in parts])\n completion = np.concatenate([part.completion_pct for part in parts])\n order = np.argsort(timestamps, kind=\"stable\")\n return ListenEvents(\n timestamps=timestamps[order],\n items=items[order],\n organic=organic[order],\n completion_pct=completion[order],\n )\n\n\ndef iter_listen_users(\n path: Path, item_index: pd.Index\n) -> Iterator[tuple[int, ListenEvents]]:\n \"\"\"Stream one uid-sorted public listen file one user at a time.\"\"\"\n parquet = pq.ParquetFile(path)\n pending_uid: int | None = None\n pending: list[ListenEvents] = []\n previous_uid = -1\n for batch in parquet.iter_batches(\n batch_size=750_000,\n columns=[\"uid\", \"timestamp\", \"item_id\", \"is_organic\", \"played_ratio_pct\"],\n ):\n frame = batch.to_pandas()\n mapped = item_index.get_indexer(frame.item_id.to_numpy(dtype=np.int64))\n keep = mapped >= 0\n if not np.any(keep):\n continue\n uids = frame.uid.to_numpy(dtype=np.int64, copy=False)[keep]\n timestamps = frame.timestamp.to_numpy(dtype=np.int64, copy=False)[keep]\n items = mapped[keep].astype(np.int32, copy=False)\n organic = frame.is_organic.to_numpy(dtype=np.uint8, copy=False)[keep]\n completion = frame.played_ratio_pct.to_numpy(dtype=np.uint16, copy=False)[keep]\n if np.any(uids[1:] < uids[:-1]):\n raise RuntimeError(f\"{path.name} is not sorted by uid\")\n boundaries = np.r_[\n 0,\n np.flatnonzero(uids[1:] != uids[:-1]) + 1,\n len(uids),\n ]\n for start, end in zip(boundaries[:-1], boundaries[1:]):\n uid = int(uids[start])\n part = ListenEvents(\n timestamps[start:end].copy(),\n items[start:end].copy(),\n organic[start:end].copy(),\n completion[start:end].copy(),\n )\n if pending_uid is None:\n pending_uid = uid\n elif uid != pending_uid:\n if pending_uid < previous_uid:\n raise RuntimeError(f\"{path.name} user order regressed\")\n yield pending_uid, _combine_listen_chunks(pending)\n previous_uid = pending_uid\n pending_uid = uid"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:36:47.754Z",
"resultDisplay": "Read lines 1-150 of 434 from ../task/current_model/two_tower_model.py",
"description": "../.../two_tower_model.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:36:47.755Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1-150 of 434 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 151.\n\n--- FILE CONTENT (truncated) ---\n\"\"\"Shared public-data primitives for the Yambda two-tower baseline.\"\"\"\nfrom __future__ import annotations\n\nimport sys\nfrom dataclasses import dataclass\nfrom pathlib import Path\nfrom typing import Iterator\n\nimport numpy as np\nimport pandas as pd\nimport pyarrow.parquet as pq\nimport torch\nfrom torch import nn\n\n\nDAY = 86_400\nTRAIN_END = 24_786_800\nVALIDATION_END = 25_393_400\nLISTEN_POSITIVE_THRESHOLD = 50\nPROFILE_LOOKBACK_DAYS = 120\nLIKE_ITEMS = 48\nDISLIKE_ITEMS = 24\nTOP_N = 100\n\n\n@dataclass(frozen=True)\nclass TowerConfig:\n sequence_length: int = 100\n embedding_dim: int = 192\n hidden_dim: int = 192\n user_hidden_1: int = 512\n user_hidden_2: int = 256\n item_hidden_1: int = 512\n item_hidden_2: int = 256\n dropout: float = 0.05\n epochs: int = 1\n learning_rate: float = 1.0e-3\n weight_decay: float = 1.0e-5\n gradient_clip: float = 2.0\n uniform_negatives: int = 24\n inbatch_negatives: int = 8\n inbatch_correction_power: float = 0.5\n inbatch_correction_min: float = 0.25\n inbatch_correction_max: float = 4.0\n seed: int = 29\n\n\n@dataclass(frozen=True)\nclass ListenEvents:\n timestamps: np.ndarray\n items: np.ndarray\n organic: np.ndarray\n completion_pct: np.ndarray\n\n\n@dataclass(frozen=True)\nclass ExplicitEvents:\n timestamps: np.ndarray\n items: np.ndarray\n channel: np.ndarray\n active: np.ndarray\n organic_like: np.ndarray\n\n\ndef log(message: str) -> None:\n print(message, file=sys.stderr, flush=True)\n\n\ndef build_item_features(catalog: pd.DataFrame) -> np.ndarray:\n \"\"\"Return four standardized features built only from the public catalog.\"\"\"\n positives = catalog.positive_listens.to_numpy(dtype=np.float64, copy=False)\n organic = catalog.organic_positive_listens.to_numpy(dtype=np.float64, copy=False)\n popularity_rank = catalog.popularity_rank.to_numpy(dtype=np.float64, copy=False)\n features = np.column_stack(\n [\n np.log1p(positives),\n np.log1p(organic),\n organic / np.maximum(positives, 1.0),\n 1.0 - (popularity_rank - 1.0) / max(len(catalog) - 1, 1),\n ]\n )\n means = features.mean(axis=0, keepdims=True)\n scales = np.maximum(features.std(axis=0, keepdims=True), 1.0e-6)\n standardized = ((features - means) / scales).astype(np.float32)\n if not np.isfinite(standardized).all():\n raise RuntimeError(\"catalog features contain non-finite values\")\n return standardized\n\n\ndef _combine_listen_chunks(parts: list[ListenEvents]) -> ListenEvents:\n if len(parts) == 1:\n return parts[0]\n timestamps = np.concatenate([part.timestamps for part in parts])\n items = np.concatenate([part.items for part in parts])\n organic = np.concatenate([part.organic for part in parts])\n completion = np.concatenate([part.completion_pct for part in parts])\n order = np.argsort(timestamps, kind=\"stable\")\n return ListenEvents(\n timestamps=timestamps[order],\n items=items[order],\n organic=organic[order],\n completion_pct=completion[order],\n )\n\n\ndef iter_listen_users(\n path: Path, item_index: pd.Index\n) -> Iterator[tuple[int, ListenEvents]]:\n \"\"\"Stream one uid-sorted public listen file one user at a time.\"\"\"\n parquet = pq.ParquetFile(path)\n pending_uid: int | None = None\n pending: list[ListenEvents] = []\n previous_uid = -1\n for batch in parquet.iter_batches(\n batch_size=750_000,\n columns=[\"uid\", \"timestamp\", \"item_id\", \"is_organic\", \"played_ratio_pct\"],\n ):\n frame = batch.to_pandas()\n mapped = item_index.get_indexer(frame.item_id.to_numpy(dtype=np.int64))\n keep = mapped >= 0\n if not np.any(keep):\n continue\n uids = frame.uid.to_numpy(dtype=np.int64, copy=False)[keep]\n timestamps = frame.timestamp.to_numpy(dtype=np.int64, copy=False)[keep]\n items = mapped[keep].astype(np.int32, copy=False)\n organic = frame.is_organic.to_numpy(dtype=np.uint8, copy=False)[keep]\n completion = frame.played_ratio_pct.to_numpy(dtype=np.uint16, copy=False)[keep]\n if np.any(uids[1:] < uids[:-1]):\n raise RuntimeError(f\"{path.name} is not sorted by uid\")\n boundaries = np.r_[\n 0,\n np.flatnonzero(uids[1:] != uids[:-1]) + 1,\n len(uids),\n ]\n for start, end in zip(boundaries[:-1], boundaries[1:]):\n uid = int(uids[start])\n part = ListenEvents(\n timestamps[start:end].copy(),\n items[start:end].copy(),\n organic[start:end].copy(),\n completion[start:end].copy(),\n )\n if pending_uid is None:\n pending_uid = uid\n elif uid != pending_uid:\n if pending_uid < previous_uid:\n raise RuntimeError(f\"{path.name} user order regressed\")\n yield pending_uid, _combine_listen_chunks(pending)\n previous_uid = pending_uid\n pending_uid = uid",
"isError": false
},
"responseMessageId": "ba52bf80-6e42-4f9e-b372-f2cd25dd7b45"
}
]
},
{
"id": 4,
"role": "agent",
"timestamp": "2026-09-29T05:36:50.225Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:36:50.225Z",
"callId": "run_shell_command__call_4829114",
"name": "run_shell_command",
"input": {
"command": "nproc && free -h",
"description": "Check CPU cores and available memory"
},
"recorded": {
"id": "run_shell_command__call_4829114",
"name": "run_shell_command",
"args": {
"command": "nproc && free -h",
"description": "Check CPU cores and available memory"
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_4829114",
"name": "run_shell_command",
"response": {
"output": "\nOutput: 8\n/usr/bin/bash: line 4: free: command not found\nExit Code: 127\nProcess Group PGID: 1618\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:36:50.249Z",
"resultDisplay": "8\n/usr/bin/bash: line 4: free: command not found",
"description": "nproc && free -h",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T05:36:50.250Z",
"content": "\nOutput: 8\n/usr/bin/bash: line 4: free: command not found\nExit Code: 127\nProcess Group PGID: 1618\n",
"isError": false
},
"responseMessageId": "300cf581-d68e-4a7a-b337-8dc7ecf360a3"
}
]
},
{
"id": 5,
"role": "agent",
"timestamp": "2026-09-29T05:36:52.365Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:36:52.365Z",
"callId": "run_shell_command__call_3322442",
"name": "run_shell_command",
"input": {
"command": "python3 -c \"import psutil; print('RAM (GB):', psutil.virtual_memory().total / (1024**3)); print('PyTorch threads:', __import__('torch').get_num_threads())\"",
"description": "Check RAM and PyTorch default threads"
},
"recorded": {
"id": "run_shell_command__call_3322442",
"name": "run_shell_command",
"args": {
"command": "python3 -c \"import psutil; print('RAM (GB):', psutil.virtual_memory().total / (1024**3)); print('PyTorch threads:', __import__('torch').get_num_threads())\"",
"description": "Check RAM and PyTorch default threads"
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_3322442",
"name": "run_shell_command",
"response": {
"output": "\nOutput: Traceback (most recent call last):\n File \"\", line 1, in \nModuleNotFoundError: No module named 'psutil'\nExit Code: 1\nProcess Group PGID: 1622\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:36:52.396Z",
"resultDisplay": "Traceback (most recent call last):\n File \"\", line 1, in \nModuleNotFoundError: No module named 'psutil'",
"description": "Check RAM and PyTorch default threads",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T05:36:52.397Z",
"content": "\nOutput: Traceback (most recent call last):\n File \"\", line 1, in \nModuleNotFoundError: No module named 'psutil'\nExit Code: 1\nProcess Group PGID: 1622\n",
"isError": false
},
"responseMessageId": "12d8b045-42c1-472c-acf8-9f5489f955f6"
}
]
},
{
"id": 6,
"role": "agent",
"timestamp": "2026-09-29T05:37:05.566Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:37:05.566Z",
"callId": "run_shell_command__call_4836042",
"name": "run_shell_command",
"input": {
"command": "python3 -c \"import torch; print('Torch threads:', torch.get_num_threads()); f = open('/proc/meminfo').readlines(); print(f[0], f[1], f[2])\"",
"description": "Check /proc/meminfo and PyTorch threads"
},
"recorded": {
"id": "run_shell_command__call_4836042",
"name": "run_shell_command",
"args": {
"command": "python3 -c \"import torch; print('Torch threads:', torch.get_num_threads()); f = open('/proc/meminfo').readlines(); print(f[0], f[1], f[2])\"",
"description": "Check /proc/meminfo and PyTorch threads"
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_4836042",
"name": "run_shell_command",
"response": {
"output": "\nOutput: Torch threads: 4\nMemTotal: 15988968 kB\n MemFree: 1286700 kB\n MemAvailable: 14349892 kB\nProcess Group PGID: 1625\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:37:08.471Z",
"resultDisplay": "Torch threads: 4\nMemTotal: 15988968 kB\n MemFree: 1286700 kB\n MemAvailable: 14349892 kB",
"description": "python3 -c \"import torch; print('Torch threads:', torch.get_num_threads()); f = open('/proc/meminfo').readlines(); print(f[0], f[1], f[2])\"",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T05:37:08.472Z",
"content": "\nOutput: Torch threads: 4\nMemTotal: 15988968 kB\n MemFree: 1286700 kB\n MemAvailable: 14349892 kB\nProcess Group PGID: 1625\n",
"isError": false
},
"responseMessageId": "10c44d55-78be-455d-a334-8fe7d27af68d"
}
]
},
{
"id": 7,
"role": "agent",
"timestamp": "2026-09-29T05:37:11.911Z",
"blocks": [
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:37:11.911Z",
"text": "Analyzing Current Script\n\nI'm currently examining the `train_current_model.py` and `two_tower_model.py` scripts to understand their functionality. My immediate plan is to execute the baseline reproduction command to gauge its runtime and performance metrics. My system has 16 GB RAM and 8 vCPUs."
},
{
"kind": "tool",
"timestamp": "2026-09-29T05:37:11.911Z",
"callId": "run_shell_command__call_227960",
"name": "run_shell_command",
"input": {
"command": "mkdir -p /app/solution && time python /task/current_model/train_current_model.py --stage validation --public-dir /task --output /app/solution/production_validation.parquet",
"description": "Run the baseline training and ranking on validation stage"
},
"recorded": {
"id": "run_shell_command__call_227960",
"name": "run_shell_command",
"args": {
"command": "mkdir -p /app/solution && time python /task/current_model/train_current_model.py --stage validation --public-dir /task --output /app/solution/production_validation.parquet",
"description": "Run the baseline training and ranking on validation stage"
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_227960",
"name": "run_shell_command",
"response": {
"output": "\nOutput: loaded explicit events: rows=720,270 users=8,220\nprepared users=1,000 tokens=2,090,746 states=35,727 elapsed=4.9s\nprepared users=2,000 tokens=4,301,266 states=71,242 elapsed=9.9s\nprepared users=3,000 tokens=6,486,146 states=109,471 elapsed=15.1s\nprepared users=4,000 tokens=8,793,043 states=148,210 elapsed=20.5s\nprepared users=5,000 tokens=10,992,205 states=182,596 elapsed=25.3s\nprepared users=6,000 tokens=13,061,727 states=217,148 elapsed=30.0s\nprepared users=7,000 tokens=15,334,418 states=256,914 elapsed=35.5s\nprepared users=8,000 tokens=17,462,672 states=296,703 elapsed=40.8s\nprepared users=9,000 tokens=19,568,405 states=331,551 elapsed=45.5s\ntarget users without qualifying catalog sequence=7\nprepared stage=validation sequence_users=9,135 tokens=20,056,329 next_item_targets=20,047,194 states=339,133\nchronological BCE window: train=1,052,953 validation=130,103 selected=1,183,056 discarded_old_targets=18,864,138 all=20,047,194\ninbatch correction: min=0.250 mean_on_targets=1.001 max=4.000\ntarget weighting=organic-strong min=0.250 mean=0.575 max=1.000\nfull-history false-negative filter: users=9,135 unique_user_items=5,524,275 bytes=34,256,250\nepoch=1/1 batches=250 targets=987,417 batch_bce=0.547529 running_bce=0.541614 elapsed=200.7s\nepoch=1/1 train_loss=0.53984 validation_loss=0.49664 pair_accuracy=0.755 train_targets=1,052,953 validation_targets=130,103 train_seconds=214.0 validation_seconds=15.6\nrestored best validation-BCE checkpoint: epoch=1 loss=0.496636\nranked users=1,024/8,456\nranked users=2,048/8,456\nranked users=3,072/8,456\nranked users=4,096/8,456\nranked users=5,120/8,456\nranked users=6,144/8,456\nranked users=7,168/8,456\nranked users=8,192/8,456\nranked users=8,456/8,456\n{\n \"baseline\": \"yambda-windowed-sequence-two-tower-base\",\n \"stage\": \"validation\",\n \"output\": \"/app/solution/production_validation.parquet\",\n \"rows\": 845600,\n \"users\": 8456,\n \"runtime_seconds\": 364.792,\n \"public_inputs_only\": true,\n \"architecture\": {\n \"family\": \"pure-windowed-sequence-two-tower\",\n \"sequence_training\": \"every-adjacent-target-within-recent-chronological-window\",\n \"sequence_encoder\": \"one-layer-GRU-dense-timestep-loss\",\n \"bptt_hidden_state\": \"carried-and-detached\",\n \"model_selection\": \"chronological-validation-BCE-early-stopping\",\n \"loss_execution\": \"gradient-accumulated-microbatches\",\n \"user_id_embedding\": false,\n \"user_mlp\": \"582-512-256-192\",\n \"item_mlp\": \"196-512-256-192\",\n \"normalization\": \"none\",\n \"score\": \"raw-dot-product\",\n \"objective\": \"sample-weighted-BCE-at-every-timestep\",\n \"negative_sampling\": \"24-uniform-plus-8-propensity-corrected-inbatch\",\n \"false_negative_filter\": \"full-training-history\",\n \"target_weighting\": \"organic-strong\"\n },\n \"config\": {\n \"bptt_steps\": 100,\n \"user_batch_size\": 64,\n \"embedding_dim\": 192,\n \"hidden_dim\": 192,\n \"user_hidden_1\": 512,\n \"user_hidden_2\": 256,\n \"item_hidden_1\": 512,\n \"item_hidden_2\": 256,\n \"dropout\": 0.05,\n \"epochs\": 1,\n \"learning_rate\": 0.001,\n \"weight_decay\": 1e-05,\n \"gradient_clip\": 2.0,\n \"loss_microbatch_size\": 65536,\n \"uniform_negatives\": 24,\n \"inbatch_negatives\": 8,\n \"inbatch_correction_power\": 0.5,\n \"inbatch_correction_min\": 0.25,\n \"inbatch_correction_max\": 4.0,\n \"negative_filter_scope\": \"full-training-history\",\n \"target_weighting\": \"organic-strong\",\n \"ordinary_target_weight\": 0.25,\n \"organic_strong_target_weight\": 1.0,\n \"organic_like_target_weight\": 3.0,\n \"discovery_horizon_days\": 7,\n \"validation_fraction\": 0.1,\n \"max_train_targets_per_user\": 128,\n \"max_validation_targets_per_user\": 16,\n \"early_stopping_patience\": 1,\n \"early_stopping_min_delta\": 0.0001,\n \"seed\": 29\n },\n \"sequence_tokens\": 20056329,\n \"all_next_item_targets\": 20047194,\n \"selected_training_targets\": 1052953,\n \"selected_validation_targets\": 130103,\n \"discarded_old_targets\": 18864138,\n \"training_history\": [\n {\n \"epoch\": 1.0,\n \"loss\": 0.5398380454468475,\n \"positive_logit\": 0.7427799638768455,\n \"negative_logit\": -0.622224704133723,\n \"pair_accuracy\": 0.7547732721688433,\n \"targets\": 1052953.0,\n \"target_weight_sum\": 632551.0,\n \"batches\": 266.0,\n \"seconds\": 213.955934511,\n \"validation_loss\": 0.49663622024592563,\n \"validation_targets\": 130103.0,\n \"validation_target_weight_sum\": 78769.25,\n \"validation_seconds\": 15.638400071999968\n }\n ],\n \"embedding_diagnostics\": {\n \"query_norm_mean\": 1.6030172109603882,\n \"query_norm_std\": 0.7697498202323914,\n \"item_norm_mean\": 1.7570279836654663,\n \"item_norm_std\": 0.8279486298561096\n },\n \"torch_version\": \"2.8.0+cpu\",\n \"device\": \"cpu\"\n}\n\nreal\t6m9.134s\nuser\t31m9.260s\nsys\t1m42.522s\nProcess Group PGID: 1635\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:43:21.063Z",
"resultDisplay": "loaded explicit events: rows=720,270 users=8,220\nprepared users=1,000 tokens=2,090,746 states=35,727 elapsed=4.9s\nprepared users=2,000 tokens=4,301,266 states=71,242 elapsed=9.9s\nprepared users=3,000 tokens=6,486,146 states=109,471 elapsed=15.1s\nprepared users=4,000 tokens=8,793,043 states=148,210 elapsed=20.5s\nprepared users=5,000 tokens=10,992,205 states=182,596 elapsed=25.3s\nprepared users=6,000 tokens=13,061,727 states=217,148 elapsed=30.0s\nprepared users=7,000 tokens=15,334,418 states=256,914 elapsed=35.5s\nprepared users=8,000 tokens=17,462,672 states=296,703 elapsed=40.8s\nprepared users=9,000 tokens=19,568,405 states=331,551 elapsed=45.5s\ntarget users without qualifying catalog sequence=7\nprepared stage=validation sequence_users=9,135 tokens=20,056,329 next_item_targets=20,047,194 states=339,133\nchronological BCE window: train=1,052,953 validation=130,103 selected=1,183,056 discarded_old_targets=18,864,138 all=20,047,194\ninbatch correction: min=0.250 mean_on_targets=1.001 max=4.000\ntarget weighting=organic-strong min=0.250 mean=0.575 max=1.000\nfull-history false-negative filter: users=9,135 unique_user_items=5,524,275 bytes=34,256,250\nepoch=1/1 batches=250 targets=987,417 batch_bce=0.547529 running_bce=0.541614 elapsed=200.7s\nepoch=1/1 train_loss=0.53984 validation_loss=0.49664 pair_accuracy=0.755 train_targets=1,052,953 validation_targets=130,103 train_seconds=214.0 validation_seconds=15.6\nrestored best validation-BCE checkpoint: epoch=1 loss=0.496636\nranked users=1,024/8,456\nranked users=2,048/8,456\nranked users=3,072/8,456\nranked users=4,096/8,456\nranked users=5,120/8,456\nranked users=6,144/8,456\nranked users=7,168/8,456\nranked users=8,192/8,456\nranked users=8,456/8,456\n{\n \"baseline\": \"yambda-windowed-sequence-two-tower-base\",\n \"stage\": \"validation\",\n \"output\": \"/app/solution/production_validation.parquet\",\n \"rows\": 845600,\n \"users\": 8456,\n \"runtime_seconds\": 364.792,\n \"public_inputs_only\": true,\n \"architecture\": {\n \"family\": \"pure-windowed-sequence-two-tower\",\n \"sequence_training\": \"every-adjacent-target-within-recent-chronological-window\",\n \"sequence_encoder\": \"one-layer-GRU-dense-timestep-loss\",\n \"bptt_hidden_state\": \"carried-and-detached\",\n \"model_selection\": \"chronological-validation-BCE-early-stopping\",\n \"loss_execution\": \"gradient-accumulated-microbatches\",\n \"user_id_embedding\": false,\n \"user_mlp\": \"582-512-256-192\",\n \"item_mlp\": \"196-512-256-192\",\n \"normalization\": \"none\",\n \"score\": \"raw-dot-product\",\n \"objective\": \"sample-weighted-BCE-at-every-timestep\",\n \"negative_sampling\": \"24-uniform-plus-8-propensity-corrected-inbatch\",\n \"false_negative_filter\": \"full-training-history\",\n \"target_weighting\": \"organic-strong\"\n },\n \"config\": {\n \"bptt_steps\": 100,\n \"user_batch_size\": 64,\n \"embedding_dim\": 192,\n \"hidden_dim\": 192,\n \"user_hidden_1\": 512,\n \"user_hidden_2\": 256,\n \"item_hidden_1\": 512,\n \"item_hidden_2\": 256,\n \"dropout\": 0.05,\n \"epochs\": 1,\n \"learning_rate\": 0.001,\n \"weight_decay\": 1e-05,\n \"gradient_clip\": 2.0,\n \"loss_microbatch_size\": 65536,\n \"uniform_negatives\": 24,\n \"inbatch_negatives\": 8,\n \"inbatch_correction_power\": 0.5,\n \"inbatch_correction_min\": 0.25,\n \"inbatch_correction_max\": 4.0,\n \"negative_filter_scope\": \"full-training-history\",\n \"target_weighting\": \"organic-strong\",\n \"ordinary_target_weight\": 0.25,\n \"organic_strong_target_weight\": 1.0,\n \"organic_like_target_weight\": 3.0,\n \"discovery_horizon_days\": 7,\n \"validation_fraction\": 0.1,\n \"max_train_targets_per_user\": 128,\n \"max_validation_targets_per_user\": 16,\n \"early_stopping_patience\": 1,\n \"early_stopping_min_delta\": 0.0001,\n \"seed\": 29\n },\n \"sequence_tokens\": 20056329,\n \"all_next_item_targets\": 20047194,\n \"selected_training_targets\": 1052953,\n \"selected_validation_targets\": 130103,\n \"discarded_old_targets\": 18864138,\n \"training_history\": [\n {\n \"epoch\": 1.0,\n \"loss\": 0.5398380454468475,\n \"positive_logit\": 0.7427799638768455,\n \"negative_logit\": -0.622224704133723,\n \"pair_accuracy\": 0.7547732721688433,\n \"targets\": 1052953.0,\n \"target_weight_sum\": 632551.0,\n \"batches\": 266.0,\n \"seconds\": 213.955934511,\n \"validation_loss\": 0.49663622024592563,\n \"validation_targets\": 130103.0,\n \"validation_target_weight_sum\": 78769.25,\n \"validation_seconds\": 15.638400071999968\n }\n ],\n \"embedding_diagnostics\": {\n \"query_norm_mean\": 1.6030172109603882,\n \"query_norm_std\": 0.7697498202323914,\n \"item_norm_mean\": 1.7570279836654663,\n \"item_norm_std\": 0.8279486298561096\n },\n \"torch_version\": \"2.8.0+cpu\",\n \"device\": \"cpu\"\n}\n\nreal\t6m9.134s\nuser\t31m9.260s\nsys\t1m42.522s",
"description": "Run the baseline training and ranking on validation stage",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T05:43:21.064Z",
"content": "\nOutput: loaded explicit events: rows=720,270 users=8,220\nprepared users=1,000 tokens=2,090,746 states=35,727 elapsed=4.9s\nprepared users=2,000 tokens=4,301,266 states=71,242 elapsed=9.9s\nprepared users=3,000 tokens=6,486,146 states=109,471 elapsed=15.1s\nprepared users=4,000 tokens=8,793,043 states=148,210 elapsed=20.5s\nprepared users=5,000 tokens=10,992,205 states=182,596 elapsed=25.3s\nprepared users=6,000 tokens=13,061,727 states=217,148 elapsed=30.0s\nprepared users=7,000 tokens=15,334,418 states=256,914 elapsed=35.5s\nprepared users=8,000 tokens=17,462,672 states=296,703 elapsed=40.8s\nprepared users=9,000 tokens=19,568,405 states=331,551 elapsed=45.5s\ntarget users without qualifying catalog sequence=7\nprepared stage=validation sequence_users=9,135 tokens=20,056,329 next_item_targets=20,047,194 states=339,133\nchronological BCE window: train=1,052,953 validation=130,103 selected=1,183,056 discarded_old_targets=18,864,138 all=20,047,194\ninbatch correction: min=0.250 mean_on_targets=1.001 max=4.000\ntarget weighting=organic-strong min=0.250 mean=0.575 max=1.000\nfull-history false-negative filter: users=9,135 unique_user_items=5,524,275 bytes=34,256,250\nepoch=1/1 batches=250 targets=987,417 batch_bce=0.547529 running_bce=0.541614 elapsed=200.7s\nepoch=1/1 train_loss=0.53984 validation_loss=0.49664 pair_accuracy=0.755 train_targets=1,052,953 validation_targets=130,103 train_seconds=214.0 validation_seconds=15.6\nrestored best validation-BCE checkpoint: epoch=1 loss=0.496636\nranked users=1,024/8,456\nranked users=2,048/8,456\nranked users=3,072/8,456\nranked users=4,096/8,456\nranked users=5,120/8,456\nranked users=6,144/8,456\nranked users=7,168/8,456\nranked users=8,192/8,456\nranked users=8,456/8,456\n{\n \"baseline\": \"yambda-windowed-sequence-two-tower-base\",\n \"stage\": \"validation\",\n \"output\": \"/app/solution/production_validation.parquet\",\n \"rows\": 845600,\n \"users\": 8456,\n \"runtime_seconds\": 364.792,\n \"public_inputs_only\": true,\n \"architecture\": {\n \"family\": \"pure-windowed-sequence-two-tower\",\n \"sequence_training\": \"every-adjacent-target-within-recent-chronological-window\",\n \"sequence_encoder\": \"one-layer-GRU-dense-timestep-loss\",\n \"bptt_hidden_state\": \"carried-and-detached\",\n \"model_selection\": \"chronological-validation-BCE-early-stopping\",\n \"loss_execution\": \"gradient-accumulated-microbatches\",\n \"user_id_embedding\": false,\n \"user_mlp\": \"582-512-256-192\",\n \"item_mlp\": \"196-512-256-192\",\n \"normalization\": \"none\",\n \"score\": \"raw-dot-product\",\n \"objective\": \"sample-weighted-BCE-at-every-timestep\",\n \"negative_sampling\": \"24-uniform-plus-8-propensity-corrected-inbatch\",\n \"false_negative_filter\": \"full-training-history\",\n \"target_weighting\": \"organic-strong\"\n },\n \"config\": {\n \"bptt_steps\": 100,\n \"user_batch_size\": 64,\n \"embedding_dim\": 192,\n \"hidden_dim\": 192,\n \"user_hidden_1\": 512,\n \"user_hidden_2\": 256,\n \"item_hidden_1\": 512,\n \"item_hidden_2\": 256,\n \"dropout\": 0.05,\n \"epochs\": 1,\n \"learning_rate\": 0.001,\n \"weight_decay\": 1e-05,\n \"gradient_clip\": 2.0,\n \"loss_microbatch_size\": 65536,\n \"uniform_negatives\": 24,\n \"inbatch_negatives\": 8,\n \"inbatch_correction_power\": 0.5,\n \"inbatch_correction_min\": 0.25,\n \"inbatch_correction_max\": 4.0,\n \"negative_filter_scope\": \"full-training-history\",\n \"target_weighting\": \"organic-strong\",\n \"ordinary_target_weight\": 0.25,\n \"organic_strong_target_weight\": 1.0,\n \"organic_like_target_weight\": 3.0,\n \"discovery_horizon_days\": 7,\n \"validation_fraction\": 0.1,\n \"max_train_targets_per_user\": 128,\n \"max_validation_targets_per_user\": 16,\n \"early_stopping_patience\": 1,\n \"early_stopping_min_delta\": 0.0001,\n \"seed\": 29\n },\n \"sequence_tokens\": 20056329,\n \"all_next_item_targets\": 20047194,\n \"selected_training_targets\": 1052953,\n \"selected_validation_targets\": 130103,\n \"discarded_old_targets\": 18864138,\n \"training_history\": [\n {\n \"epoch\": 1.0,\n \"loss\": 0.5398380454468475,\n \"positive_logit\": 0.7427799638768455,\n \"negative_logit\": -0.622224704133723,\n \"pair_accuracy\": 0.7547732721688433,\n \"targets\": 1052953.0,\n \"target_weight_sum\": 632551.0,\n \"batches\": 266.0,\n \"seconds\": 213.955934511,\n \"validation_loss\": 0.49663622024592563,\n \"validation_targets\": 130103.0,\n \"validation_target_weight_sum\": 78769.25,\n \"validation_seconds\": 15.638400071999968\n }\n ],\n \"embedding_diagnostics\": {\n \"query_norm_mean\": 1.6030172109603882,\n \"query_norm_std\": 0.7697498202323914,\n \"item_norm_mean\": 1.7570279836654663,\n \"item_norm_std\": 0.8279486298561096\n },\n \"torch_version\": \"2.8.0+cpu\",\n \"device\": \"cpu\"\n}\n\nreal\t6m9.134s\nuser\t31m9.260s\nsys\t1m42.522s\nProcess Group PGID: 1635\n",
"isError": false
},
"responseMessageId": "61c80697-2bf6-4498-8da5-f030cd368f45"
}
]
},
{
"id": 8,
"role": "agent",
"timestamp": "2026-09-29T05:43:23.430Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:43:23.430Z",
"callId": "run_shell_command__call_34567",
"name": "run_shell_command",
"input": {
"command": "python /task/evaluate_public.py /app/solution/production_validation.parquet",
"description": "Evaluate baseline validation ranking"
},
"recorded": {
"id": "run_shell_command__call_34567",
"name": "run_shell_command",
"args": {
"command": "python /task/evaluate_public.py /app/solution/production_validation.parquet",
"description": "Evaluate baseline validation ranking"
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_34567",
"name": "run_shell_command",
"response": {
"output": "\nOutput: {\n \"model\": \"/app/solution/production_validation.parquet\",\n \"scores\": {\n \"quality_ndcg@10\": 0.02128,\n \"quality_users\": 4263\n }\n}\nProcess Group PGID: 1695\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:43:26.823Z",
"resultDisplay": "{\n \"model\": \"/app/solution/production_validation.parquet\",\n \"scores\": {\n \"quality_ndcg@10\": 0.02128,\n \"quality_users\": 4263\n }\n}",
"description": "python /task/evaluate_public.py /app/solution/production_validation.parquet",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T05:43:26.824Z",
"content": "\nOutput: {\n \"model\": \"/app/solution/production_validation.parquet\",\n \"scores\": {\n \"quality_ndcg@10\": 0.02128,\n \"quality_users\": 4263\n }\n}\nProcess Group PGID: 1695\n",
"isError": false
},
"responseMessageId": "21a3aad9-15b1-4171-a736-bdfff4795fd6"
}
]
},
{
"id": 9,
"role": "agent",
"timestamp": "2026-09-29T05:43:37.216Z",
"blocks": [
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:43:37.216Z",
"text": "Establishing a Valid Foundation\n\nI'm focused on ensuring the initial submission is robust. My current thought process involves verifying the generation of a valid submission early on and thoroughly testing its integrity to secure the best possible baseline performance, aiming to preserve it."
},
{
"kind": "tool",
"timestamp": "2026-09-29T05:43:37.216Z",
"callId": "read_file__call_103328",
"name": "read_file",
"input": {
"file_path": "/task/current_model/train_current_model.py",
"start_line": 151,
"end_line": 300
},
"recorded": {
"id": "read_file__call_103328",
"name": "read_file",
"args": {
"file_path": "/task/current_model/train_current_model.py",
"start_line": 151,
"end_line": 300
},
"result": [
{
"functionResponse": {
"id": "read_file__call_103328",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 151-300 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 301.\n\n--- FILE CONTENT (truncated) ---\n \"\"\"Map the dense trainer configuration onto the unchanged tower.\"\"\"\n return base.TowerConfig(\n sequence_length=config.bptt_steps,\n embedding_dim=config.embedding_dim,\n hidden_dim=config.hidden_dim,\n user_hidden_1=config.user_hidden_1,\n user_hidden_2=config.user_hidden_2,\n item_hidden_1=config.item_hidden_1,\n item_hidden_2=config.item_hidden_2,\n dropout=config.dropout,\n epochs=config.epochs,\n learning_rate=config.learning_rate,\n weight_decay=config.weight_decay,\n gradient_clip=config.gradient_clip,\n uniform_negatives=config.uniform_negatives,\n inbatch_negatives=config.inbatch_negatives,\n inbatch_correction_power=config.inbatch_correction_power,\n inbatch_correction_min=config.inbatch_correction_min,\n inbatch_correction_max=config.inbatch_correction_max,\n seed=config.seed,\n )\n\n\ndef append_state_snapshot(\n active_likes: dict[int, int],\n active_dislikes: dict[int, int],\n n_items: int,\n liked_states: list[np.ndarray],\n disliked_states: list[np.ndarray],\n state_counts: list[tuple[int, int]],\n) -> int:\n liked_states.append(base.pack_active_items(active_likes, base.LIKE_ITEMS, n_items))\n disliked_states.append(\n base.pack_active_items(active_dislikes, base.DISLIKE_ITEMS, n_items)\n )\n state_counts.append((len(active_likes), len(active_dislikes)))\n return len(liked_states) - 1\n\n\ndef causal_activity_features(\n timestamps: np.ndarray,\n items: np.ndarray,\n organic: np.ndarray,\n completion: np.ndarray,\n) -> np.ndarray:\n \"\"\"Compute the first four reviewed user features at every event cutoff.\"\"\"\n length = len(items)\n result = np.zeros((length, 4), dtype=np.float16)\n if not length:\n return result\n left_rows = np.searchsorted(\n timestamps,\n timestamps - base.PROFILE_LOOKBACK_DAYS * base.DAY,\n side=\"left\",\n )\n rows = np.arange(length, dtype=np.int64)\n counts = rows - left_rows + 1\n organic_sum = np.cumsum(organic.astype(np.float64))\n completion_sum = np.cumsum(np.minimum(completion, 100).astype(np.float64))\n before_organic = np.where(\n left_rows > 0, organic_sum[np.maximum(left_rows - 1, 0)], 0.0\n )\n before_completion = np.where(\n left_rows > 0, completion_sum[np.maximum(left_rows - 1, 0)], 0.0\n )\n\n unique_counts = np.empty(length, dtype=np.int32)\n frequencies: dict[int, int] = {}\n left = 0\n for right, raw_item in enumerate(items):\n item = int(raw_item)\n frequencies[item] = frequencies.get(item, 0) + 1\n required_left = int(left_rows[right])\n while left < required_left:\n old_item = int(items[left])\n remaining = frequencies[old_item] - 1\n if remaining:\n frequencies[old_item] = remaining\n else:\n del frequencies[old_item]\n left += 1\n unique_counts[right] = len(frequencies)\n\n result[:, 0] = (np.log1p(counts) / 8.0).astype(np.float16)\n result[:, 1] = (np.log1p(unique_counts) / 6.0).astype(np.float16)\n result[:, 2] = (\n (organic_sum - before_organic) / np.maximum(counts, 1)\n ).astype(np.float16)\n result[:, 3] = (\n (completion_sum - before_completion)\n / np.maximum(counts, 1)\n / 100.0\n ).astype(np.float16)\n return result\n\n\ndef build_sequence_parts(\n listens: base.ListenEvents,\n explicit: base.ExplicitEvents,\n n_items: int,\n liked_states: list[np.ndarray],\n disliked_states: list[np.ndarray],\n state_counts: list[tuple[int, int]],\n) -> tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray, np.ndarray] | None:\n qualifying = listens.completion_pct >= base.LISTEN_POSITIVE_THRESHOLD\n timestamps = listens.timestamps[qualifying].astype(np.int64, copy=False)\n items = listens.items[qualifying].astype(np.uint16, copy=False)\n organic = listens.organic[qualifying].astype(np.float32, copy=False)\n completion = listens.completion_pct[qualifying].astype(np.float32, copy=False)\n if not len(items):\n return None\n\n event_features = np.zeros((len(items), 3), dtype=np.float16)\n event_features[:, 0] = organic.astype(np.float16)\n event_features[:, 1] = (\n np.minimum(completion, 100.0) / 100.0\n ).astype(np.float16)\n if len(items) > 1:\n gaps = np.maximum(0, timestamps[1:] - timestamps[:-1])\n event_features[1:, 2] = np.exp(\n -gaps.astype(np.float64) / (60.0 * base.DAY)\n ).astype(np.float16)\n\n activity = causal_activity_features(\n timestamps, items, organic, completion\n )\n states = np.zeros(len(items), dtype=np.uint32)\n active_likes: dict[int, int] = {}\n active_dislikes: dict[int, int] = {}\n explicit_position = 0\n current_state = 0\n for row, timestamp in enumerate(timestamps):\n changed = False\n while (\n explicit_position < len(explicit.timestamps)\n and int(explicit.timestamps[explicit_position]) < int(timestamp)\n ):\n item = int(explicit.items[explicit_position])\n event_timestamp = int(explicit.timestamps[explicit_position])\n destination = (\n active_likes\n if int(explicit.channel[explicit_position]) == 0\n else active_dislikes\n )\n if int(explicit.active[explicit_position]) == 1:\n destination[item] = event_timestamp\n else:\n destination.pop(item, None)\n explicit_position += 1\n changed = True"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:43:37.223Z",
"resultDisplay": "Read lines 151-300 of 2165 from ../task/current_model/train_current_model.py",
"description": "../.../train_current_model.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:43:37.224Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 151-300 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 301.\n\n--- FILE CONTENT (truncated) ---\n \"\"\"Map the dense trainer configuration onto the unchanged tower.\"\"\"\n return base.TowerConfig(\n sequence_length=config.bptt_steps,\n embedding_dim=config.embedding_dim,\n hidden_dim=config.hidden_dim,\n user_hidden_1=config.user_hidden_1,\n user_hidden_2=config.user_hidden_2,\n item_hidden_1=config.item_hidden_1,\n item_hidden_2=config.item_hidden_2,\n dropout=config.dropout,\n epochs=config.epochs,\n learning_rate=config.learning_rate,\n weight_decay=config.weight_decay,\n gradient_clip=config.gradient_clip,\n uniform_negatives=config.uniform_negatives,\n inbatch_negatives=config.inbatch_negatives,\n inbatch_correction_power=config.inbatch_correction_power,\n inbatch_correction_min=config.inbatch_correction_min,\n inbatch_correction_max=config.inbatch_correction_max,\n seed=config.seed,\n )\n\n\ndef append_state_snapshot(\n active_likes: dict[int, int],\n active_dislikes: dict[int, int],\n n_items: int,\n liked_states: list[np.ndarray],\n disliked_states: list[np.ndarray],\n state_counts: list[tuple[int, int]],\n) -> int:\n liked_states.append(base.pack_active_items(active_likes, base.LIKE_ITEMS, n_items))\n disliked_states.append(\n base.pack_active_items(active_dislikes, base.DISLIKE_ITEMS, n_items)\n )\n state_counts.append((len(active_likes), len(active_dislikes)))\n return len(liked_states) - 1\n\n\ndef causal_activity_features(\n timestamps: np.ndarray,\n items: np.ndarray,\n organic: np.ndarray,\n completion: np.ndarray,\n) -> np.ndarray:\n \"\"\"Compute the first four reviewed user features at every event cutoff.\"\"\"\n length = len(items)\n result = np.zeros((length, 4), dtype=np.float16)\n if not length:\n return result\n left_rows = np.searchsorted(\n timestamps,\n timestamps - base.PROFILE_LOOKBACK_DAYS * base.DAY,\n side=\"left\",\n )\n rows = np.arange(length, dtype=np.int64)\n counts = rows - left_rows + 1\n organic_sum = np.cumsum(organic.astype(np.float64))\n completion_sum = np.cumsum(np.minimum(completion, 100).astype(np.float64))\n before_organic = np.where(\n left_rows > 0, organic_sum[np.maximum(left_rows - 1, 0)], 0.0\n )\n before_completion = np.where(\n left_rows > 0, completion_sum[np.maximum(left_rows - 1, 0)], 0.0\n )\n\n unique_counts = np.empty(length, dtype=np.int32)\n frequencies: dict[int, int] = {}\n left = 0\n for right, raw_item in enumerate(items):\n item = int(raw_item)\n frequencies[item] = frequencies.get(item, 0) + 1\n required_left = int(left_rows[right])\n while left < required_left:\n old_item = int(items[left])\n remaining = frequencies[old_item] - 1\n if remaining:\n frequencies[old_item] = remaining\n else:\n del frequencies[old_item]\n left += 1\n unique_counts[right] = len(frequencies)\n\n result[:, 0] = (np.log1p(counts) / 8.0).astype(np.float16)\n result[:, 1] = (np.log1p(unique_counts) / 6.0).astype(np.float16)\n result[:, 2] = (\n (organic_sum - before_organic) / np.maximum(counts, 1)\n ).astype(np.float16)\n result[:, 3] = (\n (completion_sum - before_completion)\n / np.maximum(counts, 1)\n / 100.0\n ).astype(np.float16)\n return result\n\n\ndef build_sequence_parts(\n listens: base.ListenEvents,\n explicit: base.ExplicitEvents,\n n_items: int,\n liked_states: list[np.ndarray],\n disliked_states: list[np.ndarray],\n state_counts: list[tuple[int, int]],\n) -> tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray, np.ndarray] | None:\n qualifying = listens.completion_pct >= base.LISTEN_POSITIVE_THRESHOLD\n timestamps = listens.timestamps[qualifying].astype(np.int64, copy=False)\n items = listens.items[qualifying].astype(np.uint16, copy=False)\n organic = listens.organic[qualifying].astype(np.float32, copy=False)\n completion = listens.completion_pct[qualifying].astype(np.float32, copy=False)\n if not len(items):\n return None\n\n event_features = np.zeros((len(items), 3), dtype=np.float16)\n event_features[:, 0] = organic.astype(np.float16)\n event_features[:, 1] = (\n np.minimum(completion, 100.0) / 100.0\n ).astype(np.float16)\n if len(items) > 1:\n gaps = np.maximum(0, timestamps[1:] - timestamps[:-1])\n event_features[1:, 2] = np.exp(\n -gaps.astype(np.float64) / (60.0 * base.DAY)\n ).astype(np.float16)\n\n activity = causal_activity_features(\n timestamps, items, organic, completion\n )\n states = np.zeros(len(items), dtype=np.uint32)\n active_likes: dict[int, int] = {}\n active_dislikes: dict[int, int] = {}\n explicit_position = 0\n current_state = 0\n for row, timestamp in enumerate(timestamps):\n changed = False\n while (\n explicit_position < len(explicit.timestamps)\n and int(explicit.timestamps[explicit_position]) < int(timestamp)\n ):\n item = int(explicit.items[explicit_position])\n event_timestamp = int(explicit.timestamps[explicit_position])\n destination = (\n active_likes\n if int(explicit.channel[explicit_position]) == 0\n else active_dislikes\n )\n if int(explicit.active[explicit_position]) == 1:\n destination[item] = event_timestamp\n else:\n destination.pop(item, None)\n explicit_position += 1\n changed = True",
"isError": false
},
"responseMessageId": "e459f874-6ff5-4ff5-b395-b28af3cc5555"
}
]
},
{
"id": 10,
"role": "agent",
"timestamp": "2026-09-29T05:43:38.999Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:43:38.999Z",
"callId": "read_file__call_158195",
"name": "read_file",
"input": {
"end_line": 600,
"file_path": "/task/current_model/train_current_model.py",
"start_line": 301
},
"recorded": {
"id": "read_file__call_158195",
"name": "read_file",
"args": {
"end_line": 600,
"file_path": "/task/current_model/train_current_model.py",
"start_line": 301
},
"result": [
{
"functionResponse": {
"id": "read_file__call_158195",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 301-600 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 601.\n\n--- FILE CONTENT (truncated) ---\n if changed:\n current_state = append_state_snapshot(\n active_likes,\n active_dislikes,\n n_items,\n liked_states,\n disliked_states,\n state_counts,\n )\n states[row] = np.uint32(current_state)\n return timestamps, items.copy(), event_features, activity, states\n\n\ndef final_user_features(\n timestamps: np.ndarray,\n items: np.ndarray,\n event_features: np.ndarray,\n history_end: int,\n like_count: int,\n dislike_count: int,\n) -> np.ndarray:\n start = int(\n np.searchsorted(\n timestamps,\n history_end - base.PROFILE_LOOKBACK_DAYS * base.DAY,\n side=\"left\",\n )\n )\n result = np.zeros(6, dtype=np.float32)\n recent_items = items[start:]\n if len(recent_items):\n result[0] = np.log1p(len(recent_items)) / 8.0\n result[1] = np.log1p(len(np.unique(recent_items))) / 6.0\n result[2] = float(np.mean(event_features[start:, 0]))\n result[3] = float(np.mean(event_features[start:, 1]))\n result[4] = np.log1p(like_count) / 6.0\n result[5] = np.log1p(dislike_count) / 4.0\n return result\n\n\ndef prepare_data(\n public: Path, stage: str, config: Config\n) -> tuple[SequenceData, CandidateSet]:\n data = public / \"data\"\n catalog = pd.read_parquet(data / \"candidate_catalog.parquet\").sort_values(\n \"popularity_rank\"\n )\n item_ids = catalog.item_id.to_numpy(dtype=np.int64, copy=True)\n item_index = pd.Index(item_ids)\n n_items = len(item_ids)\n if n_items >= np.iinfo(np.uint16).max:\n raise RuntimeError(\"uint16 item storage is too small for this catalog\")\n\n if stage == \"validation\":\n prefixes = (\"train\",)\n history_end = base.TRAIN_END\n candidate_name = \"validation_eligible_candidates.npz\"\n state_name = \"validation_preference_state.parquet\"\n else:\n prefixes = (\"train\", \"val\")\n history_end = base.VALIDATION_END\n candidate_name = \"test_eligible_candidates.npz\"\n state_name = \"test_preference_state.parquet\"\n eligibility = CandidateSet.read(data / candidate_name)\n if not np.array_equal(eligibility.item_ids, item_ids):\n raise RuntimeError(\"catalog and candidate artifact item order differ\")\n\n explicit_by_user = base.load_explicit_events(data, prefixes, item_index)\n inference_likes, inference_dislikes, like_counts, dislike_counts = (\n base.load_inference_state(\n data / state_name,\n item_index,\n eligibility.uids,\n n_items,\n )\n )\n inference_rows = {int(uid): row for row, uid in enumerate(eligibility.uids)}\n inference_user_features = np.zeros((len(eligibility.uids), 6), dtype=np.float32)\n inference_sequence_rows = np.full(len(eligibility.uids), -1, dtype=np.int32)\n\n # Snapshot zero is the shared empty explicit state.\n liked_states = [np.full(base.LIKE_ITEMS, n_items, dtype=np.uint16)]\n disliked_states = [\n np.full(base.DISLIKE_ITEMS, n_items, dtype=np.uint16)\n ]\n state_counts: list[tuple[int, int]] = [(0, 0)]\n user_ids: list[int] = []\n offsets = [0]\n timestamp_parts: list[np.ndarray] = []\n item_parts: list[np.ndarray] = []\n event_parts: list[np.ndarray] = []\n activity_parts: list[np.ndarray] = []\n state_parts: list[np.ndarray] = []\n\n listen_paths = [data / f\"{prefix}_listens.parquet\" for prefix in prefixes]\n started = time.monotonic()\n for seen_users, (uid, listens) in enumerate(\n base.merge_listen_streams(listen_paths, item_index), start=1\n ):\n parts = build_sequence_parts(\n listens,\n explicit_by_user.get(uid, base.empty_explicit_events()),\n n_items,\n liked_states,\n disliked_states,\n state_counts,\n )\n if parts is None:\n continue\n sequence_row = len(user_ids)\n user_ids.append(uid)\n timestamp_parts.append(parts[0].astype(np.uint32, copy=False))\n item_parts.append(parts[1])\n event_parts.append(parts[2])\n activity_parts.append(parts[3])\n state_parts.append(parts[4])\n offsets.append(offsets[-1] + len(parts[1]))\n inference_row = inference_rows.get(uid)\n if inference_row is not None:\n inference_sequence_rows[inference_row] = sequence_row\n inference_user_features[inference_row] = final_user_features(\n parts[0],\n parts[1],\n parts[2],\n history_end,\n int(like_counts[inference_row]),\n int(dislike_counts[inference_row]),\n )\n if seen_users % 1_000 == 0:\n log(\n f\"prepared users={seen_users:,} tokens={offsets[-1]:,} \"\n f\"states={len(liked_states):,} \"\n f\"elapsed={time.monotonic() - started:.1f}s\"\n )\n\n missing = np.flatnonzero(inference_sequence_rows < 0)\n if len(missing):\n inference_user_features[missing, 4] = np.log1p(like_counts[missing]) / 6.0\n inference_user_features[missing, 5] = (\n np.log1p(dislike_counts[missing]) / 4.0\n )\n log(f\"target users without qualifying catalog sequence={len(missing):,}\")\n\n prepared = SequenceData(\n item_ids=item_ids,\n item_features=base.build_item_features(catalog),\n user_ids=np.asarray(user_ids, dtype=np.int64),\n offsets=np.asarray(offsets, dtype=np.int64),\n sequence_timestamps=np.concatenate(timestamp_parts),\n sequence_items=np.concatenate(item_parts),\n event_features=np.concatenate(event_parts),\n activity_features=np.concatenate(activity_parts),\n state_ids=np.concatenate(state_parts),\n liked_states=np.stack(liked_states),\n disliked_states=np.stack(disliked_states),\n state_counts=np.asarray(state_counts, dtype=np.int32),\n inference_user_rows=inference_sequence_rows,\n inference_liked_items=inference_likes,\n inference_disliked_items=inference_dislikes,\n inference_user_features=inference_user_features,\n )\n log(\n f\"prepared stage={stage} sequence_users={len(prepared.user_ids):,} \"\n f\"tokens={len(prepared.sequence_items):,} \"\n f\"next_item_targets={prepared.target_count:,} \"\n f\"states={len(prepared.liked_states):,}\"\n )\n return prepared, eligibility\n\n\ndef cache_metadata(stage: str, config: Config) -> dict[str, object]:\n return {\n \"cache_version\": CACHE_VERSION,\n \"stage\": stage,\n \"positive_listen_threshold\": base.LISTEN_POSITIVE_THRESHOLD,\n \"profile_lookback_days\": base.PROFILE_LOOKBACK_DAYS,\n \"event_time_feature\": \"exp-negative-gap-over-60-days\",\n }\n\n\ndef save_cache(\n path: Path, prepared: SequenceData, stage: str, config: Config\n) -> None:\n path.parent.mkdir(parents=True, exist_ok=True)\n np.savez(\n path,\n metadata=np.asarray(json.dumps(cache_metadata(stage, config), sort_keys=True)),\n **prepared.__dict__,\n )\n log(f\"saved public-derived full-sequence cache: {path}\")\n\n\ndef load_cache(path: Path, stage: str, config: Config) -> SequenceData:\n with np.load(path, allow_pickle=False) as artifact:\n actual = json.loads(str(artifact[\"metadata\"].item()))\n expected = cache_metadata(stage, config)\n if actual != expected:\n raise RuntimeError(f\"sequence cache differs: {actual} != {expected}\")\n values = {\n field: artifact[field].copy()\n for field in SequenceData.__dataclass_fields__\n }\n prepared = SequenceData(**values)\n log(\n f\"loaded full-sequence cache: tokens={len(prepared.sequence_items):,} \"\n f\"targets={prepared.target_count:,}\"\n )\n return prepared\n\n\ndef load_or_prepare(\n public: Path,\n stage: str,\n config: Config,\n cache: Path | None,\n) -> tuple[SequenceData, CandidateSet]:\n candidate_name = (\n \"validation_eligible_candidates.npz\"\n if stage == \"validation\"\n else \"test_eligible_candidates.npz\"\n )\n eligibility = CandidateSet.read(public / \"data\" / candidate_name)\n if cache is not None and cache.is_file():\n prepared = load_cache(cache, stage, config)\n if not np.array_equal(prepared.item_ids, eligibility.item_ids):\n raise RuntimeError(\"cached catalog differs from candidate artifact\")\n return prepared, eligibility\n prepared, built_eligibility = prepare_data(public, stage, config)\n if cache is not None:\n save_cache(cache, prepared, stage, config)\n return prepared, built_eligibility\n\n\ndef build_item_correction(\n prepared: SequenceData,\n config: Config,\n target_starts: np.ndarray,\n target_counts: np.ndarray,\n) -> torch.Tensor:\n \"\"\"Estimate in-batch propensities from the selected training targets.\"\"\"\n counts = np.zeros(len(prepared.item_ids), dtype=np.float64)\n for user_row, count in enumerate(target_counts):\n if not count:\n continue\n offset = int(prepared.offsets[user_row])\n target_start = int(target_starts[user_row])\n counts += np.bincount(\n prepared.sequence_items[\n offset\n + target_start\n + 1 : offset\n + target_start\n + int(count)\n + 1\n ].astype(np.int64),\n minlength=len(prepared.item_ids),\n )\n total = float(counts.sum())\n probabilities = (counts + 1.0) / (total + len(counts))\n raw = np.power(\n (1.0 / len(counts)) / probabilities,\n config.inbatch_correction_power,\n )\n normalization = float(np.sum(counts * raw) / max(total, 1.0))\n correction = np.clip(\n raw / max(normalization, 1.0e-12),\n config.inbatch_correction_min,\n config.inbatch_correction_max,\n ).astype(np.float32)\n mean_on_targets = float(np.sum(counts * correction) / max(total, 1.0))\n log(\n f\"inbatch correction: min={correction.min():.3f} \"\n f\"mean_on_targets={mean_on_targets:.3f} max={correction.max():.3f}\"\n )\n return torch.from_numpy(correction)\n\n\ndef build_known_positive_filter(\n prepared: SequenceData, training_counts: np.ndarray\n) -> np.ndarray:\n \"\"\"Pack each user's complete public-training positive-item set.\"\"\"\n width = (len(prepared.item_ids) + 7) // 8\n packed = np.zeros((len(prepared.user_ids), width), dtype=np.uint8)\n retained = 0\n for user_row, target_count in enumerate(training_counts):\n if not target_count:\n continue\n offset = int(prepared.offsets[user_row])\n # T adjacent targets consume sequence positions 0 through T. The\n # chronological BCE suffix starts at T + 1 and is intentionally absent.\n items = np.unique(\n prepared.sequence_items[\n offset : offset + int(target_count) + 1\n ].astype(np.int64)\n )\n byte_columns = items >> 3\n bit_values = np.left_shift(\n np.uint8(1), (items & 7).astype(np.uint8)\n )\n np.bitwise_or.at(packed[user_row], byte_columns, bit_values)"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:43:39.004Z",
"resultDisplay": "Read lines 301-600 of 2165 from ../task/current_model/train_current_model.py",
"description": "../.../train_current_model.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:43:39.004Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 301-600 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 601.\n\n--- FILE CONTENT (truncated) ---\n if changed:\n current_state = append_state_snapshot(\n active_likes,\n active_dislikes,\n n_items,\n liked_states,\n disliked_states,\n state_counts,\n )\n states[row] = np.uint32(current_state)\n return timestamps, items.copy(), event_features, activity, states\n\n\ndef final_user_features(\n timestamps: np.ndarray,\n items: np.ndarray,\n event_features: np.ndarray,\n history_end: int,\n like_count: int,\n dislike_count: int,\n) -> np.ndarray:\n start = int(\n np.searchsorted(\n timestamps,\n history_end - base.PROFILE_LOOKBACK_DAYS * base.DAY,\n side=\"left\",\n )\n )\n result = np.zeros(6, dtype=np.float32)\n recent_items = items[start:]\n if len(recent_items):\n result[0] = np.log1p(len(recent_items)) / 8.0\n result[1] = np.log1p(len(np.unique(recent_items))) / 6.0\n result[2] = float(np.mean(event_features[start:, 0]))\n result[3] = float(np.mean(event_features[start:, 1]))\n result[4] = np.log1p(like_count) / 6.0\n result[5] = np.log1p(dislike_count) / 4.0\n return result\n\n\ndef prepare_data(\n public: Path, stage: str, config: Config\n) -> tuple[SequenceData, CandidateSet]:\n data = public / \"data\"\n catalog = pd.read_parquet(data / \"candidate_catalog.parquet\").sort_values(\n \"popularity_rank\"\n )\n item_ids = catalog.item_id.to_numpy(dtype=np.int64, copy=True)\n item_index = pd.Index(item_ids)\n n_items = len(item_ids)\n if n_items >= np.iinfo(np.uint16).max:\n raise RuntimeError(\"uint16 item storage is too small for this catalog\")\n\n if stage == \"validation\":\n prefixes = (\"train\",)\n history_end = base.TRAIN_END\n candidate_name = \"validation_eligible_candidates.npz\"\n state_name = \"validation_preference_state.parquet\"\n else:\n prefixes = (\"train\", \"val\")\n history_end = base.VALIDATION_END\n candidate_name = \"test_eligible_candidates.npz\"\n state_name = \"test_preference_state.parquet\"\n eligibility = CandidateSet.read(data / candidate_name)\n if not np.array_equal(eligibility.item_ids, item_ids):\n raise RuntimeError(\"catalog and candidate artifact item order differ\")\n\n explicit_by_user = base.load_explicit_events(data, prefixes, item_index)\n inference_likes, inference_dislikes, like_counts, dislike_counts = (\n base.load_inference_state(\n data / state_name,\n item_index,\n eligibility.uids,\n n_items,\n )\n )\n inference_rows = {int(uid): row for row, uid in enumerate(eligibility.uids)}\n inference_user_features = np.zeros((len(eligibility.uids), 6), dtype=np.float32)\n inference_sequence_rows = np.full(len(eligibility.uids), -1, dtype=np.int32)\n\n # Snapshot zero is the shared empty explicit state.\n liked_states = [np.full(base.LIKE_ITEMS, n_items, dtype=np.uint16)]\n disliked_states = [\n np.full(base.DISLIKE_ITEMS, n_items, dtype=np.uint16)\n ]\n state_counts: list[tuple[int, int]] = [(0, 0)]\n user_ids: list[int] = []\n offsets = [0]\n timestamp_parts: list[np.ndarray] = []\n item_parts: list[np.ndarray] = []\n event_parts: list[np.ndarray] = []\n activity_parts: list[np.ndarray] = []\n state_parts: list[np.ndarray] = []\n\n listen_paths = [data / f\"{prefix}_listens.parquet\" for prefix in prefixes]\n started = time.monotonic()\n for seen_users, (uid, listens) in enumerate(\n base.merge_listen_streams(listen_paths, item_index), start=1\n ):\n parts = build_sequence_parts(\n listens,\n explicit_by_user.get(uid, base.empty_explicit_events()),\n n_items,\n liked_states,\n disliked_states,\n state_counts,\n )\n if parts is None:\n continue\n sequence_row = len(user_ids)\n user_ids.append(uid)\n timestamp_parts.append(parts[0].astype(np.uint32, copy=False))\n item_parts.append(parts[1])\n event_parts.append(parts[2])\n activity_parts.append(parts[3])\n state_parts.append(parts[4])\n offsets.append(offsets[-1] + len(parts[1]))\n inference_row = inference_rows.get(uid)\n if inference_row is not None:\n inference_sequence_rows[inference_row] = sequence_row\n inference_user_features[inference_row] = final_user_features(\n parts[0],\n parts[1],\n parts[2],\n history_end,\n int(like_counts[inference_row]),\n int(dislike_counts[inference_row]),\n )\n if seen_users % 1_000 == 0:\n log(\n f\"prepared users={seen_users:,} tokens={offsets[-1]:,} \"\n f\"states={len(liked_states):,} \"\n f\"elapsed={time.monotonic() - started:.1f}s\"\n )\n\n missing = np.flatnonzero(inference_sequence_rows < 0)\n if len(missing):\n inference_user_features[missing, 4] = np.log1p(like_counts[missing]) / 6.0\n inference_user_features[missing, 5] = (\n np.log1p(dislike_counts[missing]) / 4.0\n )\n log(f\"target users without qualifying catalog sequence={len(missing):,}\")\n\n prepared = SequenceData(\n item_ids=item_ids,\n item_features=base.build_item_features(catalog),\n user_ids=np.asarray(user_ids, dtype=np.int64),\n offsets=np.asarray(offsets, dtype=np.int64),\n sequence_timestamps=np.concatenate(timestamp_parts),\n sequence_items=np.concatenate(item_parts),\n event_features=np.concatenate(event_parts),\n activity_features=np.concatenate(activity_parts),\n state_ids=np.concatenate(state_parts),\n liked_states=np.stack(liked_states),\n disliked_states=np.stack(disliked_states),\n state_counts=np.asarray(state_counts, dtype=np.int32),\n inference_user_rows=inference_sequence_rows,\n inference_liked_items=inference_likes,\n inference_disliked_items=inference_dislikes,\n inference_user_features=inference_user_features,\n )\n log(\n f\"prepared stage={stage} sequence_users={len(prepared.user_ids):,} \"\n f\"tokens={len(prepared.sequence_items):,} \"\n f\"next_item_targets={prepared.target_count:,} \"\n f\"states={len(prepared.liked_states):,}\"\n )\n return prepared, eligibility\n\n\ndef cache_metadata(stage: str, config: Config) -> dict[str, object]:\n return {\n \"cache_version\": CACHE_VERSION,\n \"stage\": stage,\n \"positive_listen_threshold\": base.LISTEN_POSITIVE_THRESHOLD,\n \"profile_lookback_days\": base.PROFILE_LOOKBACK_DAYS,\n \"event_time_feature\": \"exp-negative-gap-over-60-days\",\n }\n\n\ndef save_cache(\n path: Path, prepared: SequenceData, stage: str, config: Config\n) -> None:\n path.parent.mkdir(parents=True, exist_ok=True)\n np.savez(\n path,\n metadata=np.asarray(json.dumps(cache_metadata(stage, config), sort_keys=True)),\n **prepared.__dict__,\n )\n log(f\"saved public-derived full-sequence cache: {path}\")\n\n\ndef load_cache(path: Path, stage: str, config: Config) -> SequenceData:\n with np.load(path, allow_pickle=False) as artifact:\n actual = json.loads(str(artifact[\"metadata\"].item()))\n expected = cache_metadata(stage, config)\n if actual != expected:\n raise RuntimeError(f\"sequence cache differs: {actual} != {expected}\")\n values = {\n field: artifact[field].copy()\n for field in SequenceData.__dataclass_fields__\n }\n prepared = SequenceData(**values)\n log(\n f\"loaded full-sequence cache: tokens={len(prepared.sequence_items):,} \"\n f\"targets={prepared.target_count:,}\"\n )\n return prepared\n\n\ndef load_or_prepare(\n public: Path,\n stage: str,\n config: Config,\n cache: Path | None,\n) -> tuple[SequenceData, CandidateSet]:\n candidate_name = (\n \"validation_eligible_candidates.npz\"\n if stage == \"validation\"\n else \"test_eligible_candidates.npz\"\n )\n eligibility = CandidateSet.read(public / \"data\" / candidate_name)\n if cache is not None and cache.is_file():\n prepared = load_cache(cache, stage, config)\n if not np.array_equal(prepared.item_ids, eligibility.item_ids):\n raise RuntimeError(\"cached catalog differs from candidate artifact\")\n return prepared, eligibility\n prepared, built_eligibility = prepare_data(public, stage, config)\n if cache is not None:\n save_cache(cache, prepared, stage, config)\n return prepared, built_eligibility\n\n\ndef build_item_correction(\n prepared: SequenceData,\n config: Config,\n target_starts: np.ndarray,\n target_counts: np.ndarray,\n) -> torch.Tensor:\n \"\"\"Estimate in-batch propensities from the selected training targets.\"\"\"\n counts = np.zeros(len(prepared.item_ids), dtype=np.float64)\n for user_row, count in enumerate(target_counts):\n if not count:\n continue\n offset = int(prepared.offsets[user_row])\n target_start = int(target_starts[user_row])\n counts += np.bincount(\n prepared.sequence_items[\n offset\n + target_start\n + 1 : offset\n + target_start\n + int(count)\n + 1\n ].astype(np.int64),\n minlength=len(prepared.item_ids),\n )\n total = float(counts.sum())\n probabilities = (counts + 1.0) / (total + len(counts))\n raw = np.power(\n (1.0 / len(counts)) / probabilities,\n config.inbatch_correction_power,\n )\n normalization = float(np.sum(counts * raw) / max(total, 1.0))\n correction = np.clip(\n raw / max(normalization, 1.0e-12),\n config.inbatch_correction_min,\n config.inbatch_correction_max,\n ).astype(np.float32)\n mean_on_targets = float(np.sum(counts * correction) / max(total, 1.0))\n log(\n f\"inbatch correction: min={correction.min():.3f} \"\n f\"mean_on_targets={mean_on_targets:.3f} max={correction.max():.3f}\"\n )\n return torch.from_numpy(correction)\n\n\ndef build_known_positive_filter(\n prepared: SequenceData, training_counts: np.ndarray\n) -> np.ndarray:\n \"\"\"Pack each user's complete public-training positive-item set.\"\"\"\n width = (len(prepared.item_ids) + 7) // 8\n packed = np.zeros((len(prepared.user_ids), width), dtype=np.uint8)\n retained = 0\n for user_row, target_count in enumerate(training_counts):\n if not target_count:\n continue\n offset = int(prepared.offsets[user_row])\n # T adjacent targets consume sequence positions 0 through T. The\n # chronological BCE suffix starts at T + 1 and is intentionally absent.\n items = np.unique(\n prepared.sequence_items[\n offset : offset + int(target_count) + 1\n ].astype(np.int64)\n )\n byte_columns = items >> 3\n bit_values = np.left_shift(\n np.uint8(1), (items & 7).astype(np.uint8)\n )\n np.bitwise_or.at(packed[user_row], byte_columns, bit_values)",
"isError": false
},
"responseMessageId": "0af9a5d6-1016-45cc-a6a0-682d7964cb31"
}
]
},
{
"id": 11,
"role": "agent",
"timestamp": "2026-09-29T05:43:41.316Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:43:41.316Z",
"callId": "read_file__call_81033",
"name": "read_file",
"input": {
"file_path": "/task/current_model/train_current_model.py",
"start_line": 1900,
"end_line": 2165
},
"recorded": {
"id": "read_file__call_81033",
"name": "read_file",
"args": {
"file_path": "/task/current_model/train_current_model.py",
"start_line": 1900,
"end_line": 2165
},
"result": [
{
"functionResponse": {
"id": "read_file__call_81033",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1900-2165 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 2166.\n\n--- FILE CONTENT (truncated) ---\n log(f\"ranked users={end:,}/{len(eligibility.uids):,}\")\n query_norm = torch.linalg.vector_norm(queries, dim=1).numpy()\n item_norm = torch.linalg.vector_norm(items, dim=1).numpy()\n diagnostics = {\n \"query_norm_mean\": float(query_norm.mean()),\n \"query_norm_std\": float(query_norm.std()),\n \"item_norm_mean\": float(item_norm.mean()),\n \"item_norm_std\": float(item_norm.std()),\n }\n return pd.DataFrame(rows, columns=[\"uid\", \"item_id\", \"rank\"]), diagnostics\n\n\ndef write_ranking(frame: pd.DataFrame, output: Path) -> None:\n output.parent.mkdir(parents=True, exist_ok=True)\n if output.suffix == \".parquet\":\n frame.to_parquet(output, index=False, compression=\"zstd\")\n elif output.suffix == \".csv\":\n frame.to_csv(output, index=False)\n else:\n raise ValueError(\"--output must end in .parquet or .csv\")\n\n\ndef main() -> int:\n parser = argparse.ArgumentParser(description=__doc__)\n parser.add_argument(\n \"--stage\", choices=(\"validation\", \"submission\"), default=\"validation\"\n )\n parser.add_argument(\"--output\", type=Path, required=True)\n parser.add_argument(\"--public-dir\", type=Path, required=True)\n parser.add_argument(\"--prepared-cache\", type=Path)\n parser.add_argument(\"--device\", choices=(\"cpu\", \"mps\"), default=\"cpu\")\n parser.add_argument(\"--threads\", type=int, default=8)\n parser.add_argument(\"--bptt-steps\", type=int, default=Config.bptt_steps)\n parser.add_argument(\n \"--user-batch-size\", type=int, default=Config.user_batch_size\n )\n parser.add_argument(\"--epochs\", type=int, default=Config.epochs)\n parser.add_argument(\"--learning-rate\", type=float, default=Config.learning_rate)\n parser.add_argument(\n \"--loss-microbatch-size\",\n type=int,\n default=Config.loss_microbatch_size,\n )\n parser.add_argument(\n \"--negative-filter-scope\",\n choices=(\"full-training-history\", \"last-100\"),\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n )\n parser.add_argument(\n \"--ordinary-target-weight\",\n type=float,\n default=Config.ordinary_target_weight,\n )\n parser.add_argument(\n \"--organic-strong-target-weight\",\n type=float,\n default=Config.organic_strong_target_weight,\n )\n parser.add_argument(\n \"--organic-like-target-weight\",\n type=float,\n default=Config.organic_like_target_weight,\n )\n parser.add_argument(\n \"--discovery-horizon-days\",\n type=int,\n default=Config.discovery_horizon_days,\n )\n parser.add_argument(\n \"--validation-fraction\",\n type=float,\n default=Config.validation_fraction,\n )\n parser.add_argument(\n \"--max-train-targets-per-user\",\n type=int,\n default=Config.max_train_targets_per_user,\n help=\"retain this many recent pre-validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--max-validation-targets-per-user\",\n type=int,\n default=Config.max_validation_targets_per_user,\n help=\"retain this many final validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--early-stopping-patience\",\n type=int,\n default=Config.early_stopping_patience,\n )\n parser.add_argument(\n \"--early-stopping-min-delta\",\n type=float,\n default=Config.early_stopping_min_delta,\n )\n parser.add_argument(\"--checkpoint-dir\", type=Path)\n parser.add_argument(\"--resume-checkpoint\", type=Path)\n parser.add_argument(\"--seed\", type=int, default=Config.seed)\n parser.add_argument(\"--ranking-batch-size\", type=int, default=64)\n parser.add_argument(\"--prepare-only\", action=\"store_true\")\n args = parser.parse_args()\n\n if args.threads < 1 or args.bptt_steps < 2 or args.user_batch_size < 1:\n parser.error(\"threads, BPTT steps, and user batch size must be positive\")\n if (\n args.epochs < 1\n or args.learning_rate <= 0\n or args.loss_microbatch_size < 1\n or args.early_stopping_patience < 1\n or args.early_stopping_min_delta < 0\n or args.ordinary_target_weight <= 0\n or args.organic_strong_target_weight <= 0\n or args.organic_like_target_weight <= 0\n or args.discovery_horizon_days < 1\n or args.max_train_targets_per_user < 0\n or args.max_validation_targets_per_user < 0\n ):\n parser.error(\"training counts, learning rate, and patience are invalid\")\n if not 0.0 < args.validation_fraction < 1.0:\n parser.error(\"validation fraction must be strictly between zero and one\")\n public = args.public_dir.resolve()\n if not (public / \"data\" / \"candidate_catalog.parquet\").is_file():\n parser.error(f\"public task data not found under {public}\")\n config = Config(\n bptt_steps=args.bptt_steps,\n user_batch_size=args.user_batch_size,\n epochs=args.epochs,\n learning_rate=args.learning_rate,\n loss_microbatch_size=args.loss_microbatch_size,\n negative_filter_scope=args.negative_filter_scope,\n target_weighting=args.target_weighting,\n ordinary_target_weight=args.ordinary_target_weight,\n organic_strong_target_weight=args.organic_strong_target_weight,\n organic_like_target_weight=args.organic_like_target_weight,\n discovery_horizon_days=args.discovery_horizon_days,\n validation_fraction=args.validation_fraction,\n max_train_targets_per_user=args.max_train_targets_per_user,\n max_validation_targets_per_user=args.max_validation_targets_per_user,\n early_stopping_patience=args.early_stopping_patience,\n early_stopping_min_delta=args.early_stopping_min_delta,\n seed=args.seed,\n )\n torch.set_num_threads(args.threads)\n torch.set_num_interop_threads(1)\n seed_everything(config.seed)\n device = torch.device(args.device)\n if device.type == \"mps\" and not torch.backends.mps.is_available():\n parser.error(\"MPS was requested but is unavailable\")\n\n started = time.monotonic()\n prepared, eligibility = load_or_prepare(\n public,\n args.stage,\n config,\n args.prepared_cache.resolve() if args.prepared_cache else None,\n )\n target_partition = chronological_target_partition(\n prepared,\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n if args.prepare_only:\n print(\n json.dumps(\n {\n \"stage\": args.stage,\n \"sequence_users\": len(prepared.user_ids),\n \"sequence_tokens\": len(prepared.sequence_items),\n \"next_item_targets\": prepared.target_count,\n \"selected_training_targets\": int(\n target_partition.training_counts.sum()\n ),\n \"selected_validation_targets\": int(\n target_partition.validation_counts.sum()\n ),\n \"discarded_old_targets\": int(\n target_partition.window_starts.sum()\n ),\n \"target_users\": len(eligibility.uids),\n \"runtime_seconds\": round(time.monotonic() - started, 3),\n \"config\": asdict(config),\n },\n indent=2,\n )\n )\n return 0\n\n model, training_history = train_model(\n public,\n args.stage,\n prepared,\n config,\n device,\n args.checkpoint_dir.resolve() if args.checkpoint_dir else None,\n args.resume_checkpoint.resolve() if args.resume_checkpoint else None,\n )\n ranking, embedding_diagnostics = rank_candidates(\n model,\n prepared,\n eligibility,\n config,\n device,\n args.ranking_batch_size,\n )\n write_ranking(ranking, args.output.resolve())\n print(\n json.dumps(\n {\n \"baseline\": \"yambda-windowed-sequence-two-tower-base\",\n \"stage\": args.stage,\n \"output\": str(args.output.resolve()),\n \"rows\": len(ranking),\n \"users\": int(ranking.uid.nunique()),\n \"runtime_seconds\": round(time.monotonic() - started, 3),\n \"public_inputs_only\": True,\n \"architecture\": {\n \"family\": \"pure-windowed-sequence-two-tower\",\n \"sequence_training\": (\n \"every-adjacent-target-within-recent-chronological-window\"\n ),\n \"sequence_encoder\": \"one-layer-GRU-dense-timestep-loss\",\n \"bptt_hidden_state\": \"carried-and-detached\",\n \"model_selection\": \"chronological-validation-BCE-early-stopping\",\n \"loss_execution\": \"gradient-accumulated-microbatches\",\n \"user_id_embedding\": False,\n \"user_mlp\": \"582-512-256-192\",\n \"item_mlp\": \"196-512-256-192\",\n \"normalization\": \"none\",\n \"score\": \"raw-dot-product\",\n \"objective\": \"sample-weighted-BCE-at-every-timestep\",\n \"negative_sampling\": (\n \"24-uniform-plus-8-propensity-corrected-inbatch\"\n ),\n \"false_negative_filter\": config.negative_filter_scope,\n \"target_weighting\": config.target_weighting,\n },\n \"config\": asdict(config),\n \"sequence_tokens\": len(prepared.sequence_items),\n \"all_next_item_targets\": prepared.target_count,\n \"selected_training_targets\": int(\n target_partition.training_counts.sum()\n ),\n \"selected_validation_targets\": int(\n target_partition.validation_counts.sum()\n ),\n \"discarded_old_targets\": int(target_partition.window_starts.sum()),\n \"training_history\": training_history,\n \"embedding_diagnostics\": embedding_diagnostics,\n \"torch_version\": torch.__version__,\n \"device\": str(device),\n },\n indent=2,\n )\n )\n return 0\n\n\nif __name__ == \"__main__\":\n raise SystemExit(main())\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:43:41.321Z",
"resultDisplay": "Read lines 1900-2165 of 2165 from ../task/current_model/train_current_model.py",
"description": "../.../train_current_model.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:43:41.321Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1900-2165 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 2166.\n\n--- FILE CONTENT (truncated) ---\n log(f\"ranked users={end:,}/{len(eligibility.uids):,}\")\n query_norm = torch.linalg.vector_norm(queries, dim=1).numpy()\n item_norm = torch.linalg.vector_norm(items, dim=1).numpy()\n diagnostics = {\n \"query_norm_mean\": float(query_norm.mean()),\n \"query_norm_std\": float(query_norm.std()),\n \"item_norm_mean\": float(item_norm.mean()),\n \"item_norm_std\": float(item_norm.std()),\n }\n return pd.DataFrame(rows, columns=[\"uid\", \"item_id\", \"rank\"]), diagnostics\n\n\ndef write_ranking(frame: pd.DataFrame, output: Path) -> None:\n output.parent.mkdir(parents=True, exist_ok=True)\n if output.suffix == \".parquet\":\n frame.to_parquet(output, index=False, compression=\"zstd\")\n elif output.suffix == \".csv\":\n frame.to_csv(output, index=False)\n else:\n raise ValueError(\"--output must end in .parquet or .csv\")\n\n\ndef main() -> int:\n parser = argparse.ArgumentParser(description=__doc__)\n parser.add_argument(\n \"--stage\", choices=(\"validation\", \"submission\"), default=\"validation\"\n )\n parser.add_argument(\"--output\", type=Path, required=True)\n parser.add_argument(\"--public-dir\", type=Path, required=True)\n parser.add_argument(\"--prepared-cache\", type=Path)\n parser.add_argument(\"--device\", choices=(\"cpu\", \"mps\"), default=\"cpu\")\n parser.add_argument(\"--threads\", type=int, default=8)\n parser.add_argument(\"--bptt-steps\", type=int, default=Config.bptt_steps)\n parser.add_argument(\n \"--user-batch-size\", type=int, default=Config.user_batch_size\n )\n parser.add_argument(\"--epochs\", type=int, default=Config.epochs)\n parser.add_argument(\"--learning-rate\", type=float, default=Config.learning_rate)\n parser.add_argument(\n \"--loss-microbatch-size\",\n type=int,\n default=Config.loss_microbatch_size,\n )\n parser.add_argument(\n \"--negative-filter-scope\",\n choices=(\"full-training-history\", \"last-100\"),\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n )\n parser.add_argument(\n \"--ordinary-target-weight\",\n type=float,\n default=Config.ordinary_target_weight,\n )\n parser.add_argument(\n \"--organic-strong-target-weight\",\n type=float,\n default=Config.organic_strong_target_weight,\n )\n parser.add_argument(\n \"--organic-like-target-weight\",\n type=float,\n default=Config.organic_like_target_weight,\n )\n parser.add_argument(\n \"--discovery-horizon-days\",\n type=int,\n default=Config.discovery_horizon_days,\n )\n parser.add_argument(\n \"--validation-fraction\",\n type=float,\n default=Config.validation_fraction,\n )\n parser.add_argument(\n \"--max-train-targets-per-user\",\n type=int,\n default=Config.max_train_targets_per_user,\n help=\"retain this many recent pre-validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--max-validation-targets-per-user\",\n type=int,\n default=Config.max_validation_targets_per_user,\n help=\"retain this many final validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--early-stopping-patience\",\n type=int,\n default=Config.early_stopping_patience,\n )\n parser.add_argument(\n \"--early-stopping-min-delta\",\n type=float,\n default=Config.early_stopping_min_delta,\n )\n parser.add_argument(\"--checkpoint-dir\", type=Path)\n parser.add_argument(\"--resume-checkpoint\", type=Path)\n parser.add_argument(\"--seed\", type=int, default=Config.seed)\n parser.add_argument(\"--ranking-batch-size\", type=int, default=64)\n parser.add_argument(\"--prepare-only\", action=\"store_true\")\n args = parser.parse_args()\n\n if args.threads < 1 or args.bptt_steps < 2 or args.user_batch_size < 1:\n parser.error(\"threads, BPTT steps, and user batch size must be positive\")\n if (\n args.epochs < 1\n or args.learning_rate <= 0\n or args.loss_microbatch_size < 1\n or args.early_stopping_patience < 1\n or args.early_stopping_min_delta < 0\n or args.ordinary_target_weight <= 0\n or args.organic_strong_target_weight <= 0\n or args.organic_like_target_weight <= 0\n or args.discovery_horizon_days < 1\n or args.max_train_targets_per_user < 0\n or args.max_validation_targets_per_user < 0\n ):\n parser.error(\"training counts, learning rate, and patience are invalid\")\n if not 0.0 < args.validation_fraction < 1.0:\n parser.error(\"validation fraction must be strictly between zero and one\")\n public = args.public_dir.resolve()\n if not (public / \"data\" / \"candidate_catalog.parquet\").is_file():\n parser.error(f\"public task data not found under {public}\")\n config = Config(\n bptt_steps=args.bptt_steps,\n user_batch_size=args.user_batch_size,\n epochs=args.epochs,\n learning_rate=args.learning_rate,\n loss_microbatch_size=args.loss_microbatch_size,\n negative_filter_scope=args.negative_filter_scope,\n target_weighting=args.target_weighting,\n ordinary_target_weight=args.ordinary_target_weight,\n organic_strong_target_weight=args.organic_strong_target_weight,\n organic_like_target_weight=args.organic_like_target_weight,\n discovery_horizon_days=args.discovery_horizon_days,\n validation_fraction=args.validation_fraction,\n max_train_targets_per_user=args.max_train_targets_per_user,\n max_validation_targets_per_user=args.max_validation_targets_per_user,\n early_stopping_patience=args.early_stopping_patience,\n early_stopping_min_delta=args.early_stopping_min_delta,\n seed=args.seed,\n )\n torch.set_num_threads(args.threads)\n torch.set_num_interop_threads(1)\n seed_everything(config.seed)\n device = torch.device(args.device)\n if device.type == \"mps\" and not torch.backends.mps.is_available():\n parser.error(\"MPS was requested but is unavailable\")\n\n started = time.monotonic()\n prepared, eligibility = load_or_prepare(\n public,\n args.stage,\n config,\n args.prepared_cache.resolve() if args.prepared_cache else None,\n )\n target_partition = chronological_target_partition(\n prepared,\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n if args.prepare_only:\n print(\n json.dumps(\n {\n \"stage\": args.stage,\n \"sequence_users\": len(prepared.user_ids),\n \"sequence_tokens\": len(prepared.sequence_items),\n \"next_item_targets\": prepared.target_count,\n \"selected_training_targets\": int(\n target_partition.training_counts.sum()\n ),\n \"selected_validation_targets\": int(\n target_partition.validation_counts.sum()\n ),\n \"discarded_old_targets\": int(\n target_partition.window_starts.sum()\n ),\n \"target_users\": len(eligibility.uids),\n \"runtime_seconds\": round(time.monotonic() - started, 3),\n \"config\": asdict(config),\n },\n indent=2,\n )\n )\n return 0\n\n model, training_history = train_model(\n public,\n args.stage,\n prepared,\n config,\n device,\n args.checkpoint_dir.resolve() if args.checkpoint_dir else None,\n args.resume_checkpoint.resolve() if args.resume_checkpoint else None,\n )\n ranking, embedding_diagnostics = rank_candidates(\n model,\n prepared,\n eligibility,\n config,\n device,\n args.ranking_batch_size,\n )\n write_ranking(ranking, args.output.resolve())\n print(\n json.dumps(\n {\n \"baseline\": \"yambda-windowed-sequence-two-tower-base\",\n \"stage\": args.stage,\n \"output\": str(args.output.resolve()),\n \"rows\": len(ranking),\n \"users\": int(ranking.uid.nunique()),\n \"runtime_seconds\": round(time.monotonic() - started, 3),\n \"public_inputs_only\": True,\n \"architecture\": {\n \"family\": \"pure-windowed-sequence-two-tower\",\n \"sequence_training\": (\n \"every-adjacent-target-within-recent-chronological-window\"\n ),\n \"sequence_encoder\": \"one-layer-GRU-dense-timestep-loss\",\n \"bptt_hidden_state\": \"carried-and-detached\",\n \"model_selection\": \"chronological-validation-BCE-early-stopping\",\n \"loss_execution\": \"gradient-accumulated-microbatches\",\n \"user_id_embedding\": False,\n \"user_mlp\": \"582-512-256-192\",\n \"item_mlp\": \"196-512-256-192\",\n \"normalization\": \"none\",\n \"score\": \"raw-dot-product\",\n \"objective\": \"sample-weighted-BCE-at-every-timestep\",\n \"negative_sampling\": (\n \"24-uniform-plus-8-propensity-corrected-inbatch\"\n ),\n \"false_negative_filter\": config.negative_filter_scope,\n \"target_weighting\": config.target_weighting,\n },\n \"config\": asdict(config),\n \"sequence_tokens\": len(prepared.sequence_items),\n \"all_next_item_targets\": prepared.target_count,\n \"selected_training_targets\": int(\n target_partition.training_counts.sum()\n ),\n \"selected_validation_targets\": int(\n target_partition.validation_counts.sum()\n ),\n \"discarded_old_targets\": int(target_partition.window_starts.sum()),\n \"training_history\": training_history,\n \"embedding_diagnostics\": embedding_diagnostics,\n \"torch_version\": torch.__version__,\n \"device\": str(device),\n },\n indent=2,\n )\n )\n return 0\n\n\nif __name__ == \"__main__\":\n raise SystemExit(main())\n",
"isError": false
},
"responseMessageId": "a0f79bde-bb08-4de5-b8f1-4d09fe9faab8"
}
]
},
{
"id": 12,
"role": "agent",
"timestamp": "2026-09-29T05:43:45.055Z",
"blocks": [
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:43:45.055Z",
"text": "Exploring Cache Optimization\n\nI've noticed the `--prepare-only` and `--prepared-cache` flags and how `load_or_prepare` now saves the cache with the latter. My current focus is on determining if we can pre-generate this prepared cache for validation and submission. The goal is to shave off roughly 50 seconds of data preparation per training run by avoiding redundant preparation."
},
{
"kind": "tool",
"timestamp": "2026-09-29T05:43:45.055Z",
"callId": "grep_search__call_28083",
"name": "grep_search",
"input": {
"dir_path": "/task/current_model",
"pattern": "negative"
},
"recorded": {
"id": "grep_search__call_28083",
"name": "grep_search",
"args": {
"dir_path": "/task/current_model",
"pattern": "negative"
},
"result": [
{
"functionResponse": {
"id": "grep_search__call_28083",
"name": "grep_search",
"response": {
"output": "Found 100 matches for pattern \"negative\" in path \"/task/current_model\" (results limited to 100 matches for performance):\n---\nFile: train_current_model.py\nL14: and 24 uniform plus eight propensity-corrected unique in-batch negatives. Every\nL56: uniform_negatives: int = 24\nL57: inbatch_negatives: int = 8\nL61: negative_filter_scope: str = \"full-training-history\"\nL165: uniform_negatives=config.uniform_negatives,\nL166: inbatch_negatives=config.inbatch_negatives,\nL477: \"event_time_feature\": \"exp-negative-gap-over-60-days\",\nL603: f\"full-history false-negative filter: users={len(prepared.user_ids):,} \"\nL609: def invalid_full_history_negatives(\nL610: negatives: np.ndarray,\nL616: \"\"\"Return row-level invalid negatives using the complete training prefix.\"\"\"\nL617: byte_columns = negatives >> 3\nL618: bit_columns = negatives & 7\nL621: invalid |= negatives == targets[:, None]\nL622: invalid |= (negatives[:, :, None] == liked_items[:, None, :]).any(axis=2)\nL623: for column in range(1, negatives.shape[1]):\nL625: negatives[:, column, None] == negatives[:, :column]\nL630: def draw_full_history_filtered_negatives(\nL641: \"\"\"Draw mixed negatives and reject all known public-training positives.\"\"\"\nL647: (batch_size, config.uniform_negatives),\nL653: if config.inbatch_negatives:\nL656: config.inbatch_negatives, max(len(unique_targets) - 1, 0)\nL672: if corrected_count < config.inbatch_negatives:\nL675: (batch_size, config.inbatch_negatives - corrected_count),\nL681: negatives = torch.cat([uniform, inbatch], dim=1).numpy()\nL682: correction = np.ones_like(negatives, dtype=np.float32)\nL685: :, config.uniform_negatives : config.uniform_negatives + corrected_count\nL688: invalid = invalid_full_history_negatives(\nL689: negatives,\nL697: torch.from_numpy(negatives).to(device),\nL706: negatives[invalid] = replacements\nL709: invalid = invalid_full_history_negatives(\nL710: negatives,\nL717: raise RuntimeError(\"full-history negative resampling failed\")\nL719: torch.from_numpy(negatives).to(device),\nL724: def draw_training_negatives(\nL736: if config.negative_filter_scope == \"full-training-history\":\nL738: raise RuntimeError(\"full-history negative filter was not built\")\nL739: return draw_full_history_filtered_negatives(\nL750: return base.draw_filtered_mixed_negatives(\nL755: config.uniform_negatives,\nL756: config.inbatch_negatives,\nL816: negatives: torch.Tensor,\nL818: all_items = torch.cat([positives, negatives.reshape(-1)])\nL822: negative_vectors = encoded[inverse[len(positives) :]].reshape(\nL823: len(positives), negatives.shape[1], -1\nL825: return positive_vectors, negative_vectors\nL845: def corrected_negative_loss(\nL851: total = config.uniform_negatives + config.inbatch_negatives\nL853: if config.uniform_negatives:\nL854: result += config.uniform_negatives * raw[\nL855: :, : config.uniform_negatives\nL857: if config.inbatch_negatives:\nL858: selected = slice(config.uniform_negatives, total)\nL860: result += config.inbatch_negatives * (\nL880: raise RuntimeError(\"chronological target partition produced negative counts\")\nL1119: negatives: torch.Tensor,\nL1120: negative_weights: torch.Tensor,\nL1134: \"negative_sum\": 0.0,\nL1162: positive_vectors, negative_vectors = candidate_vectors(\nL1165: negatives[begin:end],\nL1168: negative_logits = torch.einsum(\nL1169: \"bd,bkd->bk\", query_vectors, negative_vectors\nL1172: negative_loss = corrected_negative_loss(\nL1173: negative_logits, negative_weights[begin:end], config\nL1179: selected_weights * (positive_loss + negative_loss) / 2.0\nL1190: statistics[\"negative_sum\"] += float(negative_logits.detach().sum().cpu())\nL1192: (positive_logits.detach()[:, None] > negative_logits.detach()).sum().cpu()\nL1227: negative_count = config.uniform_negatives + config.inbatch_negatives\nL1232: total_negative = 0.0\nL1293: negatives, negative_weights = draw_training_negatives(\nL1314: negatives,\nL1315: negative_weights,\nL1331: total_negative += statistics[\"negative_sum\"]\nL1353: \"negative_logit\": total_negative / max(total_targets * negative_count, 1),\nL1355: / max(total_targets * negative_count, 1),\nL1449: negatives, negative_weights = draw_training_negatives(\nL1469: negatives,\nL1470: negative_weights,\nL1567: if config.negative_filter_scope == \"full-training-history\"\nL1601: \"negative_filter_scope\", \"last-100\"\nL1629: stored_filter_scope != config.negative_filter_scope\nL1660: f\"{stored_filter_scope}->{config.negative_filter_scope}, \"\nL1944: \"--negative-filter-scope\",\nL1946: default=Config.negative_filter_scope,\nL2034: negative_filter_scope=args.negative_filter_scope,\nL2136: \"negative_sampling\": (\nL2139: \"false_negative_filter\": config.negative_filter_scope,\n---\nFile: two_tower_model.py\nL40: uniform_negatives: int = 24\nL41: inbatch_negatives: int = 8\nL346: def invalid_negative_mask(\nL347: negatives: torch.Tensor,\nL353: invalid = negatives.eq(targets[:, None])\nL354: invalid |= negatives[:, :, None].eq(context_items[:, None, :]).any(dim=2)\nL355: invalid |= negatives[:, :, None].eq(liked_items[:, None, :]).any(dim=2)\nL356: for column in range(1, negatives.shape[1]):\nL357: invalid[:, column] |= negatives[:, column, None].eq(\nL358: negatives[:, :column]\nL363: def draw_filtered_mixed_negatives(\nL374: \"\"\"Draw uniform and unique-target in-batch negatives.\"\"\"\n---"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:43:45.077Z",
"resultDisplay": {
"summary": "Found 100 matches (limited)",
"matches": [
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 14,
"line": "and 24 uniform plus eight propensity-corrected unique in-batch negatives. Every"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 56,
"line": " uniform_negatives: int = 24"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 57,
"line": " inbatch_negatives: int = 8"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 61,
"line": " negative_filter_scope: str = \"full-training-history\""
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 165,
"line": " uniform_negatives=config.uniform_negatives,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 166,
"line": " inbatch_negatives=config.inbatch_negatives,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 477,
"line": " \"event_time_feature\": \"exp-negative-gap-over-60-days\","
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 603,
"line": " f\"full-history false-negative filter: users={len(prepared.user_ids):,} \""
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 609,
"line": "def invalid_full_history_negatives("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 610,
"line": " negatives: np.ndarray,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 616,
"line": " \"\"\"Return row-level invalid negatives using the complete training prefix.\"\"\""
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 617,
"line": " byte_columns = negatives >> 3"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 618,
"line": " bit_columns = negatives & 7"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 621,
"line": " invalid |= negatives == targets[:, None]"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 622,
"line": " invalid |= (negatives[:, :, None] == liked_items[:, None, :]).any(axis=2)"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 623,
"line": " for column in range(1, negatives.shape[1]):"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 625,
"line": " negatives[:, column, None] == negatives[:, :column]"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 630,
"line": "def draw_full_history_filtered_negatives("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 641,
"line": " \"\"\"Draw mixed negatives and reject all known public-training positives.\"\"\""
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 647,
"line": " (batch_size, config.uniform_negatives),"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 653,
"line": " if config.inbatch_negatives:"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 656,
"line": " config.inbatch_negatives, max(len(unique_targets) - 1, 0)"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 672,
"line": " if corrected_count < config.inbatch_negatives:"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 675,
"line": " (batch_size, config.inbatch_negatives - corrected_count),"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 681,
"line": " negatives = torch.cat([uniform, inbatch], dim=1).numpy()"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 682,
"line": " correction = np.ones_like(negatives, dtype=np.float32)"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 685,
"line": " :, config.uniform_negatives : config.uniform_negatives + corrected_count"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 688,
"line": " invalid = invalid_full_history_negatives("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 689,
"line": " negatives,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 697,
"line": " torch.from_numpy(negatives).to(device),"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 706,
"line": " negatives[invalid] = replacements"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 709,
"line": " invalid = invalid_full_history_negatives("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 710,
"line": " negatives,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 717,
"line": " raise RuntimeError(\"full-history negative resampling failed\")"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 719,
"line": " torch.from_numpy(negatives).to(device),"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 724,
"line": "def draw_training_negatives("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 736,
"line": " if config.negative_filter_scope == \"full-training-history\":"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 738,
"line": " raise RuntimeError(\"full-history negative filter was not built\")"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 739,
"line": " return draw_full_history_filtered_negatives("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 750,
"line": " return base.draw_filtered_mixed_negatives("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 755,
"line": " config.uniform_negatives,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 756,
"line": " config.inbatch_negatives,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 816,
"line": " negatives: torch.Tensor,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 818,
"line": " all_items = torch.cat([positives, negatives.reshape(-1)])"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 822,
"line": " negative_vectors = encoded[inverse[len(positives) :]].reshape("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 823,
"line": " len(positives), negatives.shape[1], -1"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 825,
"line": " return positive_vectors, negative_vectors"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 845,
"line": "def corrected_negative_loss("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 851,
"line": " total = config.uniform_negatives + config.inbatch_negatives"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 853,
"line": " if config.uniform_negatives:"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 854,
"line": " result += config.uniform_negatives * raw["
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 855,
"line": " :, : config.uniform_negatives"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 857,
"line": " if config.inbatch_negatives:"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 858,
"line": " selected = slice(config.uniform_negatives, total)"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 860,
"line": " result += config.inbatch_negatives * ("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 880,
"line": " raise RuntimeError(\"chronological target partition produced negative counts\")"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1119,
"line": " negatives: torch.Tensor,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1120,
"line": " negative_weights: torch.Tensor,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1134,
"line": " \"negative_sum\": 0.0,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1162,
"line": " positive_vectors, negative_vectors = candidate_vectors("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1165,
"line": " negatives[begin:end],"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1168,
"line": " negative_logits = torch.einsum("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1169,
"line": " \"bd,bkd->bk\", query_vectors, negative_vectors"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1172,
"line": " negative_loss = corrected_negative_loss("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1173,
"line": " negative_logits, negative_weights[begin:end], config"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1179,
"line": " selected_weights * (positive_loss + negative_loss) / 2.0"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1190,
"line": " statistics[\"negative_sum\"] += float(negative_logits.detach().sum().cpu())"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1192,
"line": " (positive_logits.detach()[:, None] > negative_logits.detach()).sum().cpu()"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1227,
"line": " negative_count = config.uniform_negatives + config.inbatch_negatives"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1232,
"line": " total_negative = 0.0"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1293,
"line": " negatives, negative_weights = draw_training_negatives("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1314,
"line": " negatives,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1315,
"line": " negative_weights,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1331,
"line": " total_negative += statistics[\"negative_sum\"]"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1353,
"line": " \"negative_logit\": total_negative / max(total_targets * negative_count, 1),"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1355,
"line": " / max(total_targets * negative_count, 1),"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1449,
"line": " negatives, negative_weights = draw_training_negatives("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1469,
"line": " negatives,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1470,
"line": " negative_weights,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1567,
"line": " if config.negative_filter_scope == \"full-training-history\""
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1601,
"line": " \"negative_filter_scope\", \"last-100\""
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1629,
"line": " stored_filter_scope != config.negative_filter_scope"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1660,
"line": " f\"{stored_filter_scope}->{config.negative_filter_scope}, \""
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1944,
"line": " \"--negative-filter-scope\","
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1946,
"line": " default=Config.negative_filter_scope,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 2034,
"line": " negative_filter_scope=args.negative_filter_scope,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 2136,
"line": " \"negative_sampling\": ("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 2139,
"line": " \"false_negative_filter\": config.negative_filter_scope,"
},
{
"filePath": "two_tower_model.py",
"absolutePath": "/task/current_model/two_tower_model.py",
"lineNumber": 40,
"line": " uniform_negatives: int = 24"
},
{
"filePath": "two_tower_model.py",
"absolutePath": "/task/current_model/two_tower_model.py",
"lineNumber": 41,
"line": " inbatch_negatives: int = 8"
},
{
"filePath": "two_tower_model.py",
"absolutePath": "/task/current_model/two_tower_model.py",
"lineNumber": 346,
"line": "def invalid_negative_mask("
},
{
"filePath": "two_tower_model.py",
"absolutePath": "/task/current_model/two_tower_model.py",
"lineNumber": 347,
"line": " negatives: torch.Tensor,"
},
{
"filePath": "two_tower_model.py",
"absolutePath": "/task/current_model/two_tower_model.py",
"lineNumber": 353,
"line": " invalid = negatives.eq(targets[:, None])"
},
{
"filePath": "two_tower_model.py",
"absolutePath": "/task/current_model/two_tower_model.py",
"lineNumber": 354,
"line": " invalid |= negatives[:, :, None].eq(context_items[:, None, :]).any(dim=2)"
},
{
"filePath": "two_tower_model.py",
"absolutePath": "/task/current_model/two_tower_model.py",
"lineNumber": 355,
"line": " invalid |= negatives[:, :, None].eq(liked_items[:, None, :]).any(dim=2)"
},
{
"filePath": "two_tower_model.py",
"absolutePath": "/task/current_model/two_tower_model.py",
"lineNumber": 356,
"line": " for column in range(1, negatives.shape[1]):"
},
{
"filePath": "two_tower_model.py",
"absolutePath": "/task/current_model/two_tower_model.py",
"lineNumber": 357,
"line": " invalid[:, column] |= negatives[:, column, None].eq("
},
{
"filePath": "two_tower_model.py",
"absolutePath": "/task/current_model/two_tower_model.py",
"lineNumber": 358,
"line": " negatives[:, :column]"
},
{
"filePath": "two_tower_model.py",
"absolutePath": "/task/current_model/two_tower_model.py",
"lineNumber": 363,
"line": "def draw_filtered_mixed_negatives("
},
{
"filePath": "two_tower_model.py",
"absolutePath": "/task/current_model/two_tower_model.py",
"lineNumber": 374,
"line": " \"\"\"Draw uniform and unique-target in-batch negatives.\"\"\""
}
]
},
"description": "'negative' within ../task/current_model",
"displayName": "SearchText",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:43:45.079Z",
"content": "Found 100 matches for pattern \"negative\" in path \"/task/current_model\" (results limited to 100 matches for performance):\n---\nFile: train_current_model.py\nL14: and 24 uniform plus eight propensity-corrected unique in-batch negatives. Every\nL56: uniform_negatives: int = 24\nL57: inbatch_negatives: int = 8\nL61: negative_filter_scope: str = \"full-training-history\"\nL165: uniform_negatives=config.uniform_negatives,\nL166: inbatch_negatives=config.inbatch_negatives,\nL477: \"event_time_feature\": \"exp-negative-gap-over-60-days\",\nL603: f\"full-history false-negative filter: users={len(prepared.user_ids):,} \"\nL609: def invalid_full_history_negatives(\nL610: negatives: np.ndarray,\nL616: \"\"\"Return row-level invalid negatives using the complete training prefix.\"\"\"\nL617: byte_columns = negatives >> 3\nL618: bit_columns = negatives & 7\nL621: invalid |= negatives == targets[:, None]\nL622: invalid |= (negatives[:, :, None] == liked_items[:, None, :]).any(axis=2)\nL623: for column in range(1, negatives.shape[1]):\nL625: negatives[:, column, None] == negatives[:, :column]\nL630: def draw_full_history_filtered_negatives(\nL641: \"\"\"Draw mixed negatives and reject all known public-training positives.\"\"\"\nL647: (batch_size, config.uniform_negatives),\nL653: if config.inbatch_negatives:\nL656: config.inbatch_negatives, max(len(unique_targets) - 1, 0)\nL672: if corrected_count < config.inbatch_negatives:\nL675: (batch_size, config.inbatch_negatives - corrected_count),\nL681: negatives = torch.cat([uniform, inbatch], dim=1).numpy()\nL682: correction = np.ones_like(negatives, dtype=np.float32)\nL685: :, config.uniform_negatives : config.uniform_negatives + corrected_count\nL688: invalid = invalid_full_history_negatives(\nL689: negatives,\nL697: torch.from_numpy(negatives).to(device),\nL706: negatives[invalid] = replacements\nL709: invalid = invalid_full_history_negatives(\nL710: negatives,\nL717: raise RuntimeError(\"full-history negative resampling failed\")\nL719: torch.from_numpy(negatives).to(device),\nL724: def draw_training_negatives(\nL736: if config.negative_filter_scope == \"full-training-history\":\nL738: raise RuntimeError(\"full-history negative filter was not built\")\nL739: return draw_full_history_filtered_negatives(\nL750: return base.draw_filtered_mixed_negatives(\nL755: config.uniform_negatives,\nL756: config.inbatch_negatives,\nL816: negatives: torch.Tensor,\nL818: all_items = torch.cat([positives, negatives.reshape(-1)])\nL822: negative_vectors = encoded[inverse[len(positives) :]].reshape(\nL823: len(positives), negatives.shape[1], -1\nL825: return positive_vectors, negative_vectors\nL845: def corrected_negative_loss(\nL851: total = config.uniform_negatives + config.inbatch_negatives\nL853: if config.uniform_negatives:\nL854: result += config.uniform_negatives * raw[\nL855: :, : config.uniform_negatives\nL857: if config.inbatch_negatives:\nL858: selected = slice(config.uniform_negatives, total)\nL860: result += config.inbatch_negatives * (\nL880: raise RuntimeError(\"chronological target partition produced negative counts\")\nL1119: negatives: torch.Tensor,\nL1120: negative_weights: torch.Tensor,\nL1134: \"negative_sum\": 0.0,\nL1162: positive_vectors, negative_vectors = candidate_vectors(\nL1165: negatives[begin:end],\nL1168: negative_logits = torch.einsum(\nL1169: \"bd,bkd->bk\", query_vectors, negative_vectors\nL1172: negative_loss = corrected_negative_loss(\nL1173: negative_logits, negative_weights[begin:end], config\nL1179: selected_weights * (positive_loss + negative_loss) / 2.0\nL1190: statistics[\"negative_sum\"] += float(negative_logits.detach().sum().cpu())\nL1192: (positive_logits.detach()[:, None] > negative_logits.detach()).sum().cpu()\nL1227: negative_count = config.uniform_negatives + config.inbatch_negatives\nL1232: total_negative = 0.0\nL1293: negatives, negative_weights = draw_training_negatives(\nL1314: negatives,\nL1315: negative_weights,\nL1331: total_negative += statistics[\"negative_sum\"]\nL1353: \"negative_logit\": total_negative / max(total_targets * negative_count, 1),\nL1355: / max(total_targets * negative_count, 1),\nL1449: negatives, negative_weights = draw_training_negatives(\nL1469: negatives,\nL1470: negative_weights,\nL1567: if config.negative_filter_scope == \"full-training-history\"\nL1601: \"negative_filter_scope\", \"last-100\"\nL1629: stored_filter_scope != config.negative_filter_scope\nL1660: f\"{stored_filter_scope}->{config.negative_filter_scope}, \"\nL1944: \"--negative-filter-scope\",\nL1946: default=Config.negative_filter_scope,\nL2034: negative_filter_scope=args.negative_filter_scope,\nL2136: \"negative_sampling\": (\nL2139: \"false_negative_filter\": config.negative_filter_scope,\n---\nFile: two_tower_model.py\nL40: uniform_negatives: int = 24\nL41: inbatch_negatives: int = 8\nL346: def invalid_negative_mask(\nL347: negatives: torch.Tensor,\nL353: invalid = negatives.eq(targets[:, None])\nL354: invalid |= negatives[:, :, None].eq(context_items[:, None, :]).any(dim=2)\nL355: invalid |= negatives[:, :, None].eq(liked_items[:, None, :]).any(dim=2)\nL356: for column in range(1, negatives.shape[1]):\nL357: invalid[:, column] |= negatives[:, column, None].eq(\nL358: negatives[:, :column]\nL363: def draw_filtered_mixed_negatives(\nL374: \"\"\"Draw uniform and unique-target in-batch negatives.\"\"\"\n---",
"isError": false
},
"responseMessageId": "89630db7-461e-4607-9da1-03bd90f002f3"
}
]
},
{
"id": 13,
"role": "agent",
"timestamp": "2026-09-29T05:43:52.912Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:43:52.912Z",
"callId": "read_file__call_5963777",
"name": "read_file",
"input": {
"start_line": 605,
"end_line": 770,
"file_path": "/task/current_model/train_current_model.py"
},
"recorded": {
"id": "read_file__call_5963777",
"name": "read_file",
"args": {
"start_line": 605,
"end_line": 770,
"file_path": "/task/current_model/train_current_model.py"
},
"result": [
{
"functionResponse": {
"id": "read_file__call_5963777",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 605-770 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 771.\n\n--- FILE CONTENT (truncated) ---\n )\n return packed\n\n\ndef invalid_full_history_negatives(\n negatives: np.ndarray,\n targets: np.ndarray,\n user_rows: np.ndarray,\n known_positive_packed: np.ndarray,\n liked_items: np.ndarray,\n) -> np.ndarray:\n \"\"\"Return row-level invalid negatives using the complete training prefix.\"\"\"\n byte_columns = negatives >> 3\n bit_columns = negatives & 7\n known = known_positive_packed[user_rows[:, None], byte_columns]\n invalid = ((known >> bit_columns) & 1).astype(bool)\n invalid |= negatives == targets[:, None]\n invalid |= (negatives[:, :, None] == liked_items[:, None, :]).any(axis=2)\n for column in range(1, negatives.shape[1]):\n invalid[:, column] |= (\n negatives[:, column, None] == negatives[:, :column]\n ).any(axis=1)\n return invalid\n\n\ndef draw_full_history_filtered_negatives(\n targets: torch.Tensor,\n user_rows: np.ndarray,\n liked_items: torch.Tensor,\n known_positive_packed: np.ndarray,\n n_items: int,\n config: Config,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n \"\"\"Draw mixed negatives and reject all known public-training positives.\"\"\"\n targets_cpu = targets.detach().cpu().numpy().astype(np.int64, copy=False)\n liked_cpu = liked_items.detach().cpu().numpy().astype(np.int64, copy=False)\n batch_size = len(targets_cpu)\n uniform = torch.randint(\n n_items,\n (batch_size, config.uniform_negatives),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.empty((batch_size, 0), dtype=torch.long)\n corrected_count = 0\n if config.inbatch_negatives:\n unique_targets = torch.unique(targets.detach().cpu())\n corrected_count = min(\n config.inbatch_negatives, max(len(unique_targets) - 1, 0)\n )\n if corrected_count:\n priorities = torch.rand(\n (batch_size, len(unique_targets)), generator=generator\n )\n priorities.masked_fill_(\n unique_targets[None, :].eq(\n torch.from_numpy(targets_cpu)[:, None]\n ),\n -1.0,\n )\n columns = torch.topk(\n priorities, k=corrected_count, dim=1\n ).indices\n inbatch = unique_targets[columns]\n if corrected_count < config.inbatch_negatives:\n filler = torch.randint(\n n_items,\n (batch_size, config.inbatch_negatives - corrected_count),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.cat([inbatch, filler], dim=1)\n\n negatives = torch.cat([uniform, inbatch], dim=1).numpy()\n correction = np.ones_like(negatives, dtype=np.float32)\n if corrected_count:\n correction[\n :, config.uniform_negatives : config.uniform_negatives + corrected_count\n ] = item_correction[inbatch[:, :corrected_count]].numpy()\n for _ in range(32):\n invalid = invalid_full_history_negatives(\n negatives,\n targets_cpu,\n user_rows,\n known_positive_packed,\n liked_cpu,\n )\n if not np.any(invalid):\n return (\n torch.from_numpy(negatives).to(device),\n torch.from_numpy(correction).to(device),\n )\n replacements = torch.randint(\n n_items,\n (int(invalid.sum()),),\n generator=generator,\n device=\"cpu\",\n ).numpy()\n negatives[invalid] = replacements\n # Rejected in-batch items become ordinary uniform draws.\n correction[invalid] = 1.0\n invalid = invalid_full_history_negatives(\n negatives,\n targets_cpu,\n user_rows,\n known_positive_packed,\n liked_cpu,\n )\n if np.any(invalid):\n raise RuntimeError(\"full-history negative resampling failed\")\n return (\n torch.from_numpy(negatives).to(device),\n torch.from_numpy(correction).to(device),\n )\n\n\ndef draw_training_negatives(\n positive_items: torch.Tensor,\n known_history: torch.Tensor,\n liked_items: torch.Tensor,\n user_rows: np.ndarray,\n known_positive_packed: np.ndarray | None,\n prepared: SequenceData,\n config: Config,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n if config.negative_filter_scope == \"full-training-history\":\n if known_positive_packed is None:\n raise RuntimeError(\"full-history negative filter was not built\")\n return draw_full_history_filtered_negatives(\n positive_items,\n user_rows,\n liked_items,\n known_positive_packed,\n len(prepared.item_ids),\n config,\n item_correction,\n generator,\n device,\n )\n return base.draw_filtered_mixed_negatives(\n positive_items,\n known_history,\n liked_items,\n len(prepared.item_ids),\n config.uniform_negatives,\n config.inbatch_negatives,\n item_correction,\n generator,\n device,\n )\n\n\ndef event_sequence(\n model: base.SequentialTwoTower,\n items: torch.Tensor,\n features: torch.Tensor,\n hidden: torch.Tensor,\n) -> tuple[torch.Tensor, torch.Tensor]:\n inputs = torch.cat([model.context_embedding(items), features], dim=-1)\n projected = model.event_projection(inputs)"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:43:52.917Z",
"resultDisplay": "Read lines 605-770 of 2165 from ../task/current_model/train_current_model.py",
"description": "../.../train_current_model.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:43:52.918Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 605-770 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 771.\n\n--- FILE CONTENT (truncated) ---\n )\n return packed\n\n\ndef invalid_full_history_negatives(\n negatives: np.ndarray,\n targets: np.ndarray,\n user_rows: np.ndarray,\n known_positive_packed: np.ndarray,\n liked_items: np.ndarray,\n) -> np.ndarray:\n \"\"\"Return row-level invalid negatives using the complete training prefix.\"\"\"\n byte_columns = negatives >> 3\n bit_columns = negatives & 7\n known = known_positive_packed[user_rows[:, None], byte_columns]\n invalid = ((known >> bit_columns) & 1).astype(bool)\n invalid |= negatives == targets[:, None]\n invalid |= (negatives[:, :, None] == liked_items[:, None, :]).any(axis=2)\n for column in range(1, negatives.shape[1]):\n invalid[:, column] |= (\n negatives[:, column, None] == negatives[:, :column]\n ).any(axis=1)\n return invalid\n\n\ndef draw_full_history_filtered_negatives(\n targets: torch.Tensor,\n user_rows: np.ndarray,\n liked_items: torch.Tensor,\n known_positive_packed: np.ndarray,\n n_items: int,\n config: Config,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n \"\"\"Draw mixed negatives and reject all known public-training positives.\"\"\"\n targets_cpu = targets.detach().cpu().numpy().astype(np.int64, copy=False)\n liked_cpu = liked_items.detach().cpu().numpy().astype(np.int64, copy=False)\n batch_size = len(targets_cpu)\n uniform = torch.randint(\n n_items,\n (batch_size, config.uniform_negatives),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.empty((batch_size, 0), dtype=torch.long)\n corrected_count = 0\n if config.inbatch_negatives:\n unique_targets = torch.unique(targets.detach().cpu())\n corrected_count = min(\n config.inbatch_negatives, max(len(unique_targets) - 1, 0)\n )\n if corrected_count:\n priorities = torch.rand(\n (batch_size, len(unique_targets)), generator=generator\n )\n priorities.masked_fill_(\n unique_targets[None, :].eq(\n torch.from_numpy(targets_cpu)[:, None]\n ),\n -1.0,\n )\n columns = torch.topk(\n priorities, k=corrected_count, dim=1\n ).indices\n inbatch = unique_targets[columns]\n if corrected_count < config.inbatch_negatives:\n filler = torch.randint(\n n_items,\n (batch_size, config.inbatch_negatives - corrected_count),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.cat([inbatch, filler], dim=1)\n\n negatives = torch.cat([uniform, inbatch], dim=1).numpy()\n correction = np.ones_like(negatives, dtype=np.float32)\n if corrected_count:\n correction[\n :, config.uniform_negatives : config.uniform_negatives + corrected_count\n ] = item_correction[inbatch[:, :corrected_count]].numpy()\n for _ in range(32):\n invalid = invalid_full_history_negatives(\n negatives,\n targets_cpu,\n user_rows,\n known_positive_packed,\n liked_cpu,\n )\n if not np.any(invalid):\n return (\n torch.from_numpy(negatives).to(device),\n torch.from_numpy(correction).to(device),\n )\n replacements = torch.randint(\n n_items,\n (int(invalid.sum()),),\n generator=generator,\n device=\"cpu\",\n ).numpy()\n negatives[invalid] = replacements\n # Rejected in-batch items become ordinary uniform draws.\n correction[invalid] = 1.0\n invalid = invalid_full_history_negatives(\n negatives,\n targets_cpu,\n user_rows,\n known_positive_packed,\n liked_cpu,\n )\n if np.any(invalid):\n raise RuntimeError(\"full-history negative resampling failed\")\n return (\n torch.from_numpy(negatives).to(device),\n torch.from_numpy(correction).to(device),\n )\n\n\ndef draw_training_negatives(\n positive_items: torch.Tensor,\n known_history: torch.Tensor,\n liked_items: torch.Tensor,\n user_rows: np.ndarray,\n known_positive_packed: np.ndarray | None,\n prepared: SequenceData,\n config: Config,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n if config.negative_filter_scope == \"full-training-history\":\n if known_positive_packed is None:\n raise RuntimeError(\"full-history negative filter was not built\")\n return draw_full_history_filtered_negatives(\n positive_items,\n user_rows,\n liked_items,\n known_positive_packed,\n len(prepared.item_ids),\n config,\n item_correction,\n generator,\n device,\n )\n return base.draw_filtered_mixed_negatives(\n positive_items,\n known_history,\n liked_items,\n len(prepared.item_ids),\n config.uniform_negatives,\n config.inbatch_negatives,\n item_correction,\n generator,\n device,\n )\n\n\ndef event_sequence(\n model: base.SequentialTwoTower,\n items: torch.Tensor,\n features: torch.Tensor,\n hidden: torch.Tensor,\n) -> tuple[torch.Tensor, torch.Tensor]:\n inputs = torch.cat([model.context_embedding(items), features], dim=-1)\n projected = model.event_projection(inputs)",
"isError": false
},
"responseMessageId": "17407898-41d7-4be0-accb-9a1ffcfa8b08"
}
]
},
{
"id": 14,
"role": "agent",
"timestamp": "2026-09-29T05:43:57.248Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:43:57.248Z",
"callId": "read_file__call_6055480",
"name": "read_file",
"input": {
"end_line": 950,
"start_line": 800,
"file_path": "/task/current_model/train_current_model.py"
},
"recorded": {
"id": "read_file__call_6055480",
"name": "read_file",
"args": {
"end_line": 950,
"start_line": 800,
"file_path": "/task/current_model/train_current_model.py"
},
"result": [
{
"functionResponse": {
"id": "read_file__call_6055480",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 800-950 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 951.\n\n--- FILE CONTENT (truncated) ---\n sequence_vectors: torch.Tensor,\n liked_vectors: torch.Tensor,\n disliked_vectors: torch.Tensor,\n user_features: torch.Tensor,\n) -> torch.Tensor:\n return model.user_tower(\n torch.cat(\n [sequence_vectors, liked_vectors, disliked_vectors, user_features],\n dim=1,\n )\n )\n\n\ndef candidate_vectors(\n model: base.SequentialTwoTower,\n positives: torch.Tensor,\n negatives: torch.Tensor,\n) -> tuple[torch.Tensor, torch.Tensor]:\n all_items = torch.cat([positives, negatives.reshape(-1)])\n unique_items, inverse = torch.unique(all_items, return_inverse=True)\n encoded = model.encode_items(unique_items)\n positive_vectors = encoded[inverse[: len(positives)]]\n negative_vectors = encoded[inverse[len(positives) :]].reshape(\n len(positives), negatives.shape[1], -1\n )\n return positive_vectors, negative_vectors\n\n\ndef history_windows(\n sequence: np.ndarray,\n start: int,\n count: int,\n width: int,\n padding_item: int,\n) -> np.ndarray:\n base_start = max(0, start - width + 1)\n segment = sequence[base_start : start + count]\n padded = np.concatenate(\n [np.full(width - 1, padding_item, dtype=np.uint16), segment]\n )\n windows = np.lib.stride_tricks.sliding_window_view(padded, width)\n offset = start - base_start\n return windows[offset : offset + count].copy()\n\n\ndef corrected_negative_loss(\n logits: torch.Tensor,\n correction: torch.Tensor,\n config: Config,\n) -> torch.Tensor:\n raw = F.softplus(logits)\n total = config.uniform_negatives + config.inbatch_negatives\n result = torch.zeros(len(logits), device=logits.device)\n if config.uniform_negatives:\n result += config.uniform_negatives * raw[\n :, : config.uniform_negatives\n ].mean(dim=1)\n if config.inbatch_negatives:\n selected = slice(config.uniform_negatives, total)\n weights = correction[:, selected]\n result += config.inbatch_negatives * (\n (raw[:, selected] * weights).sum(dim=1)\n / weights.sum(dim=1).clamp_min(1.0e-6)\n )\n return result / total\n\n\ndef retained_target_window_starts(\n total_targets: np.ndarray,\n validation_counts: np.ndarray,\n training_counts: np.ndarray,\n) -> np.ndarray:\n \"\"\"Return the discarded prefix before the retained chronological suffix.\"\"\"\n training_region_ends = total_targets - validation_counts\n window_starts = training_region_ends - training_counts\n if (\n np.any(window_starts < 0)\n or np.any(training_counts < 0)\n or np.any(validation_counts < 0)\n ):\n raise RuntimeError(\"chronological target partition produced negative counts\")\n if not np.array_equal(\n window_starts + training_counts + validation_counts, total_targets\n ):\n raise RuntimeError(\"chronological target window is not a final contiguous suffix\")\n return window_starts\n\n\ndef chronological_target_partition(\n prepared: SequenceData,\n validation_fraction: float,\n max_train_targets_per_user: int,\n max_validation_targets_per_user: int,\n) -> TargetPartition:\n \"\"\"Select a recent contiguous window with a final out-of-time BCE suffix.\"\"\"\n totals = np.maximum(prepared.lengths - 1, 0).astype(np.int64)\n validation = np.floor(totals * validation_fraction).astype(np.int64)\n validation[(totals >= 2) & (validation == 0)] = 1\n if max_validation_targets_per_user:\n validation = np.minimum(validation, max_validation_targets_per_user)\n\n history_training = totals - validation\n training = history_training.copy()\n if max_train_targets_per_user:\n training = np.minimum(training, max_train_targets_per_user)\n window_starts = retained_target_window_starts(totals, validation, training)\n return TargetPartition(\n window_starts=window_starts,\n training_counts=training,\n validation_counts=validation,\n history_training_counts=history_training,\n )\n\n\ndef load_organic_like_outcomes(\n public: Path, stage: str, item_ids: np.ndarray\n) -> dict[int, dict[int, np.ndarray]]:\n \"\"\"Load mapped organic-like timestamps from public fitting history.\"\"\"\n prefixes = (\"train\",) if stage == \"validation\" else (\"train\", \"val\")\n frames = [\n pd.read_parquet(\n public / \"data\" / f\"{prefix}_likes.parquet\",\n columns=[\"uid\", \"timestamp\", \"item_id\", \"is_organic\"],\n )\n for prefix in prefixes\n ]\n likes = pd.concat(frames, ignore_index=True)\n likes = likes[likes.is_organic == 1].copy()\n mapped = pd.Index(item_ids).get_indexer(\n likes.item_id.to_numpy(dtype=np.int64)\n )\n likes = likes.loc[mapped >= 0, [\"uid\", \"timestamp\"]].copy()\n likes[\"item\"] = mapped[mapped >= 0].astype(np.int32, copy=False)\n likes = likes.sort_values([\"uid\", \"item\", \"timestamp\"], kind=\"stable\")\n nested: dict[int, dict[int, list[int]]] = {}\n for raw_uid, raw_timestamp, raw_item in likes.itertuples(\n index=False, name=None\n ):\n nested.setdefault(int(raw_uid), {}).setdefault(int(raw_item), []).append(\n int(raw_timestamp)\n )\n result = {\n uid: {\n item: np.asarray(timestamps, dtype=np.int64)\n for item, timestamps in items.items()\n }\n for uid, items in nested.items()\n }\n log(\n f\"loaded organic-like outcomes: rows={len(likes):,} \"\n f\"users={len(result):,}\""
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:43:57.256Z",
"resultDisplay": "Read lines 800-950 of 2165 from ../task/current_model/train_current_model.py",
"description": "../.../train_current_model.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:43:57.257Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 800-950 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 951.\n\n--- FILE CONTENT (truncated) ---\n sequence_vectors: torch.Tensor,\n liked_vectors: torch.Tensor,\n disliked_vectors: torch.Tensor,\n user_features: torch.Tensor,\n) -> torch.Tensor:\n return model.user_tower(\n torch.cat(\n [sequence_vectors, liked_vectors, disliked_vectors, user_features],\n dim=1,\n )\n )\n\n\ndef candidate_vectors(\n model: base.SequentialTwoTower,\n positives: torch.Tensor,\n negatives: torch.Tensor,\n) -> tuple[torch.Tensor, torch.Tensor]:\n all_items = torch.cat([positives, negatives.reshape(-1)])\n unique_items, inverse = torch.unique(all_items, return_inverse=True)\n encoded = model.encode_items(unique_items)\n positive_vectors = encoded[inverse[: len(positives)]]\n negative_vectors = encoded[inverse[len(positives) :]].reshape(\n len(positives), negatives.shape[1], -1\n )\n return positive_vectors, negative_vectors\n\n\ndef history_windows(\n sequence: np.ndarray,\n start: int,\n count: int,\n width: int,\n padding_item: int,\n) -> np.ndarray:\n base_start = max(0, start - width + 1)\n segment = sequence[base_start : start + count]\n padded = np.concatenate(\n [np.full(width - 1, padding_item, dtype=np.uint16), segment]\n )\n windows = np.lib.stride_tricks.sliding_window_view(padded, width)\n offset = start - base_start\n return windows[offset : offset + count].copy()\n\n\ndef corrected_negative_loss(\n logits: torch.Tensor,\n correction: torch.Tensor,\n config: Config,\n) -> torch.Tensor:\n raw = F.softplus(logits)\n total = config.uniform_negatives + config.inbatch_negatives\n result = torch.zeros(len(logits), device=logits.device)\n if config.uniform_negatives:\n result += config.uniform_negatives * raw[\n :, : config.uniform_negatives\n ].mean(dim=1)\n if config.inbatch_negatives:\n selected = slice(config.uniform_negatives, total)\n weights = correction[:, selected]\n result += config.inbatch_negatives * (\n (raw[:, selected] * weights).sum(dim=1)\n / weights.sum(dim=1).clamp_min(1.0e-6)\n )\n return result / total\n\n\ndef retained_target_window_starts(\n total_targets: np.ndarray,\n validation_counts: np.ndarray,\n training_counts: np.ndarray,\n) -> np.ndarray:\n \"\"\"Return the discarded prefix before the retained chronological suffix.\"\"\"\n training_region_ends = total_targets - validation_counts\n window_starts = training_region_ends - training_counts\n if (\n np.any(window_starts < 0)\n or np.any(training_counts < 0)\n or np.any(validation_counts < 0)\n ):\n raise RuntimeError(\"chronological target partition produced negative counts\")\n if not np.array_equal(\n window_starts + training_counts + validation_counts, total_targets\n ):\n raise RuntimeError(\"chronological target window is not a final contiguous suffix\")\n return window_starts\n\n\ndef chronological_target_partition(\n prepared: SequenceData,\n validation_fraction: float,\n max_train_targets_per_user: int,\n max_validation_targets_per_user: int,\n) -> TargetPartition:\n \"\"\"Select a recent contiguous window with a final out-of-time BCE suffix.\"\"\"\n totals = np.maximum(prepared.lengths - 1, 0).astype(np.int64)\n validation = np.floor(totals * validation_fraction).astype(np.int64)\n validation[(totals >= 2) & (validation == 0)] = 1\n if max_validation_targets_per_user:\n validation = np.minimum(validation, max_validation_targets_per_user)\n\n history_training = totals - validation\n training = history_training.copy()\n if max_train_targets_per_user:\n training = np.minimum(training, max_train_targets_per_user)\n window_starts = retained_target_window_starts(totals, validation, training)\n return TargetPartition(\n window_starts=window_starts,\n training_counts=training,\n validation_counts=validation,\n history_training_counts=history_training,\n )\n\n\ndef load_organic_like_outcomes(\n public: Path, stage: str, item_ids: np.ndarray\n) -> dict[int, dict[int, np.ndarray]]:\n \"\"\"Load mapped organic-like timestamps from public fitting history.\"\"\"\n prefixes = (\"train\",) if stage == \"validation\" else (\"train\", \"val\")\n frames = [\n pd.read_parquet(\n public / \"data\" / f\"{prefix}_likes.parquet\",\n columns=[\"uid\", \"timestamp\", \"item_id\", \"is_organic\"],\n )\n for prefix in prefixes\n ]\n likes = pd.concat(frames, ignore_index=True)\n likes = likes[likes.is_organic == 1].copy()\n mapped = pd.Index(item_ids).get_indexer(\n likes.item_id.to_numpy(dtype=np.int64)\n )\n likes = likes.loc[mapped >= 0, [\"uid\", \"timestamp\"]].copy()\n likes[\"item\"] = mapped[mapped >= 0].astype(np.int32, copy=False)\n likes = likes.sort_values([\"uid\", \"item\", \"timestamp\"], kind=\"stable\")\n nested: dict[int, dict[int, list[int]]] = {}\n for raw_uid, raw_timestamp, raw_item in likes.itertuples(\n index=False, name=None\n ):\n nested.setdefault(int(raw_uid), {}).setdefault(int(raw_item), []).append(\n int(raw_timestamp)\n )\n result = {\n uid: {\n item: np.asarray(timestamps, dtype=np.int64)\n for item, timestamps in items.items()\n }\n for uid, items in nested.items()\n }\n log(\n f\"loaded organic-like outcomes: rows={len(likes):,} \"\n f\"users={len(result):,}\"",
"isError": false
},
"responseMessageId": "a911048b-29a4-4551-8336-6834674c4b40"
}
]
},
{
"id": 15,
"role": "agent",
"timestamp": "2026-09-29T05:43:59.114Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:43:59.114Z",
"callId": "read_file__call_5962667",
"name": "read_file",
"input": {
"end_line": 1250,
"file_path": "/task/current_model/train_current_model.py",
"start_line": 951
},
"recorded": {
"id": "read_file__call_5962667",
"name": "read_file",
"args": {
"end_line": 1250,
"file_path": "/task/current_model/train_current_model.py",
"start_line": 951
},
"result": [
{
"functionResponse": {
"id": "read_file__call_5962667",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 951-1250 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 1251.\n\n--- FILE CONTENT (truncated) ---\n )\n return result\n\n\ndef build_target_weights(\n public: Path,\n stage: str,\n prepared: SequenceData,\n training_counts: np.ndarray,\n config: Config,\n) -> np.ndarray:\n \"\"\"Build nonzero per-token weights without dropping adjacent targets.\"\"\"\n if config.target_weighting == \"uniform\":\n weights = np.ones(len(prepared.sequence_items), dtype=np.float32)\n else:\n weights = np.full(\n len(prepared.sequence_items),\n config.ordinary_target_weight,\n dtype=np.float32,\n )\n organic_strong = (\n (prepared.event_features[:, 0] > 0.5)\n & (prepared.event_features[:, 1] >= 0.8)\n )\n weights[organic_strong] = config.organic_strong_target_weight\n if config.target_weighting == \"discovery-graded\":\n like_outcomes = load_organic_like_outcomes(\n public, stage, prepared.item_ids\n )\n history_end = (\n base.TRAIN_END if stage == \"validation\" else base.VALIDATION_END\n )\n horizon = config.discovery_horizon_days * base.DAY\n grade_three = 0\n for user_row, uid in enumerate(prepared.user_ids):\n start = int(prepared.offsets[user_row])\n end = int(prepared.offsets[user_row + 1])\n if end - start < 2:\n continue\n last_training_target = start + int(training_counts[user_row])\n first_validation_target = last_training_target + 1\n split_timestamp = (\n int(prepared.sequence_timestamps[first_validation_target])\n if first_validation_target < end\n else history_end\n )\n user_likes = like_outcomes.get(int(uid), {})\n seen = {int(prepared.sequence_items[start])}\n for position in range(start + 1, end):\n item = int(prepared.sequence_items[position])\n novel = item not in seen\n seen.add(item)\n if not novel:\n # Repeats remain training examples, but they cannot be\n # eligible discovery outcomes at the corresponding\n # cutoff and therefore keep only the ordinary weight.\n weights[position] = config.ordinary_target_weight\n continue\n timestamps = user_likes.get(item)\n if timestamps is None:\n continue\n timestamp = int(prepared.sequence_timestamps[position])\n upper = min(\n timestamp + horizon,\n split_timestamp\n if position <= last_training_target\n else history_end,\n )\n outcome = int(np.searchsorted(timestamps, timestamp, side=\"left\"))\n if outcome < len(timestamps) and int(timestamps[outcome]) < upper:\n weights[position] = config.organic_like_target_weight\n grade_three += 1\n log(\n f\"grade-3 first-encounter targets={grade_three:,} \"\n f\"horizon_days={config.discovery_horizon_days}\"\n )\n log(\n f\"target weighting={config.target_weighting} \"\n f\"min={weights.min():.3f} mean={weights.mean():.3f} \"\n f\"max={weights.max():.3f}\"\n )\n return weights\n\n\ndef build_dense_block(\n prepared: SequenceData,\n target_weights: np.ndarray,\n users: np.ndarray,\n window_starts: np.ndarray,\n process_counts: np.ndarray,\n score_starts: np.ndarray,\n start: int,\n steps: int,\n config: Config,\n) -> tuple[np.ndarray, ...]:\n \"\"\"Build one causal sequence block and mark the target rows to score.\"\"\"\n batch_size = len(users)\n padding_item = len(prepared.item_ids)\n input_items = np.full((batch_size, steps), padding_item, dtype=np.uint16)\n event_features = np.zeros((batch_size, steps, 3), dtype=np.float16)\n activity_features = np.zeros((batch_size, steps, 4), dtype=np.float16)\n state_ids = np.zeros((batch_size, steps), dtype=np.uint32)\n targets = np.zeros((batch_size, steps), dtype=np.uint16)\n weights = np.ones((batch_size, steps), dtype=np.float32)\n scored = np.zeros((batch_size, steps), dtype=bool)\n histories = np.full(\n (batch_size, steps, config.bptt_steps),\n padding_item,\n dtype=np.uint16,\n )\n processed = np.zeros(batch_size, dtype=np.int32)\n\n lengths = prepared.lengths\n for row, user_row in enumerate(users):\n count = min(steps, max(int(process_counts[row]) - start, 0))\n if not count:\n continue\n offset = int(prepared.offsets[user_row])\n window_start = int(window_starts[row])\n absolute_start = window_start + start\n positions = slice(\n offset + absolute_start,\n offset + absolute_start + count,\n )\n target_positions = slice(\n offset + absolute_start + 1,\n offset + absolute_start + count + 1,\n )\n input_items[row, :count] = prepared.sequence_items[positions]\n event_features[row, :count] = prepared.event_features[positions]\n activity_features[row, :count] = prepared.activity_features[positions]\n state_ids[row, :count] = prepared.state_ids[positions]\n targets[row, :count] = prepared.sequence_items[target_positions]\n weights[row, :count] = target_weights[target_positions]\n score_from = min(count, max(int(score_starts[row]) - start, 0))\n scored[row, score_from:count] = True\n processed[row] = count\n user_sequence = prepared.sequence_items[\n offset : offset + int(lengths[user_row])\n ]\n histories[row, :count] = history_windows(\n user_sequence,\n absolute_start,\n count,\n config.bptt_steps,\n padding_item,\n )\n return (\n input_items,\n event_features,\n activity_features,\n state_ids,\n targets,\n scored,\n histories,\n processed,\n weights,\n )\n\n\ndef score_target_microbatches(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n sequence_vectors: torch.Tensor,\n flat_states: np.ndarray,\n flat_activity: np.ndarray,\n positive_items: torch.Tensor,\n negatives: torch.Tensor,\n negative_weights: torch.Tensor,\n sample_weights: np.ndarray,\n device: torch.device,\n backward: bool,\n) -> tuple[dict[str, float], torch.Tensor | None]:\n \"\"\"Compute one logical block's exact BCE in bounded-memory slices.\"\"\"\n target_count = len(positive_items)\n total_weight = float(sample_weights.sum())\n if total_weight <= 0:\n raise RuntimeError(\"target block has no positive sample weight\")\n sequence_gradient = torch.zeros_like(sequence_vectors) if backward else None\n statistics = {\n \"loss_sum\": 0.0,\n \"positive_sum\": 0.0,\n \"negative_sum\": 0.0,\n \"pair_correct\": 0.0,\n \"weight_sum\": 0.0,\n }\n for begin in range(0, target_count, config.loss_microbatch_size):\n end = min(begin + config.loss_microbatch_size, target_count)\n selected_states = flat_states[begin:end]\n if backward:\n sequence_input = (\n sequence_vectors[begin:end].detach().requires_grad_(True)\n )\n else:\n sequence_input = sequence_vectors[begin:end]\n liked_vectors, disliked_vectors, _ = pool_explicit_states(\n model, prepared, selected_states, device\n )\n counts = prepared.state_counts[selected_states]\n numeric = np.zeros((end - begin, 6), dtype=np.float32)\n numeric[:, :4] = flat_activity[begin:end]\n numeric[:, 4] = np.log1p(counts[:, 0]) / 6.0\n numeric[:, 5] = np.log1p(counts[:, 1]) / 4.0\n query_vectors = user_queries(\n model,\n sequence_input,\n liked_vectors,\n disliked_vectors,\n torch.from_numpy(numeric).to(device),\n )\n positive_vectors, negative_vectors = candidate_vectors(\n model,\n positive_items[begin:end],\n negatives[begin:end],\n )\n positive_logits = (query_vectors * positive_vectors).sum(dim=1)\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n positive_loss = F.softplus(-positive_logits)\n negative_loss = corrected_negative_loss(\n negative_logits, negative_weights[begin:end], config\n )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n loss_sum = (\n selected_weights * (positive_loss + negative_loss) / 2.0\n ).sum()\n if not bool(torch.isfinite(loss_sum)):\n raise RuntimeError(\"dense next-item training produced non-finite loss\")\n if backward:\n (loss_sum / total_weight).backward()\n if sequence_input.grad is None:\n raise RuntimeError(\"sequence microbatch did not produce a gradient\")\n sequence_gradient[begin:end] = sequence_input.grad\n statistics[\"loss_sum\"] += float(loss_sum.detach().cpu())\n statistics[\"positive_sum\"] += float(positive_logits.detach().sum().cpu())\n statistics[\"negative_sum\"] += float(negative_logits.detach().sum().cpu())\n statistics[\"pair_correct\"] += float(\n (positive_logits.detach()[:, None] > negative_logits.detach()).sum().cpu()\n )\n statistics[\"weight_sum\"] += float(selected_weights.sum().detach().cpu())\n return statistics, sequence_gradient\n\n\ndef release_accelerator_cache(device: torch.device) -> None:\n if device.type == \"mps\":\n torch.mps.empty_cache()\n\n\ndef run_training_epoch(\n model: base.SequentialTwoTower,\n optimizer: torch.optim.Optimizer,\n prepared: SequenceData,\n config: Config,\n target_weights: np.ndarray,\n window_starts: np.ndarray,\n target_counts: np.ndarray,\n item_correction: torch.Tensor,\n known_positive_packed: np.ndarray | None,\n generator: torch.Generator,\n device: torch.device,\n epoch: int,\n) -> dict[str, float]:\n model.train()\n train_users = np.flatnonzero(target_counts > 0)\n sorted_users = train_users[\n np.argsort(target_counts[train_users], kind=\"stable\")\n ]\n groups = [\n sorted_users[start : start + config.user_batch_size]\n for start in range(0, len(sorted_users), config.user_batch_size)\n ]\n group_order = torch.randperm(len(groups), generator=generator).numpy()\n negative_count = config.uniform_negatives + config.inbatch_negatives\n total_loss = 0.0\n total_targets = 0\n total_weight = 0.0\n total_positive = 0.0\n total_negative = 0.0\n total_pair_correct = 0.0\n batches = 0\n started = time.monotonic()\n\n for group_number in group_order:\n users = groups[int(group_number)]\n group_counts = target_counts[users]\n hidden = torch.zeros(\n 1, len(users), config.embedding_dim, device=device\n )\n max_targets = int(group_counts.max())\n for start in range(0, max_targets, config.bptt_steps):\n steps = min(config.bptt_steps, max_targets - start)\n block = build_dense_block(\n prepared,\n target_weights,\n users,\n window_starts[users],"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:43:59.119Z",
"resultDisplay": "Read lines 951-1250 of 2165 from ../task/current_model/train_current_model.py",
"description": "../.../train_current_model.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:43:59.120Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 951-1250 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 1251.\n\n--- FILE CONTENT (truncated) ---\n )\n return result\n\n\ndef build_target_weights(\n public: Path,\n stage: str,\n prepared: SequenceData,\n training_counts: np.ndarray,\n config: Config,\n) -> np.ndarray:\n \"\"\"Build nonzero per-token weights without dropping adjacent targets.\"\"\"\n if config.target_weighting == \"uniform\":\n weights = np.ones(len(prepared.sequence_items), dtype=np.float32)\n else:\n weights = np.full(\n len(prepared.sequence_items),\n config.ordinary_target_weight,\n dtype=np.float32,\n )\n organic_strong = (\n (prepared.event_features[:, 0] > 0.5)\n & (prepared.event_features[:, 1] >= 0.8)\n )\n weights[organic_strong] = config.organic_strong_target_weight\n if config.target_weighting == \"discovery-graded\":\n like_outcomes = load_organic_like_outcomes(\n public, stage, prepared.item_ids\n )\n history_end = (\n base.TRAIN_END if stage == \"validation\" else base.VALIDATION_END\n )\n horizon = config.discovery_horizon_days * base.DAY\n grade_three = 0\n for user_row, uid in enumerate(prepared.user_ids):\n start = int(prepared.offsets[user_row])\n end = int(prepared.offsets[user_row + 1])\n if end - start < 2:\n continue\n last_training_target = start + int(training_counts[user_row])\n first_validation_target = last_training_target + 1\n split_timestamp = (\n int(prepared.sequence_timestamps[first_validation_target])\n if first_validation_target < end\n else history_end\n )\n user_likes = like_outcomes.get(int(uid), {})\n seen = {int(prepared.sequence_items[start])}\n for position in range(start + 1, end):\n item = int(prepared.sequence_items[position])\n novel = item not in seen\n seen.add(item)\n if not novel:\n # Repeats remain training examples, but they cannot be\n # eligible discovery outcomes at the corresponding\n # cutoff and therefore keep only the ordinary weight.\n weights[position] = config.ordinary_target_weight\n continue\n timestamps = user_likes.get(item)\n if timestamps is None:\n continue\n timestamp = int(prepared.sequence_timestamps[position])\n upper = min(\n timestamp + horizon,\n split_timestamp\n if position <= last_training_target\n else history_end,\n )\n outcome = int(np.searchsorted(timestamps, timestamp, side=\"left\"))\n if outcome < len(timestamps) and int(timestamps[outcome]) < upper:\n weights[position] = config.organic_like_target_weight\n grade_three += 1\n log(\n f\"grade-3 first-encounter targets={grade_three:,} \"\n f\"horizon_days={config.discovery_horizon_days}\"\n )\n log(\n f\"target weighting={config.target_weighting} \"\n f\"min={weights.min():.3f} mean={weights.mean():.3f} \"\n f\"max={weights.max():.3f}\"\n )\n return weights\n\n\ndef build_dense_block(\n prepared: SequenceData,\n target_weights: np.ndarray,\n users: np.ndarray,\n window_starts: np.ndarray,\n process_counts: np.ndarray,\n score_starts: np.ndarray,\n start: int,\n steps: int,\n config: Config,\n) -> tuple[np.ndarray, ...]:\n \"\"\"Build one causal sequence block and mark the target rows to score.\"\"\"\n batch_size = len(users)\n padding_item = len(prepared.item_ids)\n input_items = np.full((batch_size, steps), padding_item, dtype=np.uint16)\n event_features = np.zeros((batch_size, steps, 3), dtype=np.float16)\n activity_features = np.zeros((batch_size, steps, 4), dtype=np.float16)\n state_ids = np.zeros((batch_size, steps), dtype=np.uint32)\n targets = np.zeros((batch_size, steps), dtype=np.uint16)\n weights = np.ones((batch_size, steps), dtype=np.float32)\n scored = np.zeros((batch_size, steps), dtype=bool)\n histories = np.full(\n (batch_size, steps, config.bptt_steps),\n padding_item,\n dtype=np.uint16,\n )\n processed = np.zeros(batch_size, dtype=np.int32)\n\n lengths = prepared.lengths\n for row, user_row in enumerate(users):\n count = min(steps, max(int(process_counts[row]) - start, 0))\n if not count:\n continue\n offset = int(prepared.offsets[user_row])\n window_start = int(window_starts[row])\n absolute_start = window_start + start\n positions = slice(\n offset + absolute_start,\n offset + absolute_start + count,\n )\n target_positions = slice(\n offset + absolute_start + 1,\n offset + absolute_start + count + 1,\n )\n input_items[row, :count] = prepared.sequence_items[positions]\n event_features[row, :count] = prepared.event_features[positions]\n activity_features[row, :count] = prepared.activity_features[positions]\n state_ids[row, :count] = prepared.state_ids[positions]\n targets[row, :count] = prepared.sequence_items[target_positions]\n weights[row, :count] = target_weights[target_positions]\n score_from = min(count, max(int(score_starts[row]) - start, 0))\n scored[row, score_from:count] = True\n processed[row] = count\n user_sequence = prepared.sequence_items[\n offset : offset + int(lengths[user_row])\n ]\n histories[row, :count] = history_windows(\n user_sequence,\n absolute_start,\n count,\n config.bptt_steps,\n padding_item,\n )\n return (\n input_items,\n event_features,\n activity_features,\n state_ids,\n targets,\n scored,\n histories,\n processed,\n weights,\n )\n\n\ndef score_target_microbatches(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n sequence_vectors: torch.Tensor,\n flat_states: np.ndarray,\n flat_activity: np.ndarray,\n positive_items: torch.Tensor,\n negatives: torch.Tensor,\n negative_weights: torch.Tensor,\n sample_weights: np.ndarray,\n device: torch.device,\n backward: bool,\n) -> tuple[dict[str, float], torch.Tensor | None]:\n \"\"\"Compute one logical block's exact BCE in bounded-memory slices.\"\"\"\n target_count = len(positive_items)\n total_weight = float(sample_weights.sum())\n if total_weight <= 0:\n raise RuntimeError(\"target block has no positive sample weight\")\n sequence_gradient = torch.zeros_like(sequence_vectors) if backward else None\n statistics = {\n \"loss_sum\": 0.0,\n \"positive_sum\": 0.0,\n \"negative_sum\": 0.0,\n \"pair_correct\": 0.0,\n \"weight_sum\": 0.0,\n }\n for begin in range(0, target_count, config.loss_microbatch_size):\n end = min(begin + config.loss_microbatch_size, target_count)\n selected_states = flat_states[begin:end]\n if backward:\n sequence_input = (\n sequence_vectors[begin:end].detach().requires_grad_(True)\n )\n else:\n sequence_input = sequence_vectors[begin:end]\n liked_vectors, disliked_vectors, _ = pool_explicit_states(\n model, prepared, selected_states, device\n )\n counts = prepared.state_counts[selected_states]\n numeric = np.zeros((end - begin, 6), dtype=np.float32)\n numeric[:, :4] = flat_activity[begin:end]\n numeric[:, 4] = np.log1p(counts[:, 0]) / 6.0\n numeric[:, 5] = np.log1p(counts[:, 1]) / 4.0\n query_vectors = user_queries(\n model,\n sequence_input,\n liked_vectors,\n disliked_vectors,\n torch.from_numpy(numeric).to(device),\n )\n positive_vectors, negative_vectors = candidate_vectors(\n model,\n positive_items[begin:end],\n negatives[begin:end],\n )\n positive_logits = (query_vectors * positive_vectors).sum(dim=1)\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n positive_loss = F.softplus(-positive_logits)\n negative_loss = corrected_negative_loss(\n negative_logits, negative_weights[begin:end], config\n )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n loss_sum = (\n selected_weights * (positive_loss + negative_loss) / 2.0\n ).sum()\n if not bool(torch.isfinite(loss_sum)):\n raise RuntimeError(\"dense next-item training produced non-finite loss\")\n if backward:\n (loss_sum / total_weight).backward()\n if sequence_input.grad is None:\n raise RuntimeError(\"sequence microbatch did not produce a gradient\")\n sequence_gradient[begin:end] = sequence_input.grad\n statistics[\"loss_sum\"] += float(loss_sum.detach().cpu())\n statistics[\"positive_sum\"] += float(positive_logits.detach().sum().cpu())\n statistics[\"negative_sum\"] += float(negative_logits.detach().sum().cpu())\n statistics[\"pair_correct\"] += float(\n (positive_logits.detach()[:, None] > negative_logits.detach()).sum().cpu()\n )\n statistics[\"weight_sum\"] += float(selected_weights.sum().detach().cpu())\n return statistics, sequence_gradient\n\n\ndef release_accelerator_cache(device: torch.device) -> None:\n if device.type == \"mps\":\n torch.mps.empty_cache()\n\n\ndef run_training_epoch(\n model: base.SequentialTwoTower,\n optimizer: torch.optim.Optimizer,\n prepared: SequenceData,\n config: Config,\n target_weights: np.ndarray,\n window_starts: np.ndarray,\n target_counts: np.ndarray,\n item_correction: torch.Tensor,\n known_positive_packed: np.ndarray | None,\n generator: torch.Generator,\n device: torch.device,\n epoch: int,\n) -> dict[str, float]:\n model.train()\n train_users = np.flatnonzero(target_counts > 0)\n sorted_users = train_users[\n np.argsort(target_counts[train_users], kind=\"stable\")\n ]\n groups = [\n sorted_users[start : start + config.user_batch_size]\n for start in range(0, len(sorted_users), config.user_batch_size)\n ]\n group_order = torch.randperm(len(groups), generator=generator).numpy()\n negative_count = config.uniform_negatives + config.inbatch_negatives\n total_loss = 0.0\n total_targets = 0\n total_weight = 0.0\n total_positive = 0.0\n total_negative = 0.0\n total_pair_correct = 0.0\n batches = 0\n started = time.monotonic()\n\n for group_number in group_order:\n users = groups[int(group_number)]\n group_counts = target_counts[users]\n hidden = torch.zeros(\n 1, len(users), config.embedding_dim, device=device\n )\n max_targets = int(group_counts.max())\n for start in range(0, max_targets, config.bptt_steps):\n steps = min(config.bptt_steps, max_targets - start)\n block = build_dense_block(\n prepared,\n target_weights,\n users,\n window_starts[users],",
"isError": false
},
"responseMessageId": "7a633a02-1439-48bc-995c-34bac2084654"
}
]
},
{
"id": 16,
"role": "agent",
"timestamp": "2026-09-29T05:44:00.897Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:44:00.897Z",
"callId": "read_file__call_5146743",
"name": "read_file",
"input": {
"file_path": "/task/current_model/train_current_model.py",
"start_line": 1251,
"end_line": 1400
},
"recorded": {
"id": "read_file__call_5146743",
"name": "read_file",
"args": {
"file_path": "/task/current_model/train_current_model.py",
"start_line": 1251,
"end_line": 1400
},
"result": [
{
"functionResponse": {
"id": "read_file__call_5146743",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1251-1400 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 1401.\n\n--- FILE CONTENT (truncated) ---\n group_counts,\n np.zeros(len(users), dtype=np.int64),\n start,\n steps,\n config,\n )\n sequence_outputs, _ = event_sequence(\n model,\n torch.from_numpy(block[0].astype(np.int64)).to(device),\n torch.from_numpy(block[1].astype(np.float32)).to(device),\n hidden,\n )\n hidden = torch.stack(\n [\n sequence_outputs[row, count - 1]\n if count\n else hidden[0, row]\n for row, count in enumerate(block[7])\n ]\n ).unsqueeze(0).detach()\n flat_scored = block[5].reshape(-1)\n if not np.any(flat_scored):\n continue\n mask = torch.from_numpy(flat_scored).to(device)\n sequence_vectors = sequence_outputs.reshape(\n -1, config.embedding_dim\n )[mask]\n flat_states = block[3].reshape(-1)[flat_scored]\n positive_items = torch.from_numpy(\n block[4].reshape(-1)[flat_scored].astype(np.int64)\n ).to(device)\n known_history = torch.from_numpy(\n block[6].reshape(-1, config.bptt_steps)[flat_scored].astype(\n np.int64\n )\n ).to(device)\n liked_items = torch.from_numpy(\n prepared.liked_states[flat_states].astype(np.int64)\n ).to(device)\n flat_user_rows = np.broadcast_to(\n users[:, None], block[5].shape\n ).reshape(-1)[flat_scored]\n negatives, negative_weights = draw_training_negatives(\n positive_items,\n known_history,\n liked_items,\n flat_user_rows,\n known_positive_packed,\n prepared,\n config,\n item_correction,\n generator,\n device,\n )\n optimizer.zero_grad(set_to_none=True)\n statistics, sequence_gradient = score_target_microbatches(\n model,\n prepared,\n config,\n sequence_vectors,\n flat_states,\n block[2].reshape(-1, 4)[flat_scored].astype(np.float32),\n positive_items,\n negatives,\n negative_weights,\n block[8].reshape(-1)[flat_scored],\n device,\n backward=True,\n )\n if sequence_gradient is None:\n raise RuntimeError(\"training block omitted its sequence gradient\")\n sequence_vectors.backward(sequence_gradient)\n nn.utils.clip_grad_norm_(model.parameters(), config.gradient_clip)\n optimizer.step()\n\n count = len(positive_items)\n total_loss += statistics[\"loss_sum\"]\n total_targets += count\n total_weight += statistics[\"weight_sum\"]\n total_positive += statistics[\"positive_sum\"]\n total_negative += statistics[\"negative_sum\"]\n total_pair_correct += statistics[\"pair_correct\"]\n batches += 1\n if batches % 25 == 0:\n release_accelerator_cache(device)\n if batches % 250 == 0:\n batch_bce = statistics[\"loss_sum\"] / max(\n statistics[\"weight_sum\"], 1.0e-12\n )\n running_bce = total_loss / max(total_weight, 1.0e-12)\n log(\n f\"epoch={epoch}/{config.epochs} batches={batches:,} \"\n f\"targets={total_targets:,} \"\n f\"batch_bce={batch_bce:.6f} \"\n f\"running_bce={running_bce:.6f} \"\n f\"elapsed={time.monotonic() - started:.1f}s\"\n )\n\n return {\n \"epoch\": float(epoch),\n \"loss\": total_loss / max(total_weight, 1.0e-12),\n \"positive_logit\": total_positive / max(total_targets, 1),\n \"negative_logit\": total_negative / max(total_targets * negative_count, 1),\n \"pair_accuracy\": total_pair_correct\n / max(total_targets * negative_count, 1),\n \"targets\": float(total_targets),\n \"target_weight_sum\": total_weight,\n \"batches\": float(batches),\n \"seconds\": time.monotonic() - started,\n }\n\n\n@torch.no_grad()\ndef evaluate_validation_bce(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n target_weights: np.ndarray,\n window_starts: np.ndarray,\n training_counts: np.ndarray,\n validation_counts: np.ndarray,\n item_correction: torch.Tensor,\n known_positive_packed: np.ndarray | None,\n device: torch.device,\n) -> dict[str, float]:\n \"\"\"Measure deterministic BCE on each user's chronological target suffix.\"\"\"\n model.eval()\n totals = training_counts + validation_counts\n users_with_validation = np.flatnonzero(validation_counts > 0)\n ordered = users_with_validation[\n np.argsort(totals[users_with_validation], kind=\"stable\")\n ]\n groups = [\n ordered[start : start + config.user_batch_size]\n for start in range(0, len(ordered), config.user_batch_size)\n ]\n generator = torch.Generator(device=\"cpu\").manual_seed(config.seed + 20_000)\n total_loss = 0.0\n total_targets = 0\n total_weight = 0.0\n started = time.monotonic()\n\n for group_number, users in enumerate(groups, start=1):\n group_totals = totals[users]\n group_starts = training_counts[users]\n hidden = torch.zeros(\n 1, len(users), config.embedding_dim, device=device\n )\n max_targets = int(group_totals.max())\n for start in range(0, max_targets, config.bptt_steps):"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:44:00.901Z",
"resultDisplay": "Read lines 1251-1400 of 2165 from ../task/current_model/train_current_model.py",
"description": "../.../train_current_model.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:44:00.902Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1251-1400 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 1401.\n\n--- FILE CONTENT (truncated) ---\n group_counts,\n np.zeros(len(users), dtype=np.int64),\n start,\n steps,\n config,\n )\n sequence_outputs, _ = event_sequence(\n model,\n torch.from_numpy(block[0].astype(np.int64)).to(device),\n torch.from_numpy(block[1].astype(np.float32)).to(device),\n hidden,\n )\n hidden = torch.stack(\n [\n sequence_outputs[row, count - 1]\n if count\n else hidden[0, row]\n for row, count in enumerate(block[7])\n ]\n ).unsqueeze(0).detach()\n flat_scored = block[5].reshape(-1)\n if not np.any(flat_scored):\n continue\n mask = torch.from_numpy(flat_scored).to(device)\n sequence_vectors = sequence_outputs.reshape(\n -1, config.embedding_dim\n )[mask]\n flat_states = block[3].reshape(-1)[flat_scored]\n positive_items = torch.from_numpy(\n block[4].reshape(-1)[flat_scored].astype(np.int64)\n ).to(device)\n known_history = torch.from_numpy(\n block[6].reshape(-1, config.bptt_steps)[flat_scored].astype(\n np.int64\n )\n ).to(device)\n liked_items = torch.from_numpy(\n prepared.liked_states[flat_states].astype(np.int64)\n ).to(device)\n flat_user_rows = np.broadcast_to(\n users[:, None], block[5].shape\n ).reshape(-1)[flat_scored]\n negatives, negative_weights = draw_training_negatives(\n positive_items,\n known_history,\n liked_items,\n flat_user_rows,\n known_positive_packed,\n prepared,\n config,\n item_correction,\n generator,\n device,\n )\n optimizer.zero_grad(set_to_none=True)\n statistics, sequence_gradient = score_target_microbatches(\n model,\n prepared,\n config,\n sequence_vectors,\n flat_states,\n block[2].reshape(-1, 4)[flat_scored].astype(np.float32),\n positive_items,\n negatives,\n negative_weights,\n block[8].reshape(-1)[flat_scored],\n device,\n backward=True,\n )\n if sequence_gradient is None:\n raise RuntimeError(\"training block omitted its sequence gradient\")\n sequence_vectors.backward(sequence_gradient)\n nn.utils.clip_grad_norm_(model.parameters(), config.gradient_clip)\n optimizer.step()\n\n count = len(positive_items)\n total_loss += statistics[\"loss_sum\"]\n total_targets += count\n total_weight += statistics[\"weight_sum\"]\n total_positive += statistics[\"positive_sum\"]\n total_negative += statistics[\"negative_sum\"]\n total_pair_correct += statistics[\"pair_correct\"]\n batches += 1\n if batches % 25 == 0:\n release_accelerator_cache(device)\n if batches % 250 == 0:\n batch_bce = statistics[\"loss_sum\"] / max(\n statistics[\"weight_sum\"], 1.0e-12\n )\n running_bce = total_loss / max(total_weight, 1.0e-12)\n log(\n f\"epoch={epoch}/{config.epochs} batches={batches:,} \"\n f\"targets={total_targets:,} \"\n f\"batch_bce={batch_bce:.6f} \"\n f\"running_bce={running_bce:.6f} \"\n f\"elapsed={time.monotonic() - started:.1f}s\"\n )\n\n return {\n \"epoch\": float(epoch),\n \"loss\": total_loss / max(total_weight, 1.0e-12),\n \"positive_logit\": total_positive / max(total_targets, 1),\n \"negative_logit\": total_negative / max(total_targets * negative_count, 1),\n \"pair_accuracy\": total_pair_correct\n / max(total_targets * negative_count, 1),\n \"targets\": float(total_targets),\n \"target_weight_sum\": total_weight,\n \"batches\": float(batches),\n \"seconds\": time.monotonic() - started,\n }\n\n\n@torch.no_grad()\ndef evaluate_validation_bce(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n target_weights: np.ndarray,\n window_starts: np.ndarray,\n training_counts: np.ndarray,\n validation_counts: np.ndarray,\n item_correction: torch.Tensor,\n known_positive_packed: np.ndarray | None,\n device: torch.device,\n) -> dict[str, float]:\n \"\"\"Measure deterministic BCE on each user's chronological target suffix.\"\"\"\n model.eval()\n totals = training_counts + validation_counts\n users_with_validation = np.flatnonzero(validation_counts > 0)\n ordered = users_with_validation[\n np.argsort(totals[users_with_validation], kind=\"stable\")\n ]\n groups = [\n ordered[start : start + config.user_batch_size]\n for start in range(0, len(ordered), config.user_batch_size)\n ]\n generator = torch.Generator(device=\"cpu\").manual_seed(config.seed + 20_000)\n total_loss = 0.0\n total_targets = 0\n total_weight = 0.0\n started = time.monotonic()\n\n for group_number, users in enumerate(groups, start=1):\n group_totals = totals[users]\n group_starts = training_counts[users]\n hidden = torch.zeros(\n 1, len(users), config.embedding_dim, device=device\n )\n max_targets = int(group_totals.max())\n for start in range(0, max_targets, config.bptt_steps):",
"isError": false
},
"responseMessageId": "e5a0a7e1-e60a-4abf-8669-b554624eceae"
}
]
},
{
"id": 17,
"role": "agent",
"timestamp": "2026-09-29T05:44:02.677Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:44:02.677Z",
"callId": "read_file__call_4839036",
"name": "read_file",
"input": {
"file_path": "/task/current_model/train_current_model.py",
"start_line": 1401,
"end_line": 1600
},
"recorded": {
"id": "read_file__call_4839036",
"name": "read_file",
"args": {
"file_path": "/task/current_model/train_current_model.py",
"start_line": 1401,
"end_line": 1600
},
"result": [
{
"functionResponse": {
"id": "read_file__call_4839036",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1401-1600 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 1601.\n\n--- FILE CONTENT (truncated) ---\n steps = min(config.bptt_steps, max_targets - start)\n block = build_dense_block(\n prepared,\n target_weights,\n users,\n window_starts[users],\n group_totals,\n group_starts,\n start,\n steps,\n config,\n )\n sequence_outputs, _ = event_sequence(\n model,\n torch.from_numpy(block[0].astype(np.int64)).to(device),\n torch.from_numpy(block[1].astype(np.float32)).to(device),\n hidden,\n )\n hidden = torch.stack(\n [\n sequence_outputs[row, count - 1]\n if count\n else hidden[0, row]\n for row, count in enumerate(block[7])\n ]\n ).unsqueeze(0)\n flat_scored = block[5].reshape(-1)\n if not np.any(flat_scored):\n continue\n mask = torch.from_numpy(flat_scored).to(device)\n sequence_vectors = sequence_outputs.reshape(\n -1, config.embedding_dim\n )[mask]\n flat_states = block[3].reshape(-1)[flat_scored]\n positive_items = torch.from_numpy(\n block[4].reshape(-1)[flat_scored].astype(np.int64)\n ).to(device)\n known_history = torch.from_numpy(\n block[6].reshape(-1, config.bptt_steps)[flat_scored].astype(\n np.int64\n )\n ).to(device)\n liked_items = torch.from_numpy(\n prepared.liked_states[flat_states].astype(np.int64)\n ).to(device)\n flat_user_rows = np.broadcast_to(\n users[:, None], block[5].shape\n ).reshape(-1)[flat_scored]\n negatives, negative_weights = draw_training_negatives(\n positive_items,\n known_history,\n liked_items,\n flat_user_rows,\n known_positive_packed,\n prepared,\n config,\n item_correction,\n generator,\n device,\n )\n statistics, _ = score_target_microbatches(\n model,\n prepared,\n config,\n sequence_vectors,\n flat_states,\n block[2].reshape(-1, 4)[flat_scored].astype(np.float32),\n positive_items,\n negatives,\n negative_weights,\n block[8].reshape(-1)[flat_scored],\n device,\n backward=False,\n )\n total_loss += statistics[\"loss_sum\"]\n total_targets += len(positive_items)\n total_weight += statistics[\"weight_sum\"]\n if group_number % 25 == 0:\n release_accelerator_cache(device)\n return {\n \"validation_loss\": total_loss / max(total_weight, 1.0e-12),\n \"validation_targets\": float(total_targets),\n \"validation_target_weight_sum\": total_weight,\n \"validation_seconds\": time.monotonic() - started,\n }\n\n\ndef snapshot_model(model: nn.Module) -> dict[str, torch.Tensor]:\n return {\n name: value.detach().cpu().clone()\n for name, value in model.state_dict().items()\n }\n\n\ndef save_epoch_checkpoint(\n checkpoint_dir: Path,\n epoch: int,\n model_state: dict[str, torch.Tensor],\n config: Config,\n diagnostics: dict[str, float],\n optimizer: torch.optim.Optimizer,\n generator: torch.Generator,\n) -> Path:\n checkpoint_dir.mkdir(parents=True, exist_ok=True)\n path = checkpoint_dir / f\"epoch-{epoch:02d}.pt\"\n torch.save(\n {\n \"epoch\": epoch,\n \"config\": asdict(config),\n \"diagnostics\": diagnostics,\n \"model_state\": model_state,\n \"optimizer_state\": optimizer.state_dict(),\n \"generator_state\": generator.get_state(),\n },\n path,\n )\n log(f\"saved epoch checkpoint: {path}\")\n return path\n\n\ndef train_model(\n public: Path,\n stage: str,\n prepared: SequenceData,\n config: Config,\n device: torch.device,\n checkpoint_dir: Path | None,\n resume_checkpoint: Path | None,\n) -> tuple[base.SequentialTwoTower, list[dict[str, float]]]:\n partition = chronological_target_partition(\n prepared,\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n training_counts = partition.training_counts\n validation_counts = partition.validation_counts\n log(\n f\"chronological BCE window: train={int(training_counts.sum()):,} \"\n f\"validation={int(validation_counts.sum()):,} \"\n f\"selected={partition.selected_count:,} \"\n f\"discarded_old_targets={int(partition.window_starts.sum()):,} \"\n f\"all={prepared.target_count:,}\"\n )\n model = base.SequentialTwoTower(prepared.item_features, model_config(config)).to(\n device\n )\n optimizer = torch.optim.AdamW(\n model.parameters(), lr=config.learning_rate, weight_decay=config.weight_decay\n )\n generator = torch.Generator(device=\"cpu\").manual_seed(config.seed + 10_000)\n item_correction = build_item_correction(\n prepared,\n config,\n partition.window_starts,\n training_counts,\n )\n target_weights = build_target_weights(\n public,\n stage,\n prepared,\n partition.history_training_counts,\n config,\n )\n known_positive_packed = (\n build_known_positive_filter(prepared, partition.history_training_counts)\n if config.negative_filter_scope == \"full-training-history\"\n else None\n )\n history: list[dict[str, float]] = []\n best_loss = float(\"inf\")\n best_epoch = 0\n best_state: dict[str, torch.Tensor] | None = None\n epochs_without_improvement = 0\n start_epoch = 0\n\n if resume_checkpoint is not None:\n payload = torch.load(\n resume_checkpoint, map_location=device, weights_only=False\n )\n stored_config = payload.get(\"config\", {})\n architecture_fields = (\n \"bptt_steps\",\n \"embedding_dim\",\n \"hidden_dim\",\n \"user_hidden_1\",\n \"user_hidden_2\",\n \"item_hidden_1\",\n \"item_hidden_2\",\n )\n mismatches = {\n name: (stored_config.get(name), getattr(config, name))\n for name in architecture_fields\n if stored_config.get(name) != getattr(config, name)\n }\n if mismatches:\n raise RuntimeError(\n f\"resume checkpoint architecture differs: {mismatches}\"\n )\n stored_filter_scope = stored_config.get("
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:44:02.681Z",
"resultDisplay": "Read lines 1401-1600 of 2165 from ../task/current_model/train_current_model.py",
"description": "../.../train_current_model.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:44:02.683Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1401-1600 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 1601.\n\n--- FILE CONTENT (truncated) ---\n steps = min(config.bptt_steps, max_targets - start)\n block = build_dense_block(\n prepared,\n target_weights,\n users,\n window_starts[users],\n group_totals,\n group_starts,\n start,\n steps,\n config,\n )\n sequence_outputs, _ = event_sequence(\n model,\n torch.from_numpy(block[0].astype(np.int64)).to(device),\n torch.from_numpy(block[1].astype(np.float32)).to(device),\n hidden,\n )\n hidden = torch.stack(\n [\n sequence_outputs[row, count - 1]\n if count\n else hidden[0, row]\n for row, count in enumerate(block[7])\n ]\n ).unsqueeze(0)\n flat_scored = block[5].reshape(-1)\n if not np.any(flat_scored):\n continue\n mask = torch.from_numpy(flat_scored).to(device)\n sequence_vectors = sequence_outputs.reshape(\n -1, config.embedding_dim\n )[mask]\n flat_states = block[3].reshape(-1)[flat_scored]\n positive_items = torch.from_numpy(\n block[4].reshape(-1)[flat_scored].astype(np.int64)\n ).to(device)\n known_history = torch.from_numpy(\n block[6].reshape(-1, config.bptt_steps)[flat_scored].astype(\n np.int64\n )\n ).to(device)\n liked_items = torch.from_numpy(\n prepared.liked_states[flat_states].astype(np.int64)\n ).to(device)\n flat_user_rows = np.broadcast_to(\n users[:, None], block[5].shape\n ).reshape(-1)[flat_scored]\n negatives, negative_weights = draw_training_negatives(\n positive_items,\n known_history,\n liked_items,\n flat_user_rows,\n known_positive_packed,\n prepared,\n config,\n item_correction,\n generator,\n device,\n )\n statistics, _ = score_target_microbatches(\n model,\n prepared,\n config,\n sequence_vectors,\n flat_states,\n block[2].reshape(-1, 4)[flat_scored].astype(np.float32),\n positive_items,\n negatives,\n negative_weights,\n block[8].reshape(-1)[flat_scored],\n device,\n backward=False,\n )\n total_loss += statistics[\"loss_sum\"]\n total_targets += len(positive_items)\n total_weight += statistics[\"weight_sum\"]\n if group_number % 25 == 0:\n release_accelerator_cache(device)\n return {\n \"validation_loss\": total_loss / max(total_weight, 1.0e-12),\n \"validation_targets\": float(total_targets),\n \"validation_target_weight_sum\": total_weight,\n \"validation_seconds\": time.monotonic() - started,\n }\n\n\ndef snapshot_model(model: nn.Module) -> dict[str, torch.Tensor]:\n return {\n name: value.detach().cpu().clone()\n for name, value in model.state_dict().items()\n }\n\n\ndef save_epoch_checkpoint(\n checkpoint_dir: Path,\n epoch: int,\n model_state: dict[str, torch.Tensor],\n config: Config,\n diagnostics: dict[str, float],\n optimizer: torch.optim.Optimizer,\n generator: torch.Generator,\n) -> Path:\n checkpoint_dir.mkdir(parents=True, exist_ok=True)\n path = checkpoint_dir / f\"epoch-{epoch:02d}.pt\"\n torch.save(\n {\n \"epoch\": epoch,\n \"config\": asdict(config),\n \"diagnostics\": diagnostics,\n \"model_state\": model_state,\n \"optimizer_state\": optimizer.state_dict(),\n \"generator_state\": generator.get_state(),\n },\n path,\n )\n log(f\"saved epoch checkpoint: {path}\")\n return path\n\n\ndef train_model(\n public: Path,\n stage: str,\n prepared: SequenceData,\n config: Config,\n device: torch.device,\n checkpoint_dir: Path | None,\n resume_checkpoint: Path | None,\n) -> tuple[base.SequentialTwoTower, list[dict[str, float]]]:\n partition = chronological_target_partition(\n prepared,\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n training_counts = partition.training_counts\n validation_counts = partition.validation_counts\n log(\n f\"chronological BCE window: train={int(training_counts.sum()):,} \"\n f\"validation={int(validation_counts.sum()):,} \"\n f\"selected={partition.selected_count:,} \"\n f\"discarded_old_targets={int(partition.window_starts.sum()):,} \"\n f\"all={prepared.target_count:,}\"\n )\n model = base.SequentialTwoTower(prepared.item_features, model_config(config)).to(\n device\n )\n optimizer = torch.optim.AdamW(\n model.parameters(), lr=config.learning_rate, weight_decay=config.weight_decay\n )\n generator = torch.Generator(device=\"cpu\").manual_seed(config.seed + 10_000)\n item_correction = build_item_correction(\n prepared,\n config,\n partition.window_starts,\n training_counts,\n )\n target_weights = build_target_weights(\n public,\n stage,\n prepared,\n partition.history_training_counts,\n config,\n )\n known_positive_packed = (\n build_known_positive_filter(prepared, partition.history_training_counts)\n if config.negative_filter_scope == \"full-training-history\"\n else None\n )\n history: list[dict[str, float]] = []\n best_loss = float(\"inf\")\n best_epoch = 0\n best_state: dict[str, torch.Tensor] | None = None\n epochs_without_improvement = 0\n start_epoch = 0\n\n if resume_checkpoint is not None:\n payload = torch.load(\n resume_checkpoint, map_location=device, weights_only=False\n )\n stored_config = payload.get(\"config\", {})\n architecture_fields = (\n \"bptt_steps\",\n \"embedding_dim\",\n \"hidden_dim\",\n \"user_hidden_1\",\n \"user_hidden_2\",\n \"item_hidden_1\",\n \"item_hidden_2\",\n )\n mismatches = {\n name: (stored_config.get(name), getattr(config, name))\n for name in architecture_fields\n if stored_config.get(name) != getattr(config, name)\n }\n if mismatches:\n raise RuntimeError(\n f\"resume checkpoint architecture differs: {mismatches}\"\n )\n stored_filter_scope = stored_config.get(",
"isError": false
},
"responseMessageId": "5e831001-5ed5-4141-b166-896c63fd4ed5"
}
]
},
{
"id": 18,
"role": "agent",
"timestamp": "2026-09-29T05:44:05.153Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:44:05.153Z",
"callId": "read_file__call_3327968",
"name": "read_file",
"input": {
"end_line": 1900,
"start_line": 1750,
"file_path": "/task/current_model/train_current_model.py"
},
"recorded": {
"id": "read_file__call_3327968",
"name": "read_file",
"args": {
"end_line": 1900,
"start_line": 1750,
"file_path": "/task/current_model/train_current_model.py"
},
"result": [
{
"functionResponse": {
"id": "read_file__call_3327968",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1750-1900 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 1901.\n\n--- FILE CONTENT (truncated) ---\n )\n else:\n epochs_without_improvement += 1\n log(\n f\"epoch={epoch}/{config.epochs} train_loss={diagnostics['loss']:.5f} \"\n f\"validation_loss={diagnostics['validation_loss']:.5f} \"\n f\"pair_accuracy={diagnostics['pair_accuracy']:.3f} \"\n f\"train_targets={int(diagnostics['targets']):,} \"\n f\"validation_targets={int(diagnostics['validation_targets']):,} \"\n f\"train_seconds={diagnostics['seconds']:.1f} \"\n f\"validation_seconds={diagnostics['validation_seconds']:.1f}\"\n )\n release_accelerator_cache(device)\n if epochs_without_improvement >= config.early_stopping_patience:\n log(\n f\"early stopping after epoch={epoch}; \"\n f\"best_epoch={best_epoch} best_validation_loss={best_loss:.6f}\"\n )\n break\n\n if best_state is None:\n raise RuntimeError(\"training did not produce a valid checkpoint\")\n model.load_state_dict(best_state)\n model.to(device).eval()\n log(\n f\"restored best validation-BCE checkpoint: epoch={best_epoch} \"\n f\"loss={best_loss:.6f}\"\n )\n return model, history\n\n\n@torch.inference_mode()\ndef inference_queries(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n device: torch.device,\n) -> torch.Tensor:\n sequence_rows = prepared.inference_user_rows\n lengths = np.where(\n sequence_rows >= 0, prepared.lengths[np.maximum(sequence_rows, 0)], 0\n )\n ordered = np.argsort(lengths, kind=\"stable\")\n result = torch.zeros(\n len(sequence_rows), config.embedding_dim, dtype=torch.float32\n )\n for group_start in range(0, len(ordered), config.user_batch_size):\n output_rows = ordered[group_start : group_start + config.user_batch_size]\n rows = sequence_rows[output_rows]\n group_lengths = lengths[output_rows]\n hidden = torch.zeros(\n 1, len(output_rows), config.embedding_dim, device=device\n )\n for start in range(0, int(group_lengths.max(initial=0)), config.bptt_steps):\n steps = min(\n config.bptt_steps,\n int(group_lengths.max(initial=0)) - start,\n )\n items = np.full(\n (len(output_rows), steps),\n len(prepared.item_ids),\n dtype=np.uint16,\n )\n features = np.zeros((len(output_rows), steps, 3), dtype=np.float16)\n counts = np.zeros(len(output_rows), dtype=np.int32)\n for local_row, sequence_row in enumerate(rows):\n if sequence_row < 0:\n continue\n count = min(steps, max(int(group_lengths[local_row]) - start, 0))\n if not count:\n continue\n offset = int(prepared.offsets[sequence_row])\n selected = slice(offset + start, offset + start + count)\n items[local_row, :count] = prepared.sequence_items[selected]\n features[local_row, :count] = prepared.event_features[selected]\n counts[local_row] = count\n outputs, _ = event_sequence(\n model,\n torch.from_numpy(items.astype(np.int64)).to(device),\n torch.from_numpy(features.astype(np.float32)).to(device),\n hidden,\n )\n next_rows: list[torch.Tensor] = []\n for local_row, count in enumerate(counts):\n next_rows.append(\n outputs[local_row, count - 1]\n if count\n else hidden[0, local_row]\n )\n hidden = torch.stack(next_rows).unsqueeze(0)\n\n liked_items = torch.from_numpy(\n prepared.inference_liked_items[output_rows].astype(np.int64)\n ).to(device)\n disliked_items = torch.from_numpy(\n prepared.inference_disliked_items[output_rows].astype(np.int64)\n ).to(device)\n queries = user_queries(\n model,\n hidden.squeeze(0),\n model._mean_context_channel(liked_items),\n model._mean_context_channel(disliked_items),\n torch.from_numpy(\n prepared.inference_user_features[output_rows].astype(np.float32)\n ).to(device),\n )\n result[torch.from_numpy(output_rows.astype(np.int64))] = queries.cpu()\n return result\n\n\n@torch.inference_mode()\ndef rank_candidates(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n eligibility: CandidateSet,\n config: Config,\n device: torch.device,\n ranking_batch_size: int,\n) -> tuple[pd.DataFrame, dict[str, float]]:\n model.eval()\n queries = inference_queries(model, prepared, config, device)\n item_parts: list[torch.Tensor] = []\n for start in range(0, len(prepared.item_ids), 2_048):\n rows = torch.arange(\n start,\n min(start + 2_048, len(prepared.item_ids)),\n dtype=torch.long,\n device=device,\n )\n item_parts.append(model.encode_items(rows).cpu())\n items = torch.cat(item_parts)\n rows: list[tuple[int, int, int]] = []\n for start in range(0, len(eligibility.uids), ranking_batch_size):\n end = min(start + ranking_batch_size, len(eligibility.uids))\n scores = queries[start:end] @ items.T\n eligible = np.unpackbits(\n eligibility.packed[start:end],\n axis=1,\n count=len(prepared.item_ids),\n bitorder=\"little\",\n ).astype(bool, copy=False)\n scores.masked_fill_(~torch.from_numpy(eligible), -torch.inf)\n selected = torch.topk(scores, k=base.TOP_N, dim=1).indices.numpy()\n for local_row, columns in enumerate(selected):\n uid = int(eligibility.uids[start + local_row])\n rows.extend(\n (uid, int(prepared.item_ids[column]), rank)\n for rank, column in enumerate(columns, start=1)\n )\n if end == len(eligibility.uids) or end % 1_024 == 0:\n log(f\"ranked users={end:,}/{len(eligibility.uids):,}\")"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:44:05.161Z",
"resultDisplay": "Read lines 1750-1900 of 2165 from ../task/current_model/train_current_model.py",
"description": "../.../train_current_model.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:44:05.162Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1750-1900 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 1901.\n\n--- FILE CONTENT (truncated) ---\n )\n else:\n epochs_without_improvement += 1\n log(\n f\"epoch={epoch}/{config.epochs} train_loss={diagnostics['loss']:.5f} \"\n f\"validation_loss={diagnostics['validation_loss']:.5f} \"\n f\"pair_accuracy={diagnostics['pair_accuracy']:.3f} \"\n f\"train_targets={int(diagnostics['targets']):,} \"\n f\"validation_targets={int(diagnostics['validation_targets']):,} \"\n f\"train_seconds={diagnostics['seconds']:.1f} \"\n f\"validation_seconds={diagnostics['validation_seconds']:.1f}\"\n )\n release_accelerator_cache(device)\n if epochs_without_improvement >= config.early_stopping_patience:\n log(\n f\"early stopping after epoch={epoch}; \"\n f\"best_epoch={best_epoch} best_validation_loss={best_loss:.6f}\"\n )\n break\n\n if best_state is None:\n raise RuntimeError(\"training did not produce a valid checkpoint\")\n model.load_state_dict(best_state)\n model.to(device).eval()\n log(\n f\"restored best validation-BCE checkpoint: epoch={best_epoch} \"\n f\"loss={best_loss:.6f}\"\n )\n return model, history\n\n\n@torch.inference_mode()\ndef inference_queries(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n device: torch.device,\n) -> torch.Tensor:\n sequence_rows = prepared.inference_user_rows\n lengths = np.where(\n sequence_rows >= 0, prepared.lengths[np.maximum(sequence_rows, 0)], 0\n )\n ordered = np.argsort(lengths, kind=\"stable\")\n result = torch.zeros(\n len(sequence_rows), config.embedding_dim, dtype=torch.float32\n )\n for group_start in range(0, len(ordered), config.user_batch_size):\n output_rows = ordered[group_start : group_start + config.user_batch_size]\n rows = sequence_rows[output_rows]\n group_lengths = lengths[output_rows]\n hidden = torch.zeros(\n 1, len(output_rows), config.embedding_dim, device=device\n )\n for start in range(0, int(group_lengths.max(initial=0)), config.bptt_steps):\n steps = min(\n config.bptt_steps,\n int(group_lengths.max(initial=0)) - start,\n )\n items = np.full(\n (len(output_rows), steps),\n len(prepared.item_ids),\n dtype=np.uint16,\n )\n features = np.zeros((len(output_rows), steps, 3), dtype=np.float16)\n counts = np.zeros(len(output_rows), dtype=np.int32)\n for local_row, sequence_row in enumerate(rows):\n if sequence_row < 0:\n continue\n count = min(steps, max(int(group_lengths[local_row]) - start, 0))\n if not count:\n continue\n offset = int(prepared.offsets[sequence_row])\n selected = slice(offset + start, offset + start + count)\n items[local_row, :count] = prepared.sequence_items[selected]\n features[local_row, :count] = prepared.event_features[selected]\n counts[local_row] = count\n outputs, _ = event_sequence(\n model,\n torch.from_numpy(items.astype(np.int64)).to(device),\n torch.from_numpy(features.astype(np.float32)).to(device),\n hidden,\n )\n next_rows: list[torch.Tensor] = []\n for local_row, count in enumerate(counts):\n next_rows.append(\n outputs[local_row, count - 1]\n if count\n else hidden[0, local_row]\n )\n hidden = torch.stack(next_rows).unsqueeze(0)\n\n liked_items = torch.from_numpy(\n prepared.inference_liked_items[output_rows].astype(np.int64)\n ).to(device)\n disliked_items = torch.from_numpy(\n prepared.inference_disliked_items[output_rows].astype(np.int64)\n ).to(device)\n queries = user_queries(\n model,\n hidden.squeeze(0),\n model._mean_context_channel(liked_items),\n model._mean_context_channel(disliked_items),\n torch.from_numpy(\n prepared.inference_user_features[output_rows].astype(np.float32)\n ).to(device),\n )\n result[torch.from_numpy(output_rows.astype(np.int64))] = queries.cpu()\n return result\n\n\n@torch.inference_mode()\ndef rank_candidates(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n eligibility: CandidateSet,\n config: Config,\n device: torch.device,\n ranking_batch_size: int,\n) -> tuple[pd.DataFrame, dict[str, float]]:\n model.eval()\n queries = inference_queries(model, prepared, config, device)\n item_parts: list[torch.Tensor] = []\n for start in range(0, len(prepared.item_ids), 2_048):\n rows = torch.arange(\n start,\n min(start + 2_048, len(prepared.item_ids)),\n dtype=torch.long,\n device=device,\n )\n item_parts.append(model.encode_items(rows).cpu())\n items = torch.cat(item_parts)\n rows: list[tuple[int, int, int]] = []\n for start in range(0, len(eligibility.uids), ranking_batch_size):\n end = min(start + ranking_batch_size, len(eligibility.uids))\n scores = queries[start:end] @ items.T\n eligible = np.unpackbits(\n eligibility.packed[start:end],\n axis=1,\n count=len(prepared.item_ids),\n bitorder=\"little\",\n ).astype(bool, copy=False)\n scores.masked_fill_(~torch.from_numpy(eligible), -torch.inf)\n selected = torch.topk(scores, k=base.TOP_N, dim=1).indices.numpy()\n for local_row, columns in enumerate(selected):\n uid = int(eligibility.uids[start + local_row])\n rows.extend(\n (uid, int(prepared.item_ids[column]), rank)\n for rank, column in enumerate(columns, start=1)\n )\n if end == len(eligibility.uids) or end % 1_024 == 0:\n log(f\"ranked users={end:,}/{len(eligibility.uids):,}\")",
"isError": false
},
"responseMessageId": "31667437-787e-42c6-8b49-171d9233b334"
}
]
},
{
"id": 19,
"role": "agent",
"timestamp": "2026-09-29T05:44:09.815Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:44:09.815Z",
"callId": "read_file__call_5692046",
"name": "read_file",
"input": {
"end_line": 350,
"start_line": 150,
"file_path": "/task/current_model/two_tower_model.py"
},
"recorded": {
"id": "read_file__call_5692046",
"name": "read_file",
"args": {
"end_line": 350,
"start_line": 150,
"file_path": "/task/current_model/two_tower_model.py"
},
"result": [
{
"functionResponse": {
"id": "read_file__call_5692046",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 150-350 of 434 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 351.\n\n--- FILE CONTENT (truncated) ---\n pending_uid = uid\n pending = []\n pending.append(part)\n if pending_uid is not None:\n yield pending_uid, _combine_listen_chunks(pending)\n\n\ndef merge_listen_streams(\n paths: list[Path], item_index: pd.Index\n) -> Iterator[tuple[int, ListenEvents]]:\n \"\"\"Merge uid-sorted public listen files without materializing them.\"\"\"\n streams = [iter(iter_listen_users(path, item_index)) for path in paths]\n heads: list[tuple[int, ListenEvents] | None] = [\n next(stream, None) for stream in streams\n ]\n while any(head is not None for head in heads):\n uid = min(head[0] for head in heads if head is not None)\n parts: list[ListenEvents] = []\n for position, head in enumerate(heads):\n if head is not None and head[0] == uid:\n parts.append(head[1])\n heads[position] = next(streams[position], None)\n yield uid, _combine_listen_chunks(parts)\n\n\ndef load_explicit_events(\n data: Path,\n prefixes: tuple[str, ...],\n item_index: pd.Index,\n) -> dict[int, ExplicitEvents]:\n \"\"\"Load public explicit actions as chronological state changes.\"\"\"\n specifications = (\n (\"likes\", 0, 1, 0),\n (\"unlikes\", 0, 0, 1),\n (\"dislikes\", 1, 1, 0),\n (\"undislikes\", 1, 0, 1),\n )\n parts: list[pd.DataFrame] = []\n for family, channel, active, tie_priority in specifications:\n for prefix in prefixes:\n frame = pd.read_parquet(\n data / f\"{prefix}_{family}.parquet\",\n columns=[\"uid\", \"timestamp\", \"item_id\", \"is_organic\"],\n )\n mapped = item_index.get_indexer(\n frame.item_id.to_numpy(dtype=np.int64, copy=False)\n )\n keep = mapped >= 0\n if not np.any(keep):\n continue\n retained = frame.loc[keep, [\"uid\", \"timestamp\", \"is_organic\"]].copy()\n retained[\"item\"] = mapped[keep].astype(np.int32, copy=False)\n retained[\"channel\"] = np.int8(channel)\n retained[\"active\"] = np.int8(active)\n retained[\"tie_priority\"] = np.int8(tie_priority)\n retained[\"organic_like\"] = np.int8(\n family == \"likes\"\n ) * retained.is_organic.astype(np.int8)\n parts.append(retained.drop(columns=\"is_organic\"))\n if not parts:\n return {}\n events = pd.concat(parts, ignore_index=True)\n events = events.sort_values(\n [\"uid\", \"timestamp\", \"tie_priority\"], kind=\"mergesort\"\n ).reset_index(drop=True)\n result: dict[int, ExplicitEvents] = {}\n for uid, rows in events.groupby(\"uid\", sort=False):\n result[int(uid)] = ExplicitEvents(\n timestamps=rows.timestamp.to_numpy(dtype=np.int64, copy=True),\n items=rows.item.to_numpy(dtype=np.int32, copy=True),\n channel=rows.channel.to_numpy(dtype=np.int8, copy=True),\n active=rows.active.to_numpy(dtype=np.int8, copy=True),\n organic_like=rows.organic_like.to_numpy(dtype=np.int8, copy=True),\n )\n log(f\"loaded explicit events: rows={len(events):,} users={len(result):,}\")\n return result\n\n\ndef empty_explicit_events() -> ExplicitEvents:\n return ExplicitEvents(\n timestamps=np.empty(0, dtype=np.int64),\n items=np.empty(0, dtype=np.int32),\n channel=np.empty(0, dtype=np.int8),\n active=np.empty(0, dtype=np.int8),\n organic_like=np.empty(0, dtype=np.int8),\n )\n\n\ndef pack_active_items(\n active: dict[int, int], width: int, padding_item: int\n) -> np.ndarray:\n packed = np.full(width, padding_item, dtype=np.uint16)\n if not active:\n return packed\n ordered = sorted(active.items(), key=lambda pair: (-pair[1], pair[0]))[:width]\n packed[: len(ordered)] = np.fromiter(\n (item for item, _ in ordered), dtype=np.uint16, count=len(ordered)\n )\n return packed\n\n\ndef load_inference_state(\n path: Path,\n item_index: pd.Index,\n uids: np.ndarray,\n n_items: int,\n) -> tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray]:\n \"\"\"Pack the task-provided active explicit state.\"\"\"\n state = pd.read_parquet(path, columns=[\"uid\", \"item_id\", \"state\"])\n mapped = item_index.get_indexer(state.item_id.to_numpy(dtype=np.int64))\n keep = mapped >= 0\n state = state.loc[keep, [\"uid\", \"state\"]].copy()\n state[\"item\"] = mapped[keep].astype(np.int32, copy=False)\n user_rows = pd.Index(uids)\n liked = np.full((len(uids), LIKE_ITEMS), n_items, dtype=np.uint16)\n disliked = np.full((len(uids), DISLIKE_ITEMS), n_items, dtype=np.uint16)\n like_counts = np.zeros(len(uids), dtype=np.int32)\n dislike_counts = np.zeros(len(uids), dtype=np.int32)\n for value, width, destination, counts in (\n (\"liked\", LIKE_ITEMS, liked, like_counts),\n (\"disliked\", DISLIKE_ITEMS, disliked, dislike_counts),\n ):\n selected = state[state.state == value]\n for uid, rows in selected.groupby(\"uid\", sort=False):\n row = int(user_rows.get_indexer([int(uid)])[0])\n if row < 0:\n continue\n items = np.sort(rows.item.unique())\n counts[row] = len(items)\n retained = items[:width]\n destination[row, : len(retained)] = retained.astype(np.uint16)\n return liked, disliked, like_counts, dislike_counts\n\n\nclass SequentialTwoTower(nn.Module):\n \"\"\"Order-sensitive user tower and independent catalog-aware item tower.\"\"\"\n\n def __init__(self, item_features: np.ndarray, config: TowerConfig) -> None:\n super().__init__()\n n_items, n_item_features = item_features.shape\n dim = config.embedding_dim\n self.n_items = n_items\n self.context_embedding = nn.Embedding(\n n_items + 1, dim, padding_idx=n_items\n )\n self.event_projection = nn.Sequential(\n nn.Linear(dim + 3, config.hidden_dim),\n nn.GELU(),\n )\n self.sequence_encoder = nn.GRU(\n input_size=config.hidden_dim,\n hidden_size=dim,\n num_layers=1,\n batch_first=True,\n )\n self.user_tower = nn.Sequential(\n nn.Linear(dim * 3 + 6, config.user_hidden_1),\n nn.GELU(),\n nn.Dropout(config.dropout),\n nn.Linear(config.user_hidden_1, config.user_hidden_2),\n nn.GELU(),\n nn.Dropout(config.dropout),\n nn.Linear(config.user_hidden_2, dim),\n )\n self.candidate_embedding = nn.Embedding(n_items, dim)\n self.register_buffer(\n \"item_features\", torch.from_numpy(item_features), persistent=True\n )\n self.item_tower = nn.Sequential(\n nn.Linear(dim + n_item_features, config.item_hidden_1),\n nn.GELU(),\n nn.Dropout(config.dropout),\n nn.Linear(config.item_hidden_1, config.item_hidden_2),\n nn.GELU(),\n nn.Dropout(config.dropout),\n nn.Linear(config.item_hidden_2, dim),\n )\n nn.init.normal_(self.context_embedding.weight, std=0.02)\n nn.init.normal_(self.candidate_embedding.weight, std=0.02)\n with torch.no_grad():\n self.context_embedding.weight[n_items].zero_()\n\n def _mean_context_channel(self, items: torch.Tensor) -> torch.Tensor:\n mask = items.ne(self.n_items)\n vectors = self.context_embedding(items)\n return (vectors * mask.unsqueeze(-1)).sum(dim=1) / mask.sum(\n dim=1, keepdim=True\n ).clamp_min(1)\n\n def encode_items(self, items: torch.Tensor) -> torch.Tensor:\n inputs = torch.cat(\n [self.candidate_embedding(items), self.item_features[items]], dim=1\n )\n return self.item_tower(inputs)\n\n\ndef invalid_negative_mask(\n negatives: torch.Tensor,\n targets: torch.Tensor,\n context_items: torch.Tensor,\n liked_items: torch.Tensor,"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:44:09.820Z",
"resultDisplay": "Read lines 150-350 of 434 from ../task/current_model/two_tower_model.py",
"description": "../.../two_tower_model.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:44:09.821Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 150-350 of 434 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 351.\n\n--- FILE CONTENT (truncated) ---\n pending_uid = uid\n pending = []\n pending.append(part)\n if pending_uid is not None:\n yield pending_uid, _combine_listen_chunks(pending)\n\n\ndef merge_listen_streams(\n paths: list[Path], item_index: pd.Index\n) -> Iterator[tuple[int, ListenEvents]]:\n \"\"\"Merge uid-sorted public listen files without materializing them.\"\"\"\n streams = [iter(iter_listen_users(path, item_index)) for path in paths]\n heads: list[tuple[int, ListenEvents] | None] = [\n next(stream, None) for stream in streams\n ]\n while any(head is not None for head in heads):\n uid = min(head[0] for head in heads if head is not None)\n parts: list[ListenEvents] = []\n for position, head in enumerate(heads):\n if head is not None and head[0] == uid:\n parts.append(head[1])\n heads[position] = next(streams[position], None)\n yield uid, _combine_listen_chunks(parts)\n\n\ndef load_explicit_events(\n data: Path,\n prefixes: tuple[str, ...],\n item_index: pd.Index,\n) -> dict[int, ExplicitEvents]:\n \"\"\"Load public explicit actions as chronological state changes.\"\"\"\n specifications = (\n (\"likes\", 0, 1, 0),\n (\"unlikes\", 0, 0, 1),\n (\"dislikes\", 1, 1, 0),\n (\"undislikes\", 1, 0, 1),\n )\n parts: list[pd.DataFrame] = []\n for family, channel, active, tie_priority in specifications:\n for prefix in prefixes:\n frame = pd.read_parquet(\n data / f\"{prefix}_{family}.parquet\",\n columns=[\"uid\", \"timestamp\", \"item_id\", \"is_organic\"],\n )\n mapped = item_index.get_indexer(\n frame.item_id.to_numpy(dtype=np.int64, copy=False)\n )\n keep = mapped >= 0\n if not np.any(keep):\n continue\n retained = frame.loc[keep, [\"uid\", \"timestamp\", \"is_organic\"]].copy()\n retained[\"item\"] = mapped[keep].astype(np.int32, copy=False)\n retained[\"channel\"] = np.int8(channel)\n retained[\"active\"] = np.int8(active)\n retained[\"tie_priority\"] = np.int8(tie_priority)\n retained[\"organic_like\"] = np.int8(\n family == \"likes\"\n ) * retained.is_organic.astype(np.int8)\n parts.append(retained.drop(columns=\"is_organic\"))\n if not parts:\n return {}\n events = pd.concat(parts, ignore_index=True)\n events = events.sort_values(\n [\"uid\", \"timestamp\", \"tie_priority\"], kind=\"mergesort\"\n ).reset_index(drop=True)\n result: dict[int, ExplicitEvents] = {}\n for uid, rows in events.groupby(\"uid\", sort=False):\n result[int(uid)] = ExplicitEvents(\n timestamps=rows.timestamp.to_numpy(dtype=np.int64, copy=True),\n items=rows.item.to_numpy(dtype=np.int32, copy=True),\n channel=rows.channel.to_numpy(dtype=np.int8, copy=True),\n active=rows.active.to_numpy(dtype=np.int8, copy=True),\n organic_like=rows.organic_like.to_numpy(dtype=np.int8, copy=True),\n )\n log(f\"loaded explicit events: rows={len(events):,} users={len(result):,}\")\n return result\n\n\ndef empty_explicit_events() -> ExplicitEvents:\n return ExplicitEvents(\n timestamps=np.empty(0, dtype=np.int64),\n items=np.empty(0, dtype=np.int32),\n channel=np.empty(0, dtype=np.int8),\n active=np.empty(0, dtype=np.int8),\n organic_like=np.empty(0, dtype=np.int8),\n )\n\n\ndef pack_active_items(\n active: dict[int, int], width: int, padding_item: int\n) -> np.ndarray:\n packed = np.full(width, padding_item, dtype=np.uint16)\n if not active:\n return packed\n ordered = sorted(active.items(), key=lambda pair: (-pair[1], pair[0]))[:width]\n packed[: len(ordered)] = np.fromiter(\n (item for item, _ in ordered), dtype=np.uint16, count=len(ordered)\n )\n return packed\n\n\ndef load_inference_state(\n path: Path,\n item_index: pd.Index,\n uids: np.ndarray,\n n_items: int,\n) -> tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray]:\n \"\"\"Pack the task-provided active explicit state.\"\"\"\n state = pd.read_parquet(path, columns=[\"uid\", \"item_id\", \"state\"])\n mapped = item_index.get_indexer(state.item_id.to_numpy(dtype=np.int64))\n keep = mapped >= 0\n state = state.loc[keep, [\"uid\", \"state\"]].copy()\n state[\"item\"] = mapped[keep].astype(np.int32, copy=False)\n user_rows = pd.Index(uids)\n liked = np.full((len(uids), LIKE_ITEMS), n_items, dtype=np.uint16)\n disliked = np.full((len(uids), DISLIKE_ITEMS), n_items, dtype=np.uint16)\n like_counts = np.zeros(len(uids), dtype=np.int32)\n dislike_counts = np.zeros(len(uids), dtype=np.int32)\n for value, width, destination, counts in (\n (\"liked\", LIKE_ITEMS, liked, like_counts),\n (\"disliked\", DISLIKE_ITEMS, disliked, dislike_counts),\n ):\n selected = state[state.state == value]\n for uid, rows in selected.groupby(\"uid\", sort=False):\n row = int(user_rows.get_indexer([int(uid)])[0])\n if row < 0:\n continue\n items = np.sort(rows.item.unique())\n counts[row] = len(items)\n retained = items[:width]\n destination[row, : len(retained)] = retained.astype(np.uint16)\n return liked, disliked, like_counts, dislike_counts\n\n\nclass SequentialTwoTower(nn.Module):\n \"\"\"Order-sensitive user tower and independent catalog-aware item tower.\"\"\"\n\n def __init__(self, item_features: np.ndarray, config: TowerConfig) -> None:\n super().__init__()\n n_items, n_item_features = item_features.shape\n dim = config.embedding_dim\n self.n_items = n_items\n self.context_embedding = nn.Embedding(\n n_items + 1, dim, padding_idx=n_items\n )\n self.event_projection = nn.Sequential(\n nn.Linear(dim + 3, config.hidden_dim),\n nn.GELU(),\n )\n self.sequence_encoder = nn.GRU(\n input_size=config.hidden_dim,\n hidden_size=dim,\n num_layers=1,\n batch_first=True,\n )\n self.user_tower = nn.Sequential(\n nn.Linear(dim * 3 + 6, config.user_hidden_1),\n nn.GELU(),\n nn.Dropout(config.dropout),\n nn.Linear(config.user_hidden_1, config.user_hidden_2),\n nn.GELU(),\n nn.Dropout(config.dropout),\n nn.Linear(config.user_hidden_2, dim),\n )\n self.candidate_embedding = nn.Embedding(n_items, dim)\n self.register_buffer(\n \"item_features\", torch.from_numpy(item_features), persistent=True\n )\n self.item_tower = nn.Sequential(\n nn.Linear(dim + n_item_features, config.item_hidden_1),\n nn.GELU(),\n nn.Dropout(config.dropout),\n nn.Linear(config.item_hidden_1, config.item_hidden_2),\n nn.GELU(),\n nn.Dropout(config.dropout),\n nn.Linear(config.item_hidden_2, dim),\n )\n nn.init.normal_(self.context_embedding.weight, std=0.02)\n nn.init.normal_(self.candidate_embedding.weight, std=0.02)\n with torch.no_grad():\n self.context_embedding.weight[n_items].zero_()\n\n def _mean_context_channel(self, items: torch.Tensor) -> torch.Tensor:\n mask = items.ne(self.n_items)\n vectors = self.context_embedding(items)\n return (vectors * mask.unsqueeze(-1)).sum(dim=1) / mask.sum(\n dim=1, keepdim=True\n ).clamp_min(1)\n\n def encode_items(self, items: torch.Tensor) -> torch.Tensor:\n inputs = torch.cat(\n [self.candidate_embedding(items), self.item_features[items]], dim=1\n )\n return self.item_tower(inputs)\n\n\ndef invalid_negative_mask(\n negatives: torch.Tensor,\n targets: torch.Tensor,\n context_items: torch.Tensor,\n liked_items: torch.Tensor,",
"isError": false
},
"responseMessageId": "d2c008ba-fec7-4018-82f0-ea8d65e3c02c"
}
]
},
{
"id": 20,
"role": "agent",
"timestamp": "2026-09-29T05:44:11.496Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:44:11.496Z",
"callId": "read_file__call_158321",
"name": "read_file",
"input": {
"start_line": 350,
"end_line": 434,
"file_path": "/task/current_model/two_tower_model.py"
},
"recorded": {
"id": "read_file__call_158321",
"name": "read_file",
"args": {
"start_line": 350,
"end_line": 434,
"file_path": "/task/current_model/two_tower_model.py"
},
"result": [
{
"functionResponse": {
"id": "read_file__call_158321",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 350-434 of 434 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 435.\n\n--- FILE CONTENT (truncated) ---\n liked_items: torch.Tensor,\n) -> torch.Tensor:\n \"\"\"Identify targets, known positives, active likes, and row duplicates.\"\"\"\n invalid = negatives.eq(targets[:, None])\n invalid |= negatives[:, :, None].eq(context_items[:, None, :]).any(dim=2)\n invalid |= negatives[:, :, None].eq(liked_items[:, None, :]).any(dim=2)\n for column in range(1, negatives.shape[1]):\n invalid[:, column] |= negatives[:, column, None].eq(\n negatives[:, :column]\n ).any(dim=1)\n return invalid\n\n\ndef draw_filtered_mixed_negatives(\n targets: torch.Tensor,\n context_items: torch.Tensor,\n liked_items: torch.Tensor,\n n_items: int,\n uniform_count: int,\n inbatch_count: int,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n \"\"\"Draw uniform and unique-target in-batch negatives.\"\"\"\n batch_size = len(targets)\n uniform = torch.randint(\n n_items,\n (batch_size, uniform_count),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.empty((batch_size, 0), dtype=torch.long, device=\"cpu\")\n corrected_count = 0\n if inbatch_count:\n unique_targets = torch.unique(targets.detach().cpu())\n corrected_count = min(inbatch_count, max(len(unique_targets) - 1, 0))\n if corrected_count:\n priorities = torch.rand(\n (batch_size, len(unique_targets)), generator=generator\n )\n priorities.masked_fill_(\n unique_targets[None, :].eq(targets.detach().cpu()[:, None]),\n -1.0,\n )\n columns = torch.topk(priorities, k=corrected_count, dim=1).indices\n inbatch = unique_targets[columns]\n if corrected_count < inbatch_count:\n filler = torch.randint(\n n_items,\n (batch_size, inbatch_count - corrected_count),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.cat([inbatch, filler], dim=1)\n\n negatives_cpu = torch.cat([uniform, inbatch], dim=1)\n correction_cpu = torch.ones_like(negatives_cpu, dtype=torch.float32)\n if corrected_count:\n correction_cpu[\n :, uniform_count : uniform_count + corrected_count\n ] = item_correction[inbatch[:, :corrected_count]]\n negatives = negatives_cpu.to(device)\n correction = correction_cpu.to(device)\n for _ in range(32):\n invalid = invalid_negative_mask(\n negatives, targets, context_items, liked_items\n )\n if not bool(invalid.any()):\n return negatives, correction\n replacements = torch.randint(\n n_items,\n (int(invalid.sum().detach().cpu()),),\n generator=generator,\n device=\"cpu\",\n ).to(device)\n negatives[invalid] = replacements\n correction[invalid] = 1.0\n invalid = invalid_negative_mask(\n negatives, targets, context_items, liked_items\n )\n if bool(invalid.any()):\n raise RuntimeError(\"mixed negative resampling failed false-negative filter\")\n return negatives, correction\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:44:11.500Z",
"resultDisplay": "Read lines 350-434 of 434 from ../task/current_model/two_tower_model.py",
"description": "../.../two_tower_model.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:44:11.501Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 350-434 of 434 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 435.\n\n--- FILE CONTENT (truncated) ---\n liked_items: torch.Tensor,\n) -> torch.Tensor:\n \"\"\"Identify targets, known positives, active likes, and row duplicates.\"\"\"\n invalid = negatives.eq(targets[:, None])\n invalid |= negatives[:, :, None].eq(context_items[:, None, :]).any(dim=2)\n invalid |= negatives[:, :, None].eq(liked_items[:, None, :]).any(dim=2)\n for column in range(1, negatives.shape[1]):\n invalid[:, column] |= negatives[:, column, None].eq(\n negatives[:, :column]\n ).any(dim=1)\n return invalid\n\n\ndef draw_filtered_mixed_negatives(\n targets: torch.Tensor,\n context_items: torch.Tensor,\n liked_items: torch.Tensor,\n n_items: int,\n uniform_count: int,\n inbatch_count: int,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n \"\"\"Draw uniform and unique-target in-batch negatives.\"\"\"\n batch_size = len(targets)\n uniform = torch.randint(\n n_items,\n (batch_size, uniform_count),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.empty((batch_size, 0), dtype=torch.long, device=\"cpu\")\n corrected_count = 0\n if inbatch_count:\n unique_targets = torch.unique(targets.detach().cpu())\n corrected_count = min(inbatch_count, max(len(unique_targets) - 1, 0))\n if corrected_count:\n priorities = torch.rand(\n (batch_size, len(unique_targets)), generator=generator\n )\n priorities.masked_fill_(\n unique_targets[None, :].eq(targets.detach().cpu()[:, None]),\n -1.0,\n )\n columns = torch.topk(priorities, k=corrected_count, dim=1).indices\n inbatch = unique_targets[columns]\n if corrected_count < inbatch_count:\n filler = torch.randint(\n n_items,\n (batch_size, inbatch_count - corrected_count),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.cat([inbatch, filler], dim=1)\n\n negatives_cpu = torch.cat([uniform, inbatch], dim=1)\n correction_cpu = torch.ones_like(negatives_cpu, dtype=torch.float32)\n if corrected_count:\n correction_cpu[\n :, uniform_count : uniform_count + corrected_count\n ] = item_correction[inbatch[:, :corrected_count]]\n negatives = negatives_cpu.to(device)\n correction = correction_cpu.to(device)\n for _ in range(32):\n invalid = invalid_negative_mask(\n negatives, targets, context_items, liked_items\n )\n if not bool(invalid.any()):\n return negatives, correction\n replacements = torch.randint(\n n_items,\n (int(invalid.sum().detach().cpu()),),\n generator=generator,\n device=\"cpu\",\n ).to(device)\n negatives[invalid] = replacements\n correction[invalid] = 1.0\n invalid = invalid_negative_mask(\n negatives, targets, context_items, liked_items\n )\n if bool(invalid.any()):\n raise RuntimeError(\"mixed negative resampling failed false-negative filter\")\n return negatives, correction\n",
"isError": false
},
"responseMessageId": "95030658-b2c0-4fac-b2a3-7d3a2d9c4248"
}
]
},
{
"id": 21,
"role": "agent",
"timestamp": "2026-09-29T05:44:13.667Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:44:13.667Z",
"callId": "read_file__call_42601",
"name": "read_file",
"input": {
"file_path": "/task/evaluate_public.py",
"start_line": 120,
"end_line": 158
},
"recorded": {
"id": "read_file__call_42601",
"name": "read_file",
"args": {
"file_path": "/task/evaluate_public.py",
"start_line": 120,
"end_line": 158
},
"result": [
{
"functionResponse": {
"id": "read_file__call_42601",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 120-158 of 158 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 159.\n\n--- FILE CONTENT (truncated) ---\n )\n return ranked_items(frame)\n\n\ndef score(\n ranked: dict[int, list[int]],\n graded: dict[int, dict[int, int]],\n) -> dict[str, float | int]:\n \"\"\"Aggregate per-user NDCG@10 over users with public relevance labels.\n\n Args:\n ranked: Ranked item IDs for every target user.\n graded: Eligible validation relevance keyed by user and item.\n\n Returns:\n The six-decimal ``quality_ndcg@10`` value and number of scored users.\n \"\"\"\n return score_rankings(ranked, graded)\n\n\ndef main() -> None:\n \"\"\"Run the public evaluator CLI and print its JSON score report.\"\"\"\n targets_frame = pd.read_parquet(DATA / \"target_users.parquet\")\n targets = set(targets_frame.uid.astype(int))\n eligibility = CandidateSet(DATA / \"validation_eligible_candidates.npz\")\n graded = validation_truth(eligibility, targets)\n\n if len(sys.argv) != 2:\n raise SystemExit(\n \"usage: evaluate_public.py \"\n )\n path = Path(sys.argv[1])\n ranked = load_submission(path, eligibility, targets)\n print(json.dumps({\"model\": str(path), \"scores\": score(ranked, graded)}, indent=2))\n\n\nif __name__ == \"__main__\":\n main()\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:44:13.672Z",
"resultDisplay": "Read lines 120-158 of 158 from ../task/evaluate_public.py",
"description": "../task/evaluate_public.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:44:13.673Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 120-158 of 158 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 159.\n\n--- FILE CONTENT (truncated) ---\n )\n return ranked_items(frame)\n\n\ndef score(\n ranked: dict[int, list[int]],\n graded: dict[int, dict[int, int]],\n) -> dict[str, float | int]:\n \"\"\"Aggregate per-user NDCG@10 over users with public relevance labels.\n\n Args:\n ranked: Ranked item IDs for every target user.\n graded: Eligible validation relevance keyed by user and item.\n\n Returns:\n The six-decimal ``quality_ndcg@10`` value and number of scored users.\n \"\"\"\n return score_rankings(ranked, graded)\n\n\ndef main() -> None:\n \"\"\"Run the public evaluator CLI and print its JSON score report.\"\"\"\n targets_frame = pd.read_parquet(DATA / \"target_users.parquet\")\n targets = set(targets_frame.uid.astype(int))\n eligibility = CandidateSet(DATA / \"validation_eligible_candidates.npz\")\n graded = validation_truth(eligibility, targets)\n\n if len(sys.argv) != 2:\n raise SystemExit(\n \"usage: evaluate_public.py \"\n )\n path = Path(sys.argv[1])\n ranked = load_submission(path, eligibility, targets)\n print(json.dumps({\"model\": str(path), \"scores\": score(ranked, graded)}, indent=2))\n\n\nif __name__ == \"__main__\":\n main()\n",
"isError": false
},
"responseMessageId": "55f3fc48-f209-483b-ae3c-a0320c3d8bfe"
}
]
},
{
"id": 22,
"role": "agent",
"timestamp": "2026-09-29T05:44:15.576Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:44:15.576Z",
"callId": "read_file__call_4833238",
"name": "read_file",
"input": {
"file_path": "/task/ranking_evaluation.py",
"start_line": 140,
"end_line": 232
},
"recorded": {
"id": "read_file__call_4833238",
"name": "read_file",
"args": {
"file_path": "/task/ranking_evaluation.py",
"start_line": 140,
"end_line": 232
},
"result": [
{
"functionResponse": {
"id": "read_file__call_4833238",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 140-232 of 232 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 233.\n\n--- FILE CONTENT (truncated) ---\n\n submitted_users = set(map(int, normalized.uid.unique()))\n if submitted_users != targets:\n raise RankingValidationError(\n f\"submission user set mismatch: missing={len(targets - submitted_users)} \"\n f\"extra={len(submitted_users - targets)}\"\n )\n counts = normalized.groupby(\"uid\").size()\n if not (counts == TOP_N).all():\n raise RankingValidationError(\n f\"submission must contain exactly {TOP_N} rows for every target user\"\n )\n if normalized.duplicated([\"uid\", \"item_id\"]).any():\n raise RankingValidationError(\n \"submission contains duplicate (uid, item_id) pairs\"\n )\n valid_ranks = set(range(1, TOP_N + 1))\n ranks_are_valid = normalized.groupby(\"uid\")[\"rank\"].apply(\n lambda values: set(map(int, values)) == valid_ranks\n )\n if not ranks_are_valid.all():\n raise RankingValidationError(\n f\"each user must have every rank from 1 through {TOP_N} exactly once\"\n )\n return normalized\n\n\ndef ranked_items(frame: pd.DataFrame) -> dict[int, list[int]]:\n \"\"\"Convert a validated ranking frame to ordered items keyed by user.\n\n Args:\n frame: Frame that has passed :func:`validate_ranking`.\n\n Returns:\n A ``uid -> item IDs`` mapping ordered by ascending rank.\n \"\"\"\n ordered = frame.sort_values([\"uid\", \"rank\"])\n return {\n int(uid): list(map(int, values))\n for uid, values in ordered.groupby(\"uid\").item_id\n }\n\n\ndef ndcg_at_k(ranked: list[int], gold: dict[int, int], k: int) -> float:\n \"\"\"Compute graded normalized discounted cumulative gain at ``k``.\n\n Relevance gain is ``2**grade - 1`` and the discount at one-based rank\n ``r`` is ``1 / log2(r + 1)``. Items absent from ``gold`` have grade zero.\n\n Args:\n ranked: Item IDs in predicted rank order.\n gold: Mapping from relevant item IDs to integer relevance grades.\n k: Maximum number of predicted positions to score.\n\n Returns:\n For a valid duplicate-free ranking, NDCG in ``[0, 1]``; returns\n ``0.0`` when the ideal gain is zero.\n \"\"\"\n gains = np.asarray([gold.get(item, 0) for item in ranked[:k]], dtype=np.float64)\n discounts = 1.0 / np.log2(np.arange(2, len(gains) + 2))\n dcg = float(np.sum((np.power(2.0, gains) - 1.0) * discounts))\n ideal = np.asarray(sorted(gold.values(), reverse=True)[:k], dtype=np.float64)\n ideal_discounts = 1.0 / np.log2(np.arange(2, len(ideal) + 2))\n idcg = float(np.sum((np.power(2.0, ideal) - 1.0) * ideal_discounts))\n return dcg / idcg if idcg else 0.0\n\n\ndef score_rankings(\n ranked: dict[int, list[int]],\n graded: dict[int, dict[int, int]],\n) -> dict[str, float | int]:\n \"\"\"Aggregate per-user NDCG@10 over users with relevance labels.\n\n Args:\n ranked: Ranked item IDs for every target user.\n graded: Relevant item grades keyed by user and item.\n\n Returns:\n The six-decimal ``quality_ndcg@10`` value and number of scored users.\n \"\"\"\n quality = float(\n np.mean(\n [\n ndcg_at_k(ranked[uid], gold, QUALITY_K)\n for uid, gold in graded.items()\n ]\n )\n )\n return {\n \"quality_ndcg@10\": round(quality, 6),\n \"quality_users\": len(graded),\n }\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:44:15.581Z",
"resultDisplay": "Read lines 140-232 of 232 from ../task/ranking_evaluation.py",
"description": "../task/ranking_evaluation.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:44:15.582Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 140-232 of 232 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 233.\n\n--- FILE CONTENT (truncated) ---\n\n submitted_users = set(map(int, normalized.uid.unique()))\n if submitted_users != targets:\n raise RankingValidationError(\n f\"submission user set mismatch: missing={len(targets - submitted_users)} \"\n f\"extra={len(submitted_users - targets)}\"\n )\n counts = normalized.groupby(\"uid\").size()\n if not (counts == TOP_N).all():\n raise RankingValidationError(\n f\"submission must contain exactly {TOP_N} rows for every target user\"\n )\n if normalized.duplicated([\"uid\", \"item_id\"]).any():\n raise RankingValidationError(\n \"submission contains duplicate (uid, item_id) pairs\"\n )\n valid_ranks = set(range(1, TOP_N + 1))\n ranks_are_valid = normalized.groupby(\"uid\")[\"rank\"].apply(\n lambda values: set(map(int, values)) == valid_ranks\n )\n if not ranks_are_valid.all():\n raise RankingValidationError(\n f\"each user must have every rank from 1 through {TOP_N} exactly once\"\n )\n return normalized\n\n\ndef ranked_items(frame: pd.DataFrame) -> dict[int, list[int]]:\n \"\"\"Convert a validated ranking frame to ordered items keyed by user.\n\n Args:\n frame: Frame that has passed :func:`validate_ranking`.\n\n Returns:\n A ``uid -> item IDs`` mapping ordered by ascending rank.\n \"\"\"\n ordered = frame.sort_values([\"uid\", \"rank\"])\n return {\n int(uid): list(map(int, values))\n for uid, values in ordered.groupby(\"uid\").item_id\n }\n\n\ndef ndcg_at_k(ranked: list[int], gold: dict[int, int], k: int) -> float:\n \"\"\"Compute graded normalized discounted cumulative gain at ``k``.\n\n Relevance gain is ``2**grade - 1`` and the discount at one-based rank\n ``r`` is ``1 / log2(r + 1)``. Items absent from ``gold`` have grade zero.\n\n Args:\n ranked: Item IDs in predicted rank order.\n gold: Mapping from relevant item IDs to integer relevance grades.\n k: Maximum number of predicted positions to score.\n\n Returns:\n For a valid duplicate-free ranking, NDCG in ``[0, 1]``; returns\n ``0.0`` when the ideal gain is zero.\n \"\"\"\n gains = np.asarray([gold.get(item, 0) for item in ranked[:k]], dtype=np.float64)\n discounts = 1.0 / np.log2(np.arange(2, len(gains) + 2))\n dcg = float(np.sum((np.power(2.0, gains) - 1.0) * discounts))\n ideal = np.asarray(sorted(gold.values(), reverse=True)[:k], dtype=np.float64)\n ideal_discounts = 1.0 / np.log2(np.arange(2, len(ideal) + 2))\n idcg = float(np.sum((np.power(2.0, ideal) - 1.0) * ideal_discounts))\n return dcg / idcg if idcg else 0.0\n\n\ndef score_rankings(\n ranked: dict[int, list[int]],\n graded: dict[int, dict[int, int]],\n) -> dict[str, float | int]:\n \"\"\"Aggregate per-user NDCG@10 over users with relevance labels.\n\n Args:\n ranked: Ranked item IDs for every target user.\n graded: Relevant item grades keyed by user and item.\n\n Returns:\n The six-decimal ``quality_ndcg@10`` value and number of scored users.\n \"\"\"\n quality = float(\n np.mean(\n [\n ndcg_at_k(ranked[uid], gold, QUALITY_K)\n for uid, gold in graded.items()\n ]\n )\n )\n return {\n \"quality_ndcg@10\": round(quality, 6),\n \"quality_users\": len(graded),\n }\n",
"isError": false
},
"responseMessageId": "f5acdec3-9f70-41bd-8a63-de47d3792525"
}
]
},
{
"id": 23,
"role": "agent",
"timestamp": "2026-09-29T05:44:22.391Z",
"blocks": [
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:44:22.391Z",
"text": "Investigating Production Model\n\nI'm currently delving into the production model, scrutinizing its retrieval performance with the provided implementation and public validation data. My aim is to identify areas for enhancement to ultimately deliver a superior ranking artifact."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:44:22.391Z",
"text": "Refining Negative Sampling\n\nI'm focusing on optimizing the negative sampling strategy within the training objective. My current work involves diagnosing limitations in the existing approach and devising improvements that can be validated against the public backtest, all while adhering to the strict latency budget and CPU constraints of the production environment."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:44:22.391Z",
"text": "Validating Development Workflow\n\nI'm clarifying the distinction between model development and final refitting stages, specifically how to utilize public validation windows. The goal is to ensure improvements are rigorously tested and that the final submission adheres precisely to the specified format, including all required columns and rankings for target users."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:44:22.391Z",
"text": "Testing Baseline Workflow\n\nI'm now prioritizing a baseline run to establish a performance benchmark. My immediate actions involve copying existing code, investigating the baseline submission stage duration, and exploring the feasibility of precomputing cached data. This will inform our subsequent optimization efforts."
},
{
"kind": "tool",
"timestamp": "2026-09-29T05:44:22.391Z",
"callId": "run_shell_command__call_146928",
"name": "run_shell_command",
"input": {
"command": "python /task/current_model/train_current_model.py --stage validation --public-dir /task --output /tmp/dummy.parquet --prepared-cache /app/solution/prepared_val.npz --prepare-only",
"description": "Prepare and cache validation sequence data"
},
"recorded": {
"id": "run_shell_command__call_146928",
"name": "run_shell_command",
"args": {
"command": "python /task/current_model/train_current_model.py --stage validation --public-dir /task --output /tmp/dummy.parquet --prepared-cache /app/solution/prepared_val.npz --prepare-only",
"description": "Prepare and cache validation sequence data"
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_146928",
"name": "run_shell_command",
"response": {
"output": "\nOutput: loaded explicit events: rows=720,270 users=8,220\nprepared users=1,000 tokens=2,090,746 states=35,727 elapsed=4.8s\nprepared users=2,000 tokens=4,301,266 states=71,242 elapsed=9.8s\nprepared users=3,000 tokens=6,486,146 states=109,471 elapsed=14.9s\nprepared users=4,000 tokens=8,793,043 states=148,210 elapsed=20.3s\nprepared users=5,000 tokens=10,992,205 states=182,596 elapsed=25.2s\nprepared users=6,000 tokens=13,061,727 states=217,148 elapsed=29.9s\nprepared users=7,000 tokens=15,334,418 states=256,914 elapsed=35.5s\nprepared users=8,000 tokens=17,462,672 states=296,703 elapsed=40.8s\nprepared users=9,000 tokens=19,568,405 states=331,551 elapsed=45.5s\ntarget users without qualifying catalog sequence=7\nprepared stage=validation sequence_users=9,135 tokens=20,056,329 next_item_targets=20,047,194 states=339,133\nsaved public-derived full-sequence cache: /app/solution/prepared_val.npz\n{\n \"stage\": \"validation\",\n \"sequence_users\": 9135,\n \"sequence_tokens\": 20056329,\n \"next_item_targets\": 20047194,\n \"selected_training_targets\": 1052953,\n \"selected_validation_targets\": 130103,\n \"discarded_old_targets\": 18864138,\n \"target_users\": 8456,\n \"runtime_seconds\": 50.735,\n \"config\": {\n \"bptt_steps\": 100,\n \"user_batch_size\": 64,\n \"embedding_dim\": 192,\n \"hidden_dim\": 192,\n \"user_hidden_1\": 512,\n \"user_hidden_2\": 256,\n \"item_hidden_1\": 512,\n \"item_hidden_2\": 256,\n \"dropout\": 0.05,\n \"epochs\": 1,\n \"learning_rate\": 0.001,\n \"weight_decay\": 1e-05,\n \"gradient_clip\": 2.0,\n \"loss_microbatch_size\": 65536,\n \"uniform_negatives\": 24,\n \"inbatch_negatives\": 8,\n \"inbatch_correction_power\": 0.5,\n \"inbatch_correction_min\": 0.25,\n \"inbatch_correction_max\": 4.0,\n \"negative_filter_scope\": \"full-training-history\",\n \"target_weighting\": \"organic-strong\",\n \"ordinary_target_weight\": 0.25,\n \"organic_strong_target_weight\": 1.0,\n \"organic_like_target_weight\": 3.0,\n \"discovery_horizon_days\": 7,\n \"validation_fraction\": 0.1,\n \"max_train_targets_per_user\": 128,\n \"max_validation_targets_per_user\": 16,\n \"early_stopping_patience\": 1,\n \"early_stopping_min_delta\": 0.0001,\n \"seed\": 29\n }\n}\nProcess Group PGID: 1723\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:45:17.158Z",
"resultDisplay": "loaded explicit events: rows=720,270 users=8,220\nprepared users=1,000 tokens=2,090,746 states=35,727 elapsed=4.8s\nprepared users=2,000 tokens=4,301,266 states=71,242 elapsed=9.8s\nprepared users=3,000 tokens=6,486,146 states=109,471 elapsed=14.9s\nprepared users=4,000 tokens=8,793,043 states=148,210 elapsed=20.3s\nprepared users=5,000 tokens=10,992,205 states=182,596 elapsed=25.2s\nprepared users=6,000 tokens=13,061,727 states=217,148 elapsed=29.9s\nprepared users=7,000 tokens=15,334,418 states=256,914 elapsed=35.5s\nprepared users=8,000 tokens=17,462,672 states=296,703 elapsed=40.8s\nprepared users=9,000 tokens=19,568,405 states=331,551 elapsed=45.5s\ntarget users without qualifying catalog sequence=7\nprepared stage=validation sequence_users=9,135 tokens=20,056,329 next_item_targets=20,047,194 states=339,133\nsaved public-derived full-sequence cache: /app/solution/prepared_val.npz\n{\n \"stage\": \"validation\",\n \"sequence_users\": 9135,\n \"sequence_tokens\": 20056329,\n \"next_item_targets\": 20047194,\n \"selected_training_targets\": 1052953,\n \"selected_validation_targets\": 130103,\n \"discarded_old_targets\": 18864138,\n \"target_users\": 8456,\n \"runtime_seconds\": 50.735,\n \"config\": {\n \"bptt_steps\": 100,\n \"user_batch_size\": 64,\n \"embedding_dim\": 192,\n \"hidden_dim\": 192,\n \"user_hidden_1\": 512,\n \"user_hidden_2\": 256,\n \"item_hidden_1\": 512,\n \"item_hidden_2\": 256,\n \"dropout\": 0.05,\n \"epochs\": 1,\n \"learning_rate\": 0.001,\n \"weight_decay\": 1e-05,\n \"gradient_clip\": 2.0,\n \"loss_microbatch_size\": 65536,\n \"uniform_negatives\": 24,\n \"inbatch_negatives\": 8,\n \"inbatch_correction_power\": 0.5,\n \"inbatch_correction_min\": 0.25,\n \"inbatch_correction_max\": 4.0,\n \"negative_filter_scope\": \"full-training-history\",\n \"target_weighting\": \"organic-strong\",\n \"ordinary_target_weight\": 0.25,\n \"organic_strong_target_weight\": 1.0,\n \"organic_like_target_weight\": 3.0,\n \"discovery_horizon_days\": 7,\n \"validation_fraction\": 0.1,\n \"max_train_targets_per_user\": 128,\n \"max_validation_targets_per_user\": 16,\n \"early_stopping_patience\": 1,\n \"early_stopping_min_delta\": 0.0001,\n \"seed\": 29\n }\n}",
"description": "Prepare and cache validation sequence data",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T05:45:17.160Z",
"content": "\nOutput: loaded explicit events: rows=720,270 users=8,220\nprepared users=1,000 tokens=2,090,746 states=35,727 elapsed=4.8s\nprepared users=2,000 tokens=4,301,266 states=71,242 elapsed=9.8s\nprepared users=3,000 tokens=6,486,146 states=109,471 elapsed=14.9s\nprepared users=4,000 tokens=8,793,043 states=148,210 elapsed=20.3s\nprepared users=5,000 tokens=10,992,205 states=182,596 elapsed=25.2s\nprepared users=6,000 tokens=13,061,727 states=217,148 elapsed=29.9s\nprepared users=7,000 tokens=15,334,418 states=256,914 elapsed=35.5s\nprepared users=8,000 tokens=17,462,672 states=296,703 elapsed=40.8s\nprepared users=9,000 tokens=19,568,405 states=331,551 elapsed=45.5s\ntarget users without qualifying catalog sequence=7\nprepared stage=validation sequence_users=9,135 tokens=20,056,329 next_item_targets=20,047,194 states=339,133\nsaved public-derived full-sequence cache: /app/solution/prepared_val.npz\n{\n \"stage\": \"validation\",\n \"sequence_users\": 9135,\n \"sequence_tokens\": 20056329,\n \"next_item_targets\": 20047194,\n \"selected_training_targets\": 1052953,\n \"selected_validation_targets\": 130103,\n \"discarded_old_targets\": 18864138,\n \"target_users\": 8456,\n \"runtime_seconds\": 50.735,\n \"config\": {\n \"bptt_steps\": 100,\n \"user_batch_size\": 64,\n \"embedding_dim\": 192,\n \"hidden_dim\": 192,\n \"user_hidden_1\": 512,\n \"user_hidden_2\": 256,\n \"item_hidden_1\": 512,\n \"item_hidden_2\": 256,\n \"dropout\": 0.05,\n \"epochs\": 1,\n \"learning_rate\": 0.001,\n \"weight_decay\": 1e-05,\n \"gradient_clip\": 2.0,\n \"loss_microbatch_size\": 65536,\n \"uniform_negatives\": 24,\n \"inbatch_negatives\": 8,\n \"inbatch_correction_power\": 0.5,\n \"inbatch_correction_min\": 0.25,\n \"inbatch_correction_max\": 4.0,\n \"negative_filter_scope\": \"full-training-history\",\n \"target_weighting\": \"organic-strong\",\n \"ordinary_target_weight\": 0.25,\n \"organic_strong_target_weight\": 1.0,\n \"organic_like_target_weight\": 3.0,\n \"discovery_horizon_days\": 7,\n \"validation_fraction\": 0.1,\n \"max_train_targets_per_user\": 128,\n \"max_validation_targets_per_user\": 16,\n \"early_stopping_patience\": 1,\n \"early_stopping_min_delta\": 0.0001,\n \"seed\": 29\n }\n}\nProcess Group PGID: 1723\n",
"isError": false
},
"responseMessageId": "c5e8e868-e797-4d35-ab00-84455967e15b"
}
]
},
{
"id": 24,
"role": "agent",
"timestamp": "2026-09-29T05:45:49.209Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:45:49.209Z",
"callId": "run_shell_command__call_27773",
"name": "run_shell_command",
"input": {
"description": "Check size of prepared_val.npz",
"command": "ls -lh /app/solution/prepared_val.npz"
},
"recorded": {
"id": "run_shell_command__call_27773",
"name": "run_shell_command",
"args": {
"description": "Check size of prepared_val.npz",
"command": "ls -lh /app/solution/prepared_val.npz"
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_27773",
"name": "run_shell_command",
"response": {
"output": "\nOutput: -rw-r--r-- 1 agent agent 511M Sep 29 05:45 /app/solution/prepared_val.npz\nProcess Group PGID: 1753\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:45:49.264Z",
"resultDisplay": "-rw-r--r-- 1 agent agent 511M Sep 29 05:45 /app/solution/prepared_val.npz",
"description": "ls -lh /app/solution/prepared_val.npz",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T05:45:49.265Z",
"content": "\nOutput: -rw-r--r-- 1 agent agent 511M Sep 29 05:45 /app/solution/prepared_val.npz\nProcess Group PGID: 1753\n",
"isError": false
},
"responseMessageId": "7c1522bc-ec48-4235-8293-8cb61d3a0de9"
}
]
},
{
"id": 25,
"role": "agent",
"timestamp": "2026-09-29T05:46:04.342Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:46:04.342Z",
"callId": "run_shell_command__call_23943",
"name": "run_shell_command",
"input": {
"command": "cp /task/current_model/train_current_model.py /app/solution/ && cp /task/current_model/two_tower_model.py /app/solution/",
"description": "Copy current_model files to /app/solution/"
},
"recorded": {
"id": "run_shell_command__call_23943",
"name": "run_shell_command",
"args": {
"command": "cp /task/current_model/train_current_model.py /app/solution/ && cp /task/current_model/two_tower_model.py /app/solution/",
"description": "Copy current_model files to /app/solution/"
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_23943",
"name": "run_shell_command",
"response": {
"output": "\nOutput: (empty)\nProcess Group PGID: 1756\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:46:04.364Z",
"resultDisplay": "",
"description": "cp /task/current_model/train_current_model.py /app/solution/ && cp /task/current_model/two_tower_model.py /app/solution/",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T05:46:04.365Z",
"content": "\nOutput: (empty)\nProcess Group PGID: 1756\n",
"isError": false
},
"responseMessageId": "73e68a09-4bd8-4c55-9fac-5f118f8e1f0a"
}
]
},
{
"id": 26,
"role": "agent",
"timestamp": "2026-09-29T05:46:10.404Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:46:10.404Z",
"callId": "run_shell_command__call_6056548",
"name": "run_shell_command",
"input": {
"description": "Prepare and cache submission sequence data",
"command": "python /app/solution/train_current_model.py --stage submission --public-dir /task --output /tmp/dummy_sub.parquet --prepared-cache /app/solution/prepared_sub.npz --prepare-only"
},
"recorded": {
"id": "run_shell_command__call_6056548",
"name": "run_shell_command",
"args": {
"description": "Prepare and cache submission sequence data",
"command": "python /app/solution/train_current_model.py --stage submission --public-dir /task --output /tmp/dummy_sub.parquet --prepared-cache /app/solution/prepared_sub.npz --prepare-only"
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_6056548",
"name": "run_shell_command",
"response": {
"output": "\nOutput: loaded explicit events: rows=746,504 users=8,307\nprepared users=1,000 tokens=2,167,338 states=37,073 elapsed=5.1s\nprepared users=2,000 tokens=4,450,246 states=73,784 elapsed=10.2s\nprepared users=3,000 tokens=6,703,970 states=113,190 elapsed=15.6s\nprepared users=4,000 tokens=9,087,187 states=153,283 elapsed=21.2s\nprepared users=5,000 tokens=11,354,844 states=188,776 elapsed=26.2s\nprepared users=6,000 tokens=13,485,786 states=224,384 elapsed=31.1s\nprepared users=7,000 tokens=15,826,114 states=265,318 elapsed=36.8s\nprepared users=8,000 tokens=18,028,622 states=306,474 elapsed=42.3s\nprepared users=9,000 tokens=20,195,749 states=342,272 elapsed=47.2s\ntarget users without qualifying catalog sequence=6\nprepared stage=submission sequence_users=9,153 tokens=20,712,971 next_item_targets=20,703,818 states=350,322\nsaved public-derived full-sequence cache: /app/solution/prepared_sub.npz\n{\n \"stage\": \"submission\",\n \"sequence_users\": 9153,\n \"sequence_tokens\": 20712971,\n \"next_item_targets\": 20703818,\n \"selected_training_targets\": 1067606,\n \"selected_validation_targets\": 131997,\n \"discarded_old_targets\": 19504215,\n \"target_users\": 8456,\n \"runtime_seconds\": 52.507,\n \"config\": {\n \"bptt_steps\": 100,\n \"user_batch_size\": 64,\n \"embedding_dim\": 192,\n \"hidden_dim\": 192,\n \"user_hidden_1\": 512,\n \"user_hidden_2\": 256,\n \"item_hidden_1\": 512,\n \"item_hidden_2\": 256,\n \"dropout\": 0.05,\n \"epochs\": 1,\n \"learning_rate\": 0.001,\n \"weight_decay\": 1e-05,\n \"gradient_clip\": 2.0,\n \"loss_microbatch_size\": 65536,\n \"uniform_negatives\": 24,\n \"inbatch_negatives\": 8,\n \"inbatch_correction_power\": 0.5,\n \"inbatch_correction_min\": 0.25,\n \"inbatch_correction_max\": 4.0,\n \"negative_filter_scope\": \"full-training-history\",\n \"target_weighting\": \"organic-strong\",\n \"ordinary_target_weight\": 0.25,\n \"organic_strong_target_weight\": 1.0,\n \"organic_like_target_weight\": 3.0,\n \"discovery_horizon_days\": 7,\n \"validation_fraction\": 0.1,\n \"max_train_targets_per_user\": 128,\n \"max_validation_targets_per_user\": 16,\n \"early_stopping_patience\": 1,\n \"early_stopping_min_delta\": 0.0001,\n \"seed\": 29\n }\n}\nProcess Group PGID: 1760\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:47:06.908Z",
"resultDisplay": "loaded explicit events: rows=746,504 users=8,307\nprepared users=1,000 tokens=2,167,338 states=37,073 elapsed=5.1s\nprepared users=2,000 tokens=4,450,246 states=73,784 elapsed=10.2s\nprepared users=3,000 tokens=6,703,970 states=113,190 elapsed=15.6s\nprepared users=4,000 tokens=9,087,187 states=153,283 elapsed=21.2s\nprepared users=5,000 tokens=11,354,844 states=188,776 elapsed=26.2s\nprepared users=6,000 tokens=13,485,786 states=224,384 elapsed=31.1s\nprepared users=7,000 tokens=15,826,114 states=265,318 elapsed=36.8s\nprepared users=8,000 tokens=18,028,622 states=306,474 elapsed=42.3s\nprepared users=9,000 tokens=20,195,749 states=342,272 elapsed=47.2s\ntarget users without qualifying catalog sequence=6\nprepared stage=submission sequence_users=9,153 tokens=20,712,971 next_item_targets=20,703,818 states=350,322\nsaved public-derived full-sequence cache: /app/solution/prepared_sub.npz\n{\n \"stage\": \"submission\",\n \"sequence_users\": 9153,\n \"sequence_tokens\": 20712971,\n \"next_item_targets\": 20703818,\n \"selected_training_targets\": 1067606,\n \"selected_validation_targets\": 131997,\n \"discarded_old_targets\": 19504215,\n \"target_users\": 8456,\n \"runtime_seconds\": 52.507,\n \"config\": {\n \"bptt_steps\": 100,\n \"user_batch_size\": 64,\n \"embedding_dim\": 192,\n \"hidden_dim\": 192,\n \"user_hidden_1\": 512,\n \"user_hidden_2\": 256,\n \"item_hidden_1\": 512,\n \"item_hidden_2\": 256,\n \"dropout\": 0.05,\n \"epochs\": 1,\n \"learning_rate\": 0.001,\n \"weight_decay\": 1e-05,\n \"gradient_clip\": 2.0,\n \"loss_microbatch_size\": 65536,\n \"uniform_negatives\": 24,\n \"inbatch_negatives\": 8,\n \"inbatch_correction_power\": 0.5,\n \"inbatch_correction_min\": 0.25,\n \"inbatch_correction_max\": 4.0,\n \"negative_filter_scope\": \"full-training-history\",\n \"target_weighting\": \"organic-strong\",\n \"ordinary_target_weight\": 0.25,\n \"organic_strong_target_weight\": 1.0,\n \"organic_like_target_weight\": 3.0,\n \"discovery_horizon_days\": 7,\n \"validation_fraction\": 0.1,\n \"max_train_targets_per_user\": 128,\n \"max_validation_targets_per_user\": 16,\n \"early_stopping_patience\": 1,\n \"early_stopping_min_delta\": 0.0001,\n \"seed\": 29\n }\n}",
"description": "Prepare and cache submission sequence data",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T05:47:06.910Z",
"content": "\nOutput: loaded explicit events: rows=746,504 users=8,307\nprepared users=1,000 tokens=2,167,338 states=37,073 elapsed=5.1s\nprepared users=2,000 tokens=4,450,246 states=73,784 elapsed=10.2s\nprepared users=3,000 tokens=6,703,970 states=113,190 elapsed=15.6s\nprepared users=4,000 tokens=9,087,187 states=153,283 elapsed=21.2s\nprepared users=5,000 tokens=11,354,844 states=188,776 elapsed=26.2s\nprepared users=6,000 tokens=13,485,786 states=224,384 elapsed=31.1s\nprepared users=7,000 tokens=15,826,114 states=265,318 elapsed=36.8s\nprepared users=8,000 tokens=18,028,622 states=306,474 elapsed=42.3s\nprepared users=9,000 tokens=20,195,749 states=342,272 elapsed=47.2s\ntarget users without qualifying catalog sequence=6\nprepared stage=submission sequence_users=9,153 tokens=20,712,971 next_item_targets=20,703,818 states=350,322\nsaved public-derived full-sequence cache: /app/solution/prepared_sub.npz\n{\n \"stage\": \"submission\",\n \"sequence_users\": 9153,\n \"sequence_tokens\": 20712971,\n \"next_item_targets\": 20703818,\n \"selected_training_targets\": 1067606,\n \"selected_validation_targets\": 131997,\n \"discarded_old_targets\": 19504215,\n \"target_users\": 8456,\n \"runtime_seconds\": 52.507,\n \"config\": {\n \"bptt_steps\": 100,\n \"user_batch_size\": 64,\n \"embedding_dim\": 192,\n \"hidden_dim\": 192,\n \"user_hidden_1\": 512,\n \"user_hidden_2\": 256,\n \"item_hidden_1\": 512,\n \"item_hidden_2\": 256,\n \"dropout\": 0.05,\n \"epochs\": 1,\n \"learning_rate\": 0.001,\n \"weight_decay\": 1e-05,\n \"gradient_clip\": 2.0,\n \"loss_microbatch_size\": 65536,\n \"uniform_negatives\": 24,\n \"inbatch_negatives\": 8,\n \"inbatch_correction_power\": 0.5,\n \"inbatch_correction_min\": 0.25,\n \"inbatch_correction_max\": 4.0,\n \"negative_filter_scope\": \"full-training-history\",\n \"target_weighting\": \"organic-strong\",\n \"ordinary_target_weight\": 0.25,\n \"organic_strong_target_weight\": 1.0,\n \"organic_like_target_weight\": 3.0,\n \"discovery_horizon_days\": 7,\n \"validation_fraction\": 0.1,\n \"max_train_targets_per_user\": 128,\n \"max_validation_targets_per_user\": 16,\n \"early_stopping_patience\": 1,\n \"early_stopping_min_delta\": 0.0001,\n \"seed\": 29\n }\n}\nProcess Group PGID: 1760\n",
"isError": false
},
"responseMessageId": "33ff9433-824e-4424-ba97-f5e80e5d85e2"
}
]
},
{
"id": 27,
"role": "agent",
"timestamp": "2026-09-29T05:47:29.748Z",
"blocks": [
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:47:29.748Z",
"text": "Caching Data Now\n\nI've successfully cached both `prepared_val.npz` and `prepared_sub.npz`. My immediate next step is to generate an initial baseline `submission.parquet` to ensure I have a valid submission saved before the deadline. Following that, I will proceed with training the current model."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:47:29.748Z",
"text": "Generating Submission File\n\nI'm considering generating a baseline `submission.parquet` now, even though the score isn't optimal. This ensures I have a valid submission saved, providing a fallback should anything go wrong before the deadline. I'm weighing this against continuing to investigate potential issues with the negative validation."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:47:29.748Z",
"text": "Investigating Negatives Systematically\n\nI'm shifting my focus from immediate submission generation to a deep dive into the negative sampling strategy. Running parallel training jobs on CPU seems counterproductive given resource constraints, so I'm prioritizing careful code inspection and diagnosis. My goal is to understand the current limitations and identify improvements to the negative sampling process itself before proceeding with further training."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:47:29.748Z",
"text": "Diagnosing Negative Sampling\n\nI'm now meticulously examining how my training objective selects and utilizes negative examples. My primary aim is to identify limitations within the current negative-sampling strategy and enhance it while adhering to the fixed CPU budget. I will then validate these changes using the public backtest."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:47:29.748Z",
"text": "Refining Negative Selection Logic\n\nI'm currently scrutinizing the `invalid` mask generation within the negative sampling code. My focus is on understanding its interaction with `draw_full_history_filtered_negatives` and ensuring it correctly identifies and excludes unwanted samples before I proceed with any modifications."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:47:29.748Z",
"text": "Analyzing Negative Sampling Code\n\nI'm currently scrutinizing the `invalid` mask generation within the negative sampling code. My focus is on understanding its interaction with `draw_full_history_filtered_negatives` and ensuring it correctly identifies and excludes unwanted samples before I proceed with any modifications."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:47:29.748Z",
"text": "Debugging Inbatch Negatives\n\nI'm now investigating the `inbatch` negative sampling, specifically where `targets` originates in `draw_training_negatives`. I'm checking `run_training_epoch` and the `build_dense_block` function for clues, as the current logic seems to be missing a critical piece."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:47:29.748Z",
"text": "Tracing Positive Item Origin\n\nI'm now tracing the origin of `positive_items` in the `draw_training_negatives` function. It appears to be derived from `flat_scored`, which itself originates from a BPTT block, and I need to understand how `users` and `steps` in that block determine its dimensions and content before proceeding with the negative sampling logic."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:47:29.748Z",
"text": "Understanding Target Origins\n\nI'm now analyzing how `positive_items` are derived from `flat_scored` and the `users` and `steps` within the BPTT block. This is crucial for understanding the full scope of the 6,400 targets and how they are being used to sample both uniform and in-batch negatives, which then informs the `corrected_negative_loss` calculation."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:47:29.748Z",
"text": "Analyzing Loss Functions\n\nI am currently analyzing the `corrected_negative_loss` function and its relationship with `F.softplus`. I've realized it's essentially binary cross-entropy, but I need to fully understand how it treats positive samples versus averaged negative samples before proceeding. This seems to be the core of the issue."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:47:29.748Z",
"text": "Investigating Loss Calculation\n\nI'm dissecting the `corrected_negative_loss` function, realizing it's effectively BCE. My focus is on understanding how the averaging of negative logits impacts the gradient, as a scaling factor of `1/K` appears to dilute the push from individual negatives."
},
{
"kind": "tool",
"timestamp": "2026-09-29T05:47:29.748Z",
"callId": "grep_search__call_77345",
"name": "grep_search",
"input": {
"dir_path": "/task/current_model",
"pattern": "dislike"
},
"recorded": {
"id": "grep_search__call_77345",
"name": "grep_search",
"args": {
"dir_path": "/task/current_model",
"pattern": "dislike"
},
"result": [
{
"functionResponse": {
"id": "grep_search__call_77345",
"name": "grep_search",
"response": {
"output": "Found 40 matches for pattern \"dislike\" in path \"/task/current_model\":\n---\nFile: train_current_model.py\nL110: disliked_states: np.ndarray\nL114: inference_disliked_items: np.ndarray\nL176: active_dislikes: dict[int, int],\nL179: disliked_states: list[np.ndarray],\nL183: disliked_states.append(\nL184: base.pack_active_items(active_dislikes, base.DISLIKE_ITEMS, n_items)\nL186: state_counts.append((len(active_likes), len(active_dislikes)))\nL252: disliked_states: list[np.ndarray],\nL279: active_dislikes: dict[int, int] = {}\nL293: else active_dislikes\nL304: active_dislikes,\nL307: disliked_states,\nL320: dislike_count: int,\nL337: result[5] = np.log1p(dislike_count) / 4.0\nL369: inference_likes, inference_dislikes, like_counts, dislike_counts = (\nL383: disliked_states = [\nL384: np.full(base.DISLIKE_ITEMS, n_items, dtype=np.uint16)\nL405: disliked_states,\nL427: int(dislike_counts[inference_row]),\nL440: np.log1p(dislike_counts[missing]) / 4.0\nL455: disliked_states=np.stack(disliked_states),\nL459: inference_disliked_items=inference_dislikes,\nL784: disliked_items = torch.from_numpy(\nL785: prepared.disliked_states[unique_states].astype(np.int64)\nL788: disliked_vectors = model._mean_context_channel(disliked_items)\nL793: disliked_vectors[inverse_rows],\nL802: disliked_vectors: torch.Tensor,\nL807: [sequence_vectors, liked_vectors, disliked_vectors, user_features],\nL1147: liked_vectors, disliked_vectors, _ = pool_explicit_states(\nL1159: disliked_vectors,\nL1844: disliked_items = torch.from_numpy(\nL1845: prepared.inference_disliked_items[output_rows].astype(np.int64)\nL1851: model._mean_context_channel(disliked_items),\n---\nFile: two_tower_model.py\nL22: DISLIKE_ITEMS = 24\nL184: (\"dislikes\", 1, 1, 0),\nL185: (\"undislikes\", 1, 0, 1),\nL265: disliked = np.full((len(uids), DISLIKE_ITEMS), n_items, dtype=np.uint16)\nL267: dislike_counts = np.zeros(len(uids), dtype=np.int32)\nL270: (\"disliked\", DISLIKE_ITEMS, disliked, dislike_counts),\nL281: return liked, disliked, like_counts, dislike_counts\n---"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:47:29.767Z",
"resultDisplay": {
"summary": "Found 40 matches",
"matches": [
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 110,
"line": " disliked_states: np.ndarray"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 114,
"line": " inference_disliked_items: np.ndarray"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 176,
"line": " active_dislikes: dict[int, int],"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 179,
"line": " disliked_states: list[np.ndarray],"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 183,
"line": " disliked_states.append("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 184,
"line": " base.pack_active_items(active_dislikes, base.DISLIKE_ITEMS, n_items)"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 186,
"line": " state_counts.append((len(active_likes), len(active_dislikes)))"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 252,
"line": " disliked_states: list[np.ndarray],"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 279,
"line": " active_dislikes: dict[int, int] = {}"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 293,
"line": " else active_dislikes"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 304,
"line": " active_dislikes,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 307,
"line": " disliked_states,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 320,
"line": " dislike_count: int,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 337,
"line": " result[5] = np.log1p(dislike_count) / 4.0"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 369,
"line": " inference_likes, inference_dislikes, like_counts, dislike_counts = ("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 383,
"line": " disliked_states = ["
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 384,
"line": " np.full(base.DISLIKE_ITEMS, n_items, dtype=np.uint16)"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 405,
"line": " disliked_states,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 427,
"line": " int(dislike_counts[inference_row]),"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 440,
"line": " np.log1p(dislike_counts[missing]) / 4.0"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 455,
"line": " disliked_states=np.stack(disliked_states),"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 459,
"line": " inference_disliked_items=inference_dislikes,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 784,
"line": " disliked_items = torch.from_numpy("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 785,
"line": " prepared.disliked_states[unique_states].astype(np.int64)"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 788,
"line": " disliked_vectors = model._mean_context_channel(disliked_items)"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 793,
"line": " disliked_vectors[inverse_rows],"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 802,
"line": " disliked_vectors: torch.Tensor,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 807,
"line": " [sequence_vectors, liked_vectors, disliked_vectors, user_features],"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1147,
"line": " liked_vectors, disliked_vectors, _ = pool_explicit_states("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1159,
"line": " disliked_vectors,"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1844,
"line": " disliked_items = torch.from_numpy("
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1845,
"line": " prepared.inference_disliked_items[output_rows].astype(np.int64)"
},
{
"filePath": "train_current_model.py",
"absolutePath": "/task/current_model/train_current_model.py",
"lineNumber": 1851,
"line": " model._mean_context_channel(disliked_items),"
},
{
"filePath": "two_tower_model.py",
"absolutePath": "/task/current_model/two_tower_model.py",
"lineNumber": 22,
"line": "DISLIKE_ITEMS = 24"
},
{
"filePath": "two_tower_model.py",
"absolutePath": "/task/current_model/two_tower_model.py",
"lineNumber": 184,
"line": " (\"dislikes\", 1, 1, 0),"
},
{
"filePath": "two_tower_model.py",
"absolutePath": "/task/current_model/two_tower_model.py",
"lineNumber": 185,
"line": " (\"undislikes\", 1, 0, 1),"
},
{
"filePath": "two_tower_model.py",
"absolutePath": "/task/current_model/two_tower_model.py",
"lineNumber": 265,
"line": " disliked = np.full((len(uids), DISLIKE_ITEMS), n_items, dtype=np.uint16)"
},
{
"filePath": "two_tower_model.py",
"absolutePath": "/task/current_model/two_tower_model.py",
"lineNumber": 267,
"line": " dislike_counts = np.zeros(len(uids), dtype=np.int32)"
},
{
"filePath": "two_tower_model.py",
"absolutePath": "/task/current_model/two_tower_model.py",
"lineNumber": 270,
"line": " (\"disliked\", DISLIKE_ITEMS, disliked, dislike_counts),"
},
{
"filePath": "two_tower_model.py",
"absolutePath": "/task/current_model/two_tower_model.py",
"lineNumber": 281,
"line": " return liked, disliked, like_counts, dislike_counts"
}
]
},
"description": "'dislike' within ../task/current_model",
"displayName": "SearchText",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:47:29.768Z",
"content": "Found 40 matches for pattern \"dislike\" in path \"/task/current_model\":\n---\nFile: train_current_model.py\nL110: disliked_states: np.ndarray\nL114: inference_disliked_items: np.ndarray\nL176: active_dislikes: dict[int, int],\nL179: disliked_states: list[np.ndarray],\nL183: disliked_states.append(\nL184: base.pack_active_items(active_dislikes, base.DISLIKE_ITEMS, n_items)\nL186: state_counts.append((len(active_likes), len(active_dislikes)))\nL252: disliked_states: list[np.ndarray],\nL279: active_dislikes: dict[int, int] = {}\nL293: else active_dislikes\nL304: active_dislikes,\nL307: disliked_states,\nL320: dislike_count: int,\nL337: result[5] = np.log1p(dislike_count) / 4.0\nL369: inference_likes, inference_dislikes, like_counts, dislike_counts = (\nL383: disliked_states = [\nL384: np.full(base.DISLIKE_ITEMS, n_items, dtype=np.uint16)\nL405: disliked_states,\nL427: int(dislike_counts[inference_row]),\nL440: np.log1p(dislike_counts[missing]) / 4.0\nL455: disliked_states=np.stack(disliked_states),\nL459: inference_disliked_items=inference_dislikes,\nL784: disliked_items = torch.from_numpy(\nL785: prepared.disliked_states[unique_states].astype(np.int64)\nL788: disliked_vectors = model._mean_context_channel(disliked_items)\nL793: disliked_vectors[inverse_rows],\nL802: disliked_vectors: torch.Tensor,\nL807: [sequence_vectors, liked_vectors, disliked_vectors, user_features],\nL1147: liked_vectors, disliked_vectors, _ = pool_explicit_states(\nL1159: disliked_vectors,\nL1844: disliked_items = torch.from_numpy(\nL1845: prepared.inference_disliked_items[output_rows].astype(np.int64)\nL1851: model._mean_context_channel(disliked_items),\n---\nFile: two_tower_model.py\nL22: DISLIKE_ITEMS = 24\nL184: (\"dislikes\", 1, 1, 0),\nL185: (\"undislikes\", 1, 0, 1),\nL265: disliked = np.full((len(uids), DISLIKE_ITEMS), n_items, dtype=np.uint16)\nL267: dislike_counts = np.zeros(len(uids), dtype=np.int32)\nL270: (\"disliked\", DISLIKE_ITEMS, disliked, dislike_counts),\nL281: return liked, disliked, like_counts, dislike_counts\n---",
"isError": false
},
"responseMessageId": "b62b9c32-04f5-4f2d-8f58-d4babd6671e7"
}
]
},
{
"id": 28,
"role": "agent",
"timestamp": "2026-09-29T05:48:06.683Z",
"blocks": [
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:06.683Z",
"text": "Analyzing Dislike Usage\n\nI'm currently scrutinizing the precise utilization of \"dislikes\" within the user tower, specifically confirming they function solely as an input feature and are never treated as negative examples. My immediate focus is on whether dislikes are incorporated into `invalid_full_history_negatives`, and I'm reviewing the relevant code at line 609 for clarification."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:06.683Z",
"text": "Confirming Dislike Functionality\n\nI'm verifying that dislikes act solely as input and aren't misinterpreted as negative examples. My current investigation centers on their integration into `invalid_full_history_negatives`, and I'm examining line 609 for definitive confirmation of this behavior."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:06.683Z",
"text": "Investigating Input Inclusion\n\nI'm reviewing how `prepared.sequence_items` are handled, specifically questioning if user validation prefixes or full histories are excluded. My attention is drawn to `training_counts` as a potential indicator."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:06.683Z",
"text": "Examining Dislike Handling\n\nI'm now realizing that `build_known_positive_filter` already accounts for liked items and training history to prevent their sampling as negatives. My immediate question is whether explicit dislikes (`train_dislikes`) are similarly handled, given their nature as explicitly disliked tracks. I need to determine if the discovery objective considers them."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:06.683Z",
"text": "Clarifying Negative Sample Source\n\nI'm now focusing on how negative examples are selected and utilized by the training objective. My current priority is to thoroughly examine the entire codebase for any mention of the negative-sampling strategy to identify potential limitations and areas for improvement, keeping the CPU budget in mind."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:06.683Z",
"text": "Confirming Target Identity\n\nI've realized that `targets` represents all items in a BPTT block, not just those with dislikes. This clarifies their scope for in-batch negative generation."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:06.683Z",
"text": "Sampling Uniformly Incorrectly\n\nI've discovered a significant issue: in-batch negative sampling appears to be drawing uniformly from unique targets, not proportionally to their frequency. This means popular items aren't being sampled more often, which is counter to the intended \"popularity sampling\" strategy and its benefits for log-Q."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:06.683Z",
"text": "Applying Correction Strategy\n\nI've realized that my previous assumption about uniform sampling was incorrect. The `build_item_correction` function explicitly calculates in-batch propensities based on item frequency, meaning popular items *are* sampled more often. My focus now shifts to understanding how this calculated correction is applied to the loss."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:06.683Z",
"text": "Adjusting Loss Calculations\n\nI've identified a critical error where in-batch negative sampling was uniform instead of proportional. My current focus is on how the calculated item `correction` impacts the loss, particularly when it downweights popular items, and I'm reviewing `corrected_negative_loss` for its precise application."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:06.683Z",
"text": "Analyzing Loss Weighting\n\nI'm re-examining the `corrected_negative_loss` function. It appears dividing by `weights.sum(dim=1)` normalizes the weights, effectively making them sum to one. My current focus is understanding why this loss is BCE and not a standard softmax, given the prompt's description of the model architecture and its retrieval mechanism."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:06.683Z",
"text": "Revisiting Loss Interpretation\n\nI'm reconsidering the nature of the `corrected_negative_loss`. It seems I was mistaken in expecting a softmax; it's indeed a binary cross-entropy, but the `correction` term, by scaling the logit, effectively biases it. My current focus is on the implication of this BCE loss with logit scaling."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:06.683Z",
"text": "Investigating Negative Sampling\n\nI'm currently investigating how negative examples are sampled and used, specifically questioning the uniform sampling from the entire catalog for `uniform_negatives`. My immediate focus is on how in-batch negatives are drawn from the `targets` within a BPTT block and if this process is user-specific or global."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:06.683Z",
"text": "Analyzing Loss Behavior\n\nI'm questioning the loss function's interpretation. While I initially saw BCE, the context and the term `correction` suggest it's acting more like a sampled softmax, effectively biasing logits to approximate a full softmax over candidates, which is critical for understanding the loss weighting."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:06.683Z",
"text": "Confirming Loss Type\n\nI've confirmed that the loss is indeed a BCE loss, specifically `F.softplus(-positive_logits)` for positives and a `corrected_negative_loss` for negatives. My current focus is understanding how the `correction` term within `corrected_negative_loss` effectively biases the logits to mimic a softmax behavior despite being BCE."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:06.683Z",
"text": "Clarifying Negative Logit Calculation\n\nI'm analyzing how negative logits are computed. Specifically, I'm looking at `torch.e` in the provided snippet, which seems incomplete. My goal is to understand if this is part of a larger operation for calculating all negative logits and if the `sum(dim=1)` operation applies here as well."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:06.683Z",
"text": "Analyzing Negative Logit Computation\n\nI'm now realizing the `torch.e` was a placeholder for the full operation. The actual calculation involves an einsum to compute dot products between query vectors and negative vectors, generating `(6400, 32)` negative logits, which is computationally intensive."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:06.683Z",
"text": "Evaluating Target Weighting\n\nI'm now considering the `--target-weighting` parameter and its potential impact. My focus is on the \"discovery-graded\" option, aiming to understand how it influences the training process and if it offers an improvement over the current \"organic-strong\" setting."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:06.683Z",
"text": "Evaluating Target Weighting\n\nI'm now considering the `--target-weighting` parameter and its potential impact. My focus is on the \"discovery-graded\" option, aiming to understand how it influences the training process and if it offers an improvement over the current \"organic-strong\" setting. I'm examining the relevant code block to understand the specifics of \"discovery-graded\" weighting."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:06.683Z",
"text": "Investigating Negative Selection\n\nI'm now diving deep into the core of negative sampling. My current focus is on precisely how these negatives are identified from the entire catalog, considering user listening habits and the candidate pool. I'm preparing to inspect data files directly to uncover the mechanics."
},
{
"kind": "tool",
"timestamp": "2026-09-29T05:48:06.683Z",
"callId": "run_shell_command__call_2409438",
"name": "run_shell_command",
"input": {
"description": "Inspect catalog, targets, and events",
"command": "python3 -c \"\nimport pandas as pd\nimport numpy as np\n\ncatalog = pd.read_parquet('/task/data/candidate_catalog.parquet')\nprint('Catalog shape:', catalog.shape)\nprint('Catalog columns:', catalog.columns.tolist())\nprint(catalog.head(5))\n\ntargets = pd.read_parquet('/task/data/target_users.parquet')\nprint('Target users count:', len(targets))\n\nval_likes = pd.read_parquet('/task/data/val_likes.parquet')\nval_listens = pd.read_parquet('/task/data/val_listens.parquet')\ntrain_dislikes = pd.read_parquet('/task/data/train_dislikes.parquet')\nprint('Train dislikes shape:', train_dislikes.shape)\nprint('Val likes shape:', val_likes.shape)\nprint('Val listens shape:', val_listens.shape)\n\""
},
"recorded": {
"id": "run_shell_command__call_2409438",
"name": "run_shell_command",
"args": {
"description": "Inspect catalog, targets, and events",
"command": "python3 -c \"\nimport pandas as pd\nimport numpy as np\n\ncatalog = pd.read_parquet('/task/data/candidate_catalog.parquet')\nprint('Catalog shape:', catalog.shape)\nprint('Catalog columns:', catalog.columns.tolist())\nprint(catalog.head(5))\n\ntargets = pd.read_parquet('/task/data/target_users.parquet')\nprint('Target users count:', len(targets))\n\nval_likes = pd.read_parquet('/task/data/val_likes.parquet')\nval_listens = pd.read_parquet('/task/data/val_listens.parquet')\ntrain_dislikes = pd.read_parquet('/task/data/train_dislikes.parquet')\nprint('Train dislikes shape:', train_dislikes.shape)\nprint('Val likes shape:', val_likes.shape)\nprint('Val listens shape:', val_listens.shape)\n\""
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_2409438",
"name": "run_shell_command",
"response": {
"output": "\nOutput: Catalog shape: (30000, 4)\nCatalog columns: ['item_id', 'positive_listens', 'organic_positive_listens', 'popularity_rank']\n item_id positive_listens organic_positive_listens popularity_rank\n0 5862961 26615 20620 1\n1 6901374 25536 17347 2\n2 5635052 24691 17194 3\n3 3542184 24075 17699 4\n4 9378983 24036 18195 5\nTarget users count: 8456\nTrain dislikes shape: (101075, 4)\nVal likes shape: (37795, 4)\nVal listens shape: (1501697, 6)\nProcess Group PGID: 1800\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:48:08.271Z",
"resultDisplay": "Catalog shape: (30000, 4)\nCatalog columns: ['item_id', 'positive_listens', 'organic_positive_listens', 'popularity_rank']\n item_id positive_listens organic_positive_listens popularity_rank\n0 5862961 26615 20620 1\n1 6901374 25536 17347 2\n2 5635052 24691 17194 3\n3 3542184 24075 17699 4\n4 9378983 24036 18195 5\nTarget users count: 8456\nTrain dislikes shape: (101075, 4)\nVal likes shape: (37795, 4)\nVal listens shape: (1501697, 6)",
"description": "Inspect catalog, targets, and events",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T05:48:08.273Z",
"content": "\nOutput: Catalog shape: (30000, 4)\nCatalog columns: ['item_id', 'positive_listens', 'organic_positive_listens', 'popularity_rank']\n item_id positive_listens organic_positive_listens popularity_rank\n0 5862961 26615 20620 1\n1 6901374 25536 17347 2\n2 5635052 24691 17194 3\n3 3542184 24075 17699 4\n4 9378983 24036 18195 5\nTarget users count: 8456\nTrain dislikes shape: (101075, 4)\nVal likes shape: (37795, 4)\nVal listens shape: (1501697, 6)\nProcess Group PGID: 1800\n",
"isError": false
},
"responseMessageId": "dd0e9564-a3e8-4a0f-b612-8a94d74196c0"
}
]
},
{
"id": 29,
"role": "agent",
"timestamp": "2026-09-29T05:48:11.954Z",
"blocks": [
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:11.954Z",
"text": "Investigating Ranking Mechanics\n\nI'm currently examining the ranking and scoring mechanisms within `evaluate_public.py` to understand how items are ordered. My focus is on the baseline's validation scores and predictions, specifically how they appear in `/app/solution/production_validation.parquet`, and I'm also inspecting the item distributions."
},
{
"kind": "tool",
"timestamp": "2026-09-29T05:48:11.954Z",
"callId": "run_shell_command__call_5385566",
"name": "run_shell_command",
"input": {
"description": "Analyze baseline predictions in top 10",
"command": "python3 -c \"\nimport pandas as pd\nimport numpy as np\n\ndf = pd.read_parquet('/app/solution/production_validation.parquet')\nprint('Validation ranking head:')\nprint(df.head(10))\nprint('Rank range:', df['rank'].min(), df['rank'].max())\nprint('Number of unique items ranked in top 10:', df[df['rank'] <= 10]['item_id'].nunique())\nprint('Number of unique items ranked overall:', df['item_id'].nunique())\ntop_items = df[df['rank'] <= 10]['item_id'].value_counts().head(20)\nprint('Most frequent items in top 10:')\nprint(top_items)\n\ncatalog = pd.read_parquet('/task/data/candidate_catalog.parquet').set_index('item_id')\nprint('\\nPopularity of most frequent items in top 10:')\nprint(catalog.loc[top_items.index])\n\""
},
"recorded": {
"id": "run_shell_command__call_5385566",
"name": "run_shell_command",
"args": {
"description": "Analyze baseline predictions in top 10",
"command": "python3 -c \"\nimport pandas as pd\nimport numpy as np\n\ndf = pd.read_parquet('/app/solution/production_validation.parquet')\nprint('Validation ranking head:')\nprint(df.head(10))\nprint('Rank range:', df['rank'].min(), df['rank'].max())\nprint('Number of unique items ranked in top 10:', df[df['rank'] <= 10]['item_id'].nunique())\nprint('Number of unique items ranked overall:', df['item_id'].nunique())\ntop_items = df[df['rank'] <= 10]['item_id'].value_counts().head(20)\nprint('Most frequent items in top 10:')\nprint(top_items)\n\ncatalog = pd.read_parquet('/task/data/candidate_catalog.parquet').set_index('item_id')\nprint('\\nPopularity of most frequent items in top 10:')\nprint(catalog.loc[top_items.index])\n\""
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_5385566",
"name": "run_shell_command",
"response": {
"output": "\nOutput: Validation ranking head:\n uid item_id rank\n0 100 6511826 1\n1 100 3972057 2\n2 100 2736766 3\n3 100 4948220 4\n4 100 5488609 5\n5 100 7958357 6\n6 100 8460142 7\n7 100 8250570 8\n8 100 5570016 9\n9 100 7510702 10\nRank range: 1 100\nNumber of unique items ranked in top 10: 592\nNumber of unique items ranked overall: 2575\nMost frequent items in top 10:\nitem_id\n6006390 2917\n8213481 2800\n6896222 2712\n8743390 2653\n5436843 2587\n2651314 2580\n8722394 2542\n3542184 2427\n5862961 2292\n906358 2066\n5927996 1872\n9378983 1537\n2185191 1522\n6901374 1463\n3033749 1333\n5252571 1289\n2408923 1215\n4423556 1148\n8351511 1121\n3375935 1107\nName: count, dtype: int64\n\nPopularity of most frequent items in top 10:\n positive_listens organic_positive_listens popularity_rank\nitem_id \n6006390 8970 6813 75\n8213481 22505 17071 7\n6896222 3020 2017 818\n8743390 13518 8480 22\n5436843 14897 12127 14\n2651314 16286 11801 11\n8722394 3464 1695 634\n3542184 24075 17699 4\n5862961 26615 20620 1\n906358 22238 17183 8\n5927996 14751 10424 16\n9378983 24036 18195 5\n2185191 13059 8839 25\n6901374 25536 17347 2\n3033749 16682 12267 10\n5252571 14897 11331 13\n2408923 7515 4545 111\n4423556 6544 2177 157\n8351511 14061 9739 19\n3375935 5213 3232 282\nProcess Group PGID: 1820\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:48:13.533Z",
"resultDisplay": "Validation ranking head:\n uid item_id rank\n0 100 6511826 1\n1 100 3972057 2\n2 100 2736766 3\n3 100 4948220 4\n4 100 5488609 5\n5 100 7958357 6\n6 100 8460142 7\n7 100 8250570 8\n8 100 5570016 9\n9 100 7510702 10\nRank range: 1 100\nNumber of unique items ranked in top 10: 592\nNumber of unique items ranked overall: 2575\nMost frequent items in top 10:\nitem_id\n6006390 2917\n8213481 2800\n6896222 2712\n8743390 2653\n5436843 2587\n2651314 2580\n8722394 2542\n3542184 2427\n5862961 2292\n906358 2066\n5927996 1872\n9378983 1537\n2185191 1522\n6901374 1463\n3033749 1333\n5252571 1289\n2408923 1215\n4423556 1148\n8351511 1121\n3375935 1107\nName: count, dtype: int64\n\nPopularity of most frequent items in top 10:\n positive_listens organic_positive_listens popularity_rank\nitem_id \n6006390 8970 6813 75\n8213481 22505 17071 7\n6896222 3020 2017 818\n8743390 13518 8480 22\n5436843 14897 12127 14\n2651314 16286 11801 11\n8722394 3464 1695 634\n3542184 24075 17699 4\n5862961 26615 20620 1\n906358 22238 17183 8\n5927996 14751 10424 16\n9378983 24036 18195 5\n2185191 13059 8839 25\n6901374 25536 17347 2\n3033749 16682 12267 10\n5252571 14897 11331 13\n2408923 7515 4545 111\n4423556 6544 2177 157\n8351511 14061 9739 19\n3375935 5213 3232 282",
"description": "Analyze baseline predictions in top 10",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T05:48:13.535Z",
"content": "\nOutput: Validation ranking head:\n uid item_id rank\n0 100 6511826 1\n1 100 3972057 2\n2 100 2736766 3\n3 100 4948220 4\n4 100 5488609 5\n5 100 7958357 6\n6 100 8460142 7\n7 100 8250570 8\n8 100 5570016 9\n9 100 7510702 10\nRank range: 1 100\nNumber of unique items ranked in top 10: 592\nNumber of unique items ranked overall: 2575\nMost frequent items in top 10:\nitem_id\n6006390 2917\n8213481 2800\n6896222 2712\n8743390 2653\n5436843 2587\n2651314 2580\n8722394 2542\n3542184 2427\n5862961 2292\n906358 2066\n5927996 1872\n9378983 1537\n2185191 1522\n6901374 1463\n3033749 1333\n5252571 1289\n2408923 1215\n4423556 1148\n8351511 1121\n3375935 1107\nName: count, dtype: int64\n\nPopularity of most frequent items in top 10:\n positive_listens organic_positive_listens popularity_rank\nitem_id \n6006390 8970 6813 75\n8213481 22505 17071 7\n6896222 3020 2017 818\n8743390 13518 8480 22\n5436843 14897 12127 14\n2651314 16286 11801 11\n8722394 3464 1695 634\n3542184 24075 17699 4\n5862961 26615 20620 1\n906358 22238 17183 8\n5927996 14751 10424 16\n9378983 24036 18195 5\n2185191 13059 8839 25\n6901374 25536 17347 2\n3033749 16682 12267 10\n5252571 14897 11331 13\n2408923 7515 4545 111\n4423556 6544 2177 157\n8351511 14061 9739 19\n3375935 5213 3232 282\nProcess Group PGID: 1820\n",
"isError": false
},
"responseMessageId": "48f9b25f-99ff-43e4-9d6f-a4cb58078954"
}
]
},
{
"id": 30,
"role": "agent",
"timestamp": "2026-09-29T05:48:18.210Z",
"blocks": [
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:48:18.210Z",
"text": "Analyzing Item Rankings\n\nI'm currently examining the performance of items 6006390 and 8213481. Item 6006390 is notably popular, appearing in the top 10 for over a third of users, while item 8213481 also shows strong engagement with a similar top 10 ranking."
},
{
"kind": "tool",
"timestamp": "2026-09-29T05:48:18.210Z",
"callId": "run_shell_command__call_83741",
"name": "run_shell_command",
"input": {
"command": "python3 -c \"\nimport pandas as pd\nfrom evaluate_public import validation_truth, CandidateSet\n\ntargets_frame = pd.read_parquet('/task/data/target_users.parquet')\ntargets = set(targets_frame.uid.astype(int))\neligibility = CandidateSet('/task/data/validation_eligible_candidates.npz')\ngraded = validation_truth(eligibility, targets)\n\ncatalog = pd.read_parquet('/task/data/candidate_catalog.parquet').set_index('item_id')\n\nall_relevant_items = []\ngrades = []\nfor uid, items in graded.items():\n for item, grade in items.items():\n all_relevant_items.append(item)\n grades.append(grade)\n\ns = pd.Series(all_relevant_items)\nprint('Total graded user-item interactions:', len(s))\nprint('Unique relevant items:', s.nunique())\nprint('Grade distribution:\\n', pd.Series(grades).value_counts())\nprint('Most frequent items in ground truth:')\nprint(s.value_counts().head(20))\nprint('\\nPopularity of most frequent ground truth items:')\nprint(catalog.loc[s.value_counts().head(20).index])\n\"",
"description": "Analyze validation ground truth distribution"
},
"recorded": {
"id": "run_shell_command__call_83741",
"name": "run_shell_command",
"args": {
"command": "python3 -c \"\nimport pandas as pd\nfrom evaluate_public import validation_truth, CandidateSet\n\ntargets_frame = pd.read_parquet('/task/data/target_users.parquet')\ntargets = set(targets_frame.uid.astype(int))\neligibility = CandidateSet('/task/data/validation_eligible_candidates.npz')\ngraded = validation_truth(eligibility, targets)\n\ncatalog = pd.read_parquet('/task/data/candidate_catalog.parquet').set_index('item_id')\n\nall_relevant_items = []\ngrades = []\nfor uid, items in graded.items():\n for item, grade in items.items():\n all_relevant_items.append(item)\n grades.append(grade)\n\ns = pd.Series(all_relevant_items)\nprint('Total graded user-item interactions:', len(s))\nprint('Unique relevant items:', s.nunique())\nprint('Grade distribution:\\n', pd.Series(grades).value_counts())\nprint('Most frequent items in ground truth:')\nprint(s.value_counts().head(20))\nprint('\\nPopularity of most frequent ground truth items:')\nprint(catalog.loc[s.value_counts().head(20).index])\n\"",
"description": "Analyze validation ground truth distribution"
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_83741",
"name": "run_shell_command",
"response": {
"output": "\nOutput: Traceback (most recent call last):\n File \"\", line 3, in \nModuleNotFoundError: No module named 'evaluate_public'\nExit Code: 1\nProcess Group PGID: 1839\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:48:19.696Z",
"resultDisplay": "Traceback (most recent call last):\n File \"\", line 3, in \nModuleNotFoundError: No module named 'evaluate_public'",
"description": "Analyze validation ground truth distribution",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T05:48:19.698Z",
"content": "\nOutput: Traceback (most recent call last):\n File \"\", line 3, in \nModuleNotFoundError: No module named 'evaluate_public'\nExit Code: 1\nProcess Group PGID: 1839\n",
"isError": false
},
"responseMessageId": "530643f4-e6be-45fc-867f-eebc04b9a452"
}
]
},
{
"id": 31,
"role": "agent",
"timestamp": "2026-09-29T05:48:33.370Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:48:33.370Z",
"callId": "run_shell_command__call_34569",
"name": "run_shell_command",
"input": {
"command": "python3 -c \"\nimport sys; sys.path.append('/task')\nimport pandas as pd\nfrom evaluate_public import validation_truth, CandidateSet\n\ntargets_frame = pd.read_parquet('/task/data/target_users.parquet')\ntargets = set(targets_frame.uid.astype(int))\neligibility = CandidateSet('/task/data/validation_eligible_candidates.npz')\ngraded = validation_truth(eligibility, targets)\n\ncatalog = pd.read_parquet('/task/data/candidate_catalog.parquet').set_index('item_id')\n\nall_relevant_items = []\ngrades = []\nfor uid, items in graded.items():\n for item, grade in items.items():\n all_relevant_items.append(item)\n grades.append(grade)\n\ns = pd.Series(all_relevant_items)\nprint('Total graded user-item interactions:', len(s))\nprint('Unique relevant items:', s.nunique())\nprint('Grade distribution:\\n', pd.Series(grades).value_counts())\nprint('Most frequent items in ground truth:')\nprint(s.value_counts().head(20))\nprint('\\nPopularity of most frequent ground truth items:')\nprint(catalog.loc[s.value_counts().head(20).index])\n\"",
"description": "Analyze validation ground truth distribution"
},
"recorded": {
"id": "run_shell_command__call_34569",
"name": "run_shell_command",
"args": {
"command": "python3 -c \"\nimport sys; sys.path.append('/task')\nimport pandas as pd\nfrom evaluate_public import validation_truth, CandidateSet\n\ntargets_frame = pd.read_parquet('/task/data/target_users.parquet')\ntargets = set(targets_frame.uid.astype(int))\neligibility = CandidateSet('/task/data/validation_eligible_candidates.npz')\ngraded = validation_truth(eligibility, targets)\n\ncatalog = pd.read_parquet('/task/data/candidate_catalog.parquet').set_index('item_id')\n\nall_relevant_items = []\ngrades = []\nfor uid, items in graded.items():\n for item, grade in items.items():\n all_relevant_items.append(item)\n grades.append(grade)\n\ns = pd.Series(all_relevant_items)\nprint('Total graded user-item interactions:', len(s))\nprint('Unique relevant items:', s.nunique())\nprint('Grade distribution:\\n', pd.Series(grades).value_counts())\nprint('Most frequent items in ground truth:')\nprint(s.value_counts().head(20))\nprint('\\nPopularity of most frequent ground truth items:')\nprint(catalog.loc[s.value_counts().head(20).index])\n\"",
"description": "Analyze validation ground truth distribution"
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_34569",
"name": "run_shell_command",
"response": {
"output": "\nOutput: Total graded user-item interactions: 37769\nUnique relevant items: 14124\nGrade distribution:\n 1 33352\n3 4417\nName: count, dtype: int64\nMost frequent items in ground truth:\n6399259 173\n6252104 109\n8250570 88\n2055777 83\n1265696 77\n3134183 60\n7579671 51\n1005397 46\n4948220 45\n6625048 43\n374424 42\n5570747 42\n7419525 42\n6006390 42\n8722394 41\n6762774 41\n3361052 40\n7931842 40\n5010168 39\n7227266 39\nName: count, dtype: int64\n\nPopularity of most frequent ground truth items:\n positive_listens organic_positive_listens popularity_rank\n6399259 1276 1238 3329\n6252104 855 820 5622\n8250570 1059 1029 4263\n2055777 2465 1558 1170\n1265696 681 420 7350\n3134183 1517 720 2602\n7579671 1441 928 2789\n1005397 5546 2644 240\n4948220 740 721 6675\n6625048 7133 5065 128\n374424 323 193 16388\n5570747 4953 2891 312\n7419525 476 302 11086\n6006390 8970 6813 75\n8722394 3464 1695 634\n6762774 4331 2238 413\n3361052 1969 922 1726\n7931842 1012 427 4554\n5010168 1385 1228 2947\n7227266 2823 1726 919\nProcess Group PGID: 1850\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:48:35.475Z",
"resultDisplay": "Total graded user-item interactions: 37769\nUnique relevant items: 14124\nGrade distribution:\n 1 33352\n3 4417\nName: count, dtype: int64\nMost frequent items in ground truth:\n6399259 173\n6252104 109\n8250570 88\n2055777 83\n1265696 77\n3134183 60\n7579671 51\n1005397 46\n4948220 45\n6625048 43\n374424 42\n5570747 42\n7419525 42\n6006390 42\n8722394 41\n6762774 41\n3361052 40\n7931842 40\n5010168 39\n7227266 39\nName: count, dtype: int64\n\nPopularity of most frequent ground truth items:\n positive_listens organic_positive_listens popularity_rank\n6399259 1276 1238 3329\n6252104 855 820 5622\n8250570 1059 1029 4263\n2055777 2465 1558 1170\n1265696 681 420 7350\n3134183 1517 720 2602\n7579671 1441 928 2789\n1005397 5546 2644 240\n4948220 740 721 6675\n6625048 7133 5065 128\n374424 323 193 16388\n5570747 4953 2891 312\n7419525 476 302 11086\n6006390 8970 6813 75\n8722394 3464 1695 634\n6762774 4331 2238 413\n3361052 1969 922 1726\n7931842 1012 427 4554\n5010168 1385 1228 2947\n7227266 2823 1726 919",
"description": "Analyze validation ground truth distribution",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T05:48:36.724Z",
"content": "\nOutput: Total graded user-item interactions: 37769\nUnique relevant items: 14124\nGrade distribution:\n 1 33352\n3 4417\nName: count, dtype: int64\nMost frequent items in ground truth:\n6399259 173\n6252104 109\n8250570 88\n2055777 83\n1265696 77\n3134183 60\n7579671 51\n1005397 46\n4948220 45\n6625048 43\n374424 42\n5570747 42\n7419525 42\n6006390 42\n8722394 41\n6762774 41\n3361052 40\n7931842 40\n5010168 39\n7227266 39\nName: count, dtype: int64\n\nPopularity of most frequent ground truth items:\n positive_listens organic_positive_listens popularity_rank\n6399259 1276 1238 3329\n6252104 855 820 5622\n8250570 1059 1029 4263\n2055777 2465 1558 1170\n1265696 681 420 7350\n3134183 1517 720 2602\n7579671 1441 928 2789\n1005397 5546 2644 240\n4948220 740 721 6675\n6625048 7133 5065 128\n374424 323 193 16388\n5570747 4953 2891 312\n7419525 476 302 11086\n6006390 8970 6813 75\n8722394 3464 1695 634\n6762774 4331 2238 413\n3361052 1969 922 1726\n7931842 1012 427 4554\n5010168 1385 1228 2947\n7227266 2823 1726 919\nProcess Group PGID: 1850\n",
"isError": false
},
"responseMessageId": "76f0220e-7289-4a6c-8c70-556022a2e5bf"
}
]
},
{
"id": 32,
"role": "agent",
"timestamp": "2026-09-29T05:49:55.026Z",
"blocks": [
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Analyzing Ground Truth Rankings\n\nI'm observing a curious trend where our ground truth items, the ones users genuinely engage with, often fall within a surprisingly wide and rather low ranking range, from around 1,000 to 16,000. This is within a pool of 14,124 unique relevant items."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Investigating Popularity Bias\n\nIt appears my baseline model is heavily biased towards universally popular items, consistently ranking them highly. This is particularly striking when contrasted with the sparse engagement observed for these same items in the ground truth data. I'm delving into how the two-tower architecture, specifically within the training and negative sampling phases, might be amplifying this effect, especially when high-frequency positive targets are involved."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Refining Negative Sampling\n\nI'm examining the baseline's negative sampling strategy, specifically the uniform and in-batch methods. It seems uniform negatives are drawn from the entire catalog with a vanishingly small probability for any specific item. I'm digging into the exact implementation of in-batch negatives to understand their impact."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Discovering Uniform Sampling\n\nI've uncovered a critical flaw in the in-batch negative sampling. Instead of sampling proportionally to batch frequency, the current implementation uniformly samples from unique targets. This means popular and rare items have an equal chance of being selected as negatives, severely undermining the intended purpose and likely contributing to the observed biases."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Correcting Loss Calculations\n\nI've identified a critical issue in how in-batch negatives are handled. Specifically, popular items (track 1) are being downweighted in the loss calculation. This, combined with the uniform sampling of negatives, means their impact is drastically reduced, inadvertently reinforcing popularity bias."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Revising Negative Sampling Logic\n\nI'm reassessing how in-batch negatives function. It appears my current uniform sampling from unique targets, rather than proportional to frequency, is the key issue. I'm exploring the standard literature to ensure the in-batch negative sampling and subsequent loss calculations accurately reflect proportional sampling for effective bias correction."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Adjusting Scoring Weights\n\nI'm realizing that my scoring function needs adjustment. The formula $s'(u_i, v_j) = s(u_i, v_j) - \\log p(v_j)$ is crucial for correcting sampling bias, ensuring that rankings reflect true relevance and not just popularity. I'm investigating how the codebase applies this, noting the mention of \"sample-weighted binary cross entropy\" and \"propensity-corrected unique in-batch negatives.\""
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Revising In-Batch Sampling\n\nI've realized that my in-batch negative sampling was flawed, uniformly picking from unique targets instead of proportionally to their frequency. I'm now looking at implementing proportional sampling and correcting the loss calculation to properly account for it."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Exploring Negative Candidates\n\nI'm re-evaluating how I handle negative sampling, specifically questioning whether disliked items, which are known negatives, are being properly incorporated. I'm also investigating how the system filters out encountered items during validation and testing to ensure accurate candidate pools."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Correcting Logits Transformation\n\nI'm delving into how the raw logits are transformed and weighted within the loss calculation. It appears a softplus function is applied before correction, and I need to confirm how this interacts with the propensity weighting."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Adjusting Loss Weights\n\nI've discovered that the loss calculation incorrectly downweights popular items when they appear as negatives. This mechanism, intended to correct for sampling bias, is actually amplifying popularity bias by significantly reducing the penalty for predicting popular items. I'm now focused on reversing this logic."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Correcting Popularity Bias\n\nI'm now understanding that the previous loss calculation was indeed downweighting popular items incorrectly, exacerbating the bias. I'm actively revising this to properly penalize popular items when they appear as negatives."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Revising Loss Function\n\nI'm realizing the baseline's approach to popularity correction is flawed. Instead of directly adjusting logits with $-\\log p$, it uses inverse propensity weighting within a Binary Cross Entropy loss. This inadvertently downweights popular items in the loss, reinforcing the bias I'm trying to eliminate. I'm now focused on implementing the correct logit adjustment."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Correcting Logit Calculations\n\nI'm realizing a significant error where popular items are being penalized far less when they appear as negatives. This inverse weighting is counterproductive, effectively teaching the model to favor popular items by reducing the penalty for misclassifying them. I'm now focused on rectifying this weighting to correctly penalize popular items."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Correcting Popularity Bias\n\nI'm now rectifying the loss calculation that incorrectly downweighted popular items when they served as negatives. This correction aims to properly penalize popular items, mitigating the observed popularity bias."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Correcting Loss Weights\n\nI'm realizing the loss calculation was improperly downweighting popular items when they appeared as negatives. This correction aims to properly penalize popular items, mitigating the popularity bias."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Adjusting Loss Weights\n\nI'm confirming that the loss function was indeed downweighting popular items as negatives. I'm now adjusting this to correctly penalize them, aiming to reduce popularity bias."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Analyzing Training Efficiency\n\nI'm investigating the training bottlenecks, realizing that Python loops for building dense blocks and array operations are consuming significant time. I'm exploring profiling and optimization strategies to speed up the single epoch training time, which currently takes around 3.5 minutes."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Detecting Sampling Flaws\n\nI'm examining code snippets related to invalid negative sampling. It appears a comparison `negatives[:, :, None] == liked_items[:, None, :]` is being used, and I'm scrutinizing the shapes involved to understand its precise impact and how it might be contributing to biases."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Refining Negative Selection\n\nI'm scrutinizing the intricate logic for invalid negative sampling, particularly a broad boolean comparison that generates millions of flags per microbatch. Furthermore, I'm examining `torch.unique` operations on large item sets and their subsequent encoding, aiming to optimize this for efficiency and accuracy in negative candidate selection."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Analyzing Candidate Pools\n\nI'm realizing the core task is unseen track recommendation. My validation and testing candidate sets strictly exclude previously encountered items, meaning I must focus on genuine discovery. This clarifies the importance of accurately representing interactions and the potential pitfalls of training on data that mirrors validation/test sets too closely."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Revising Weighting Logic\n\nI'm realizing the \"organic-strong\" weighting scheme assigns equal importance to all listens, including repeats. This is problematic as repeat listens cannot represent genuine discovery, yet they're being weighted equally. I need to adjust this to specifically downweight repeat listens."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Analyzing Discovery Weighting\n\nI'm examining the \"discovery-graded\" weighting scheme. It appears to fetch \"organic-like outcomes\" and uses a \"discovery horizon\" to grade interactions. I need to understand precisely how this grading is applied to historical data and if repeat listens are being correctly handled or downweighted as suspected."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Adjusting Discovery Weighting\n\nI'm refining how repeat listens are handled. The current \"ordinary\\_target\\_weight\" is applied to them, preventing them from being considered genuine discovery outcomes and thus correctly downweighting their influence."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Investigating Default Weights\n\nI'm revisiting the default weighting scheme. It seems `organic-strong` treats all listens equally, including repeats. I need to confirm if `discovery-graded` is correctly implemented to downweight these repeats and understand why `organic-strong` is the default if it doesn't align with discovery goals."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Revising Negative Sampling Strategy\n\nI'm meticulously re-examining the negative sampling process. It appears the baseline's strategy, particularly the in-batch negative selection, is uniformly sampling from unique positive targets rather than proportionally to their frequency. I'm investigating the implications of this for popularity bias and exploring literature for optimal proportional sampling methods."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Correcting In-Batch Sampling\n\nI'm realizing the in-batch negative sampling was uniformly picking from unique targets rather than proportionally to their frequency. I'm now looking at implementing proportional sampling and correcting the loss calculation to properly account for it."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Sampling Negatives Proportionaly\n\nI'm realizing that the previous in-batch negative sampling was uniformly picking from unique targets, missing the crucial frequency proportionality. I'm now exploring sampling directly from the `targets` array to achieve this, which would make popular items significantly more likely to be selected as negatives. I'm also considering mixed negative sampling to incorporate global catalog popularity."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Revising Negative Sampling\n\nI'm now focusing on the core issue: the previous in-batch negative sampling uniformly picked from unique targets, neglecting crucial frequency proportionality. I'm exploring sampling directly from the `targets` array to ensure popular items are more likely chosen as negatives."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Correcting Correction Factors\n\nI'm analyzing the `build_item_correction` function and realizing how it generates inverse propensity weights. Items with high probabilities receive small correction factors, while those with low probabilities get large ones. This seems to be applied in the loss calculation and I'm investigating its precise impact on popularity bias."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Correcting Loss Weights\n\nI'm seeing that popular items are being heavily downweighted as negatives, with weights as low as 0.25, while unpopular items are upweighted. This is fundamentally counterproductive and amplifying popularity bias by reducing the penalty for predicting popular negatives. I am correcting this logic."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Correcting Inverse Weighting\n\nI'm realizing my previous implementation of inverse propensity weighting was flawed. By heavily downweighting popular items as negatives, it was amplifying popularity bias. I am now correcting this to apply appropriate weighting, ensuring popular items are correctly penalized."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Investigating Logit Calculation\n\nI'm realizing that my current approach to logit calculation might be amplifying popularity bias. The method of downweighting popular items when they appear as negatives is counterproductive, and I'm now focused on rectifying this to ensure accurate penalization."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Understanding Retrieval Loss Functions\n\nI'm realizing that retrievers are trained using Sampled Softmax or InfoNCE (Cross-Entropy), not Binary Cross-Entropy (BCE). InfoNCE directly optimizes for ranking the positive item above negatives, focusing gradients on the hardest negatives due to the temperature parameter, unlike BCE which treats each as an independent classification."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Testing InfoNCE Correction\n\nI'm now investigating if InfoNCE or Cross-Entropy can be leveraged for bias correction. My current thoughts are centered on testing this approach by examining how it influences logit calculations and gradients compared to BCE, aiming to directly mitigate popularity bias during training."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Testing InfoNCE Approach\n\nI'm exploring how InfoNCE and Cross-Entropy loss functions can be leveraged for bias correction. My current focus is on how these functions impact logit calculations and gradients, aiming to directly address popularity bias during training by optimizing the ranking of positive items over negatives."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:49:55.026Z",
"text": "Testing Loss Functions\n\nI am now exploring how InfoNCE and Cross-Entropy loss functions can be used for bias correction. My current focus is on how these functions affect logit calculations and gradients, aiming to directly address popularity bias during training by optimizing the ranking of positive items over negatives."
},
{
"kind": "tool",
"timestamp": "2026-09-29T05:49:55.026Z",
"callId": "run_shell_command__call_161314",
"name": "run_shell_command",
"input": {
"description": "Test validation with discovery-graded target weighting",
"command": "time python /app/solution/train_current_model.py \\\n --stage validation \\\n --public-dir /task \\\n --prepared-cache /app/solution/prepared_val.npz \\\n --target-weighting discovery-graded \\\n --output /app/solution/val_discovery_graded.parquet"
},
"recorded": {
"id": "run_shell_command__call_161314",
"name": "run_shell_command",
"args": {
"description": "Test validation with discovery-graded target weighting",
"command": "time python /app/solution/train_current_model.py \\\n --stage validation \\\n --public-dir /task \\\n --prepared-cache /app/solution/prepared_val.npz \\\n --target-weighting discovery-graded \\\n --output /app/solution/val_discovery_graded.parquet"
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_161314",
"name": "run_shell_command",
"response": {
"output": "\nOutput: loaded full-sequence cache: tokens=20,056,329 targets=20,047,194\nchronological BCE window: train=1,052,953 validation=130,103 selected=1,183,056 discarded_old_targets=18,864,138 all=20,047,194\ninbatch correction: min=0.250 mean_on_targets=1.001 max=4.000\nloaded organic-like outcomes: rows=301,527 users=7,526\ngrade-3 first-encounter targets=46,417 horizon_days=7\ntarget weighting=discovery-graded min=0.250 mean=0.310 max=3.000\nfull-history false-negative filter: users=9,135 unique_user_items=5,524,275 bytes=34,256,250\nepoch=1/1 batches=250 targets=987,417 batch_bce=0.551810 running_bce=0.544138 elapsed=198.4s\nepoch=1/1 train_loss=0.54212 validation_loss=0.50483 pair_accuracy=0.760 train_targets=1,052,953 validation_targets=130,103 train_seconds=211.6 validation_seconds=15.3\nrestored best validation-BCE checkpoint: epoch=1 loss=0.504831\nranked users=1,024/8,456\nranked users=2,048/8,456\nranked users=3,072/8,456\nranked users=4,096/8,456\nranked users=5,120/8,456\nranked users=6,144/8,456\nranked users=7,168/8,456\nranked users=8,192/8,456\nranked users=8,456/8,456\n{\n \"baseline\": \"yambda-windowed-sequence-two-tower-base\",\n \"stage\": \"validation\",\n \"output\": \"/app/solution/val_discovery_graded.parquet\",\n \"rows\": 845600,\n \"users\": 8456,\n \"runtime_seconds\": 318.361,\n \"public_inputs_only\": true,\n \"architecture\": {\n \"family\": \"pure-windowed-sequence-two-tower\",\n \"sequence_training\": \"every-adjacent-target-within-recent-chronological-window\",\n \"sequence_encoder\": \"one-layer-GRU-dense-timestep-loss\",\n \"bptt_hidden_state\": \"carried-and-detached\",\n \"model_selection\": \"chronological-validation-BCE-early-stopping\",\n \"loss_execution\": \"gradient-accumulated-microbatches\",\n \"user_id_embedding\": false,\n \"user_mlp\": \"582-512-256-192\",\n \"item_mlp\": \"196-512-256-192\",\n \"normalization\": \"none\",\n \"score\": \"raw-dot-product\",\n \"objective\": \"sample-weighted-BCE-at-every-timestep\",\n \"negative_sampling\": \"24-uniform-plus-8-propensity-corrected-inbatch\",\n \"false_negative_filter\": \"full-training-history\",\n \"target_weighting\": \"discovery-graded\"\n },\n \"config\": {\n \"bptt_steps\": 100,\n \"user_batch_size\": 64,\n \"embedding_dim\": 192,\n \"hidden_dim\": 192,\n \"user_hidden_1\": 512,\n \"user_hidden_2\": 256,\n \"item_hidden_1\": 512,\n \"item_hidden_2\": 256,\n \"dropout\": 0.05,\n \"epochs\": 1,\n \"learning_rate\": 0.001,\n \"weight_decay\": 1e-05,\n \"gradient_clip\": 2.0,\n \"loss_microbatch_size\": 65536,\n \"uniform_negatives\": 24,\n \"inbatch_negatives\": 8,\n \"inbatch_correction_power\": 0.5,\n \"inbatch_correction_min\": 0.25,\n \"inbatch_correction_max\": 4.0,\n \"negative_filter_scope\": \"full-training-history\",\n \"target_weighting\": \"discovery-graded\",\n \"ordinary_target_weight\": 0.25,\n \"organic_strong_target_weight\": 1.0,\n \"organic_like_target_weight\": 3.0,\n \"discovery_horizon_days\": 7,\n \"validation_fraction\": 0.1,\n \"max_train_targets_per_user\": 128,\n \"max_validation_targets_per_user\": 16,\n \"early_stopping_patience\": 1,\n \"early_stopping_min_delta\": 0.0001,\n \"seed\": 29\n },\n \"sequence_tokens\": 20056329,\n \"all_next_item_targets\": 20047194,\n \"selected_training_targets\": 1052953,\n \"selected_validation_targets\": 130103,\n \"discarded_old_targets\": 18864138,\n \"training_history\": [\n {\n \"epoch\": 1.0,\n \"loss\": 0.5421154213454774,\n \"positive_logit\": 0.7774311381163012,\n \"negative_logit\": -0.6407561985568945,\n \"pair_accuracy\": 0.7602795423917307,\n \"targets\": 1052953.0,\n \"target_weight_sum\": 348780.0,\n \"batches\": 266.0,\n \"seconds\": 211.6158940570001,\n \"validation_loss\": 0.5048305705104994,\n \"validation_targets\": 130103.0,\n \"validation_target_weight_sum\": 42129.5,\n \"validation_seconds\": 15.314511568999933\n }\n ],\n \"embedding_diagnostics\": {\n \"query_norm_mean\": 1.7521852254867554,\n \"query_norm_std\": 0.8638167977333069,\n \"item_norm_mean\": 1.8543914556503296,\n \"item_norm_std\": 0.8459439873695374\n },\n \"torch_version\": \"2.8.0+cpu\",\n \"device\": \"cpu\"\n}\n\nreal\t5m22.684s\nuser\t29m58.539s\nsys\t1m42.847s\nProcess Group PGID: 1882\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:55:17.724Z",
"resultDisplay": "loaded full-sequence cache: tokens=20,056,329 targets=20,047,194\nchronological BCE window: train=1,052,953 validation=130,103 selected=1,183,056 discarded_old_targets=18,864,138 all=20,047,194\ninbatch correction: min=0.250 mean_on_targets=1.001 max=4.000\nloaded organic-like outcomes: rows=301,527 users=7,526\ngrade-3 first-encounter targets=46,417 horizon_days=7\ntarget weighting=discovery-graded min=0.250 mean=0.310 max=3.000\nfull-history false-negative filter: users=9,135 unique_user_items=5,524,275 bytes=34,256,250\nepoch=1/1 batches=250 targets=987,417 batch_bce=0.551810 running_bce=0.544138 elapsed=198.4s\nepoch=1/1 train_loss=0.54212 validation_loss=0.50483 pair_accuracy=0.760 train_targets=1,052,953 validation_targets=130,103 train_seconds=211.6 validation_seconds=15.3\nrestored best validation-BCE checkpoint: epoch=1 loss=0.504831\nranked users=1,024/8,456\nranked users=2,048/8,456\nranked users=3,072/8,456\nranked users=4,096/8,456\nranked users=5,120/8,456\nranked users=6,144/8,456\nranked users=7,168/8,456\nranked users=8,192/8,456\nranked users=8,456/8,456\n{\n \"baseline\": \"yambda-windowed-sequence-two-tower-base\",\n \"stage\": \"validation\",\n \"output\": \"/app/solution/val_discovery_graded.parquet\",\n \"rows\": 845600,\n \"users\": 8456,\n \"runtime_seconds\": 318.361,\n \"public_inputs_only\": true,\n \"architecture\": {\n \"family\": \"pure-windowed-sequence-two-tower\",\n \"sequence_training\": \"every-adjacent-target-within-recent-chronological-window\",\n \"sequence_encoder\": \"one-layer-GRU-dense-timestep-loss\",\n \"bptt_hidden_state\": \"carried-and-detached\",\n \"model_selection\": \"chronological-validation-BCE-early-stopping\",\n \"loss_execution\": \"gradient-accumulated-microbatches\",\n \"user_id_embedding\": false,\n \"user_mlp\": \"582-512-256-192\",\n \"item_mlp\": \"196-512-256-192\",\n \"normalization\": \"none\",\n \"score\": \"raw-dot-product\",\n \"objective\": \"sample-weighted-BCE-at-every-timestep\",\n \"negative_sampling\": \"24-uniform-plus-8-propensity-corrected-inbatch\",\n \"false_negative_filter\": \"full-training-history\",\n \"target_weighting\": \"discovery-graded\"\n },\n \"config\": {\n \"bptt_steps\": 100,\n \"user_batch_size\": 64,\n \"embedding_dim\": 192,\n \"hidden_dim\": 192,\n \"user_hidden_1\": 512,\n \"user_hidden_2\": 256,\n \"item_hidden_1\": 512,\n \"item_hidden_2\": 256,\n \"dropout\": 0.05,\n \"epochs\": 1,\n \"learning_rate\": 0.001,\n \"weight_decay\": 1e-05,\n \"gradient_clip\": 2.0,\n \"loss_microbatch_size\": 65536,\n \"uniform_negatives\": 24,\n \"inbatch_negatives\": 8,\n \"inbatch_correction_power\": 0.5,\n \"inbatch_correction_min\": 0.25,\n \"inbatch_correction_max\": 4.0,\n \"negative_filter_scope\": \"full-training-history\",\n \"target_weighting\": \"discovery-graded\",\n \"ordinary_target_weight\": 0.25,\n \"organic_strong_target_weight\": 1.0,\n \"organic_like_target_weight\": 3.0,\n \"discovery_horizon_days\": 7,\n \"validation_fraction\": 0.1,\n \"max_train_targets_per_user\": 128,\n \"max_validation_targets_per_user\": 16,\n \"early_stopping_patience\": 1,\n \"early_stopping_min_delta\": 0.0001,\n \"seed\": 29\n },\n \"sequence_tokens\": 20056329,\n \"all_next_item_targets\": 20047194,\n \"selected_training_targets\": 1052953,\n \"selected_validation_targets\": 130103,\n \"discarded_old_targets\": 18864138,\n \"training_history\": [\n {\n \"epoch\": 1.0,\n \"loss\": 0.5421154213454774,\n \"positive_logit\": 0.7774311381163012,\n \"negative_logit\": -0.6407561985568945,\n \"pair_accuracy\": 0.7602795423917307,\n \"targets\": 1052953.0,\n \"target_weight_sum\": 348780.0,\n \"batches\": 266.0,\n \"seconds\": 211.6158940570001,\n \"validation_loss\": 0.5048305705104994,\n \"validation_targets\": 130103.0,\n \"validation_target_weight_sum\": 42129.5,\n \"validation_seconds\": 15.314511568999933\n }\n ],\n \"embedding_diagnostics\": {\n \"query_norm_mean\": 1.7521852254867554,\n \"query_norm_std\": 0.8638167977333069,\n \"item_norm_mean\": 1.8543914556503296,\n \"item_norm_std\": 0.8459439873695374\n },\n \"torch_version\": \"2.8.0+cpu\",\n \"device\": \"cpu\"\n}\n\nreal\t5m22.684s\nuser\t29m58.539s\nsys\t1m42.847s",
"description": "Test validation with discovery-graded target weighting",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T05:55:17.726Z",
"content": "\nOutput: loaded full-sequence cache: tokens=20,056,329 targets=20,047,194\nchronological BCE window: train=1,052,953 validation=130,103 selected=1,183,056 discarded_old_targets=18,864,138 all=20,047,194\ninbatch correction: min=0.250 mean_on_targets=1.001 max=4.000\nloaded organic-like outcomes: rows=301,527 users=7,526\ngrade-3 first-encounter targets=46,417 horizon_days=7\ntarget weighting=discovery-graded min=0.250 mean=0.310 max=3.000\nfull-history false-negative filter: users=9,135 unique_user_items=5,524,275 bytes=34,256,250\nepoch=1/1 batches=250 targets=987,417 batch_bce=0.551810 running_bce=0.544138 elapsed=198.4s\nepoch=1/1 train_loss=0.54212 validation_loss=0.50483 pair_accuracy=0.760 train_targets=1,052,953 validation_targets=130,103 train_seconds=211.6 validation_seconds=15.3\nrestored best validation-BCE checkpoint: epoch=1 loss=0.504831\nranked users=1,024/8,456\nranked users=2,048/8,456\nranked users=3,072/8,456\nranked users=4,096/8,456\nranked users=5,120/8,456\nranked users=6,144/8,456\nranked users=7,168/8,456\nranked users=8,192/8,456\nranked users=8,456/8,456\n{\n \"baseline\": \"yambda-windowed-sequence-two-tower-base\",\n \"stage\": \"validation\",\n \"output\": \"/app/solution/val_discovery_graded.parquet\",\n \"rows\": 845600,\n \"users\": 8456,\n \"runtime_seconds\": 318.361,\n \"public_inputs_only\": true,\n \"architecture\": {\n \"family\": \"pure-windowed-sequence-two-tower\",\n \"sequence_training\": \"every-adjacent-target-within-recent-chronological-window\",\n \"sequence_encoder\": \"one-layer-GRU-dense-timestep-loss\",\n \"bptt_hidden_state\": \"carried-and-detached\",\n \"model_selection\": \"chronological-validation-BCE-early-stopping\",\n \"loss_execution\": \"gradient-accumulated-microbatches\",\n \"user_id_embedding\": false,\n \"user_mlp\": \"582-512-256-192\",\n \"item_mlp\": \"196-512-256-192\",\n \"normalization\": \"none\",\n \"score\": \"raw-dot-product\",\n \"objective\": \"sample-weighted-BCE-at-every-timestep\",\n \"negative_sampling\": \"24-uniform-plus-8-propensity-corrected-inbatch\",\n \"false_negative_filter\": \"full-training-history\",\n \"target_weighting\": \"discovery-graded\"\n },\n \"config\": {\n \"bptt_steps\": 100,\n \"user_batch_size\": 64,\n \"embedding_dim\": 192,\n \"hidden_dim\": 192,\n \"user_hidden_1\": 512,\n \"user_hidden_2\": 256,\n \"item_hidden_1\": 512,\n \"item_hidden_2\": 256,\n \"dropout\": 0.05,\n \"epochs\": 1,\n \"learning_rate\": 0.001,\n \"weight_decay\": 1e-05,\n \"gradient_clip\": 2.0,\n \"loss_microbatch_size\": 65536,\n \"uniform_negatives\": 24,\n \"inbatch_negatives\": 8,\n \"inbatch_correction_power\": 0.5,\n \"inbatch_correction_min\": 0.25,\n \"inbatch_correction_max\": 4.0,\n \"negative_filter_scope\": \"full-training-history\",\n \"target_weighting\": \"discovery-graded\",\n \"ordinary_target_weight\": 0.25,\n \"organic_strong_target_weight\": 1.0,\n \"organic_like_target_weight\": 3.0,\n \"discovery_horizon_days\": 7,\n \"validation_fraction\": 0.1,\n \"max_train_targets_per_user\": 128,\n \"max_validation_targets_per_user\": 16,\n \"early_stopping_patience\": 1,\n \"early_stopping_min_delta\": 0.0001,\n \"seed\": 29\n },\n \"sequence_tokens\": 20056329,\n \"all_next_item_targets\": 20047194,\n \"selected_training_targets\": 1052953,\n \"selected_validation_targets\": 130103,\n \"discarded_old_targets\": 18864138,\n \"training_history\": [\n {\n \"epoch\": 1.0,\n \"loss\": 0.5421154213454774,\n \"positive_logit\": 0.7774311381163012,\n \"negative_logit\": -0.6407561985568945,\n \"pair_accuracy\": 0.7602795423917307,\n \"targets\": 1052953.0,\n \"target_weight_sum\": 348780.0,\n \"batches\": 266.0,\n \"seconds\": 211.6158940570001,\n \"validation_loss\": 0.5048305705104994,\n \"validation_targets\": 130103.0,\n \"validation_target_weight_sum\": 42129.5,\n \"validation_seconds\": 15.314511568999933\n }\n ],\n \"embedding_diagnostics\": {\n \"query_norm_mean\": 1.7521852254867554,\n \"query_norm_std\": 0.8638167977333069,\n \"item_norm_mean\": 1.8543914556503296,\n \"item_norm_std\": 0.8459439873695374\n },\n \"torch_version\": \"2.8.0+cpu\",\n \"device\": \"cpu\"\n}\n\nreal\t5m22.684s\nuser\t29m58.539s\nsys\t1m42.847s\nProcess Group PGID: 1882\n",
"isError": false
},
"responseMessageId": "8c92a06a-74cd-4b13-8d2e-ad125c0f35d2"
}
]
},
{
"id": 33,
"role": "agent",
"timestamp": "2026-09-29T05:55:25.338Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T05:55:25.338Z",
"callId": "run_shell_command__call_4838876",
"name": "run_shell_command",
"input": {
"description": "Evaluate discovery-graded validation ranking",
"command": "python /task/evaluate_public.py /app/solution/val_discovery_graded.parquet"
},
"recorded": {
"id": "run_shell_command__call_4838876",
"name": "run_shell_command",
"args": {
"description": "Evaluate discovery-graded validation ranking",
"command": "python /task/evaluate_public.py /app/solution/val_discovery_graded.parquet"
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_4838876",
"name": "run_shell_command",
"response": {
"output": "\nOutput: {\n \"model\": \"/app/solution/val_discovery_graded.parquet\",\n \"scores\": {\n \"quality_ndcg@10\": 0.023522,\n \"quality_users\": 4263\n }\n}\nProcess Group PGID: 1943\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:55:28.707Z",
"resultDisplay": "{\n \"model\": \"/app/solution/val_discovery_graded.parquet\",\n \"scores\": {\n \"quality_ndcg@10\": 0.023522,\n \"quality_users\": 4263\n }\n}",
"description": "python /task/evaluate_public.py /app/solution/val_discovery_graded.parquet",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T05:55:28.709Z",
"content": "\nOutput: {\n \"model\": \"/app/solution/val_discovery_graded.parquet\",\n \"scores\": {\n \"quality_ndcg@10\": 0.023522,\n \"quality_users\": 4263\n }\n}\nProcess Group PGID: 1943\n",
"isError": false
},
"responseMessageId": "b7cda05d-8572-47d6-81e3-634bad7d4fb0"
}
]
},
{
"id": 34,
"role": "agent",
"timestamp": "2026-09-29T05:55:32.368Z",
"blocks": [
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:55:32.368Z",
"text": "Improving NDCG@10 Performance\n\nI'm pleased to report a significant 10.5% relative improvement in NDCG@10, reaching 0.02352, within a swift five-minute runtime. My next steps involve documenting these findings and shifting focus to diagnosing and experimenting with negative sampling strategies."
},
{
"kind": "tool",
"timestamp": "2026-09-29T05:55:32.368Z",
"callId": "update_topic__call_4844570",
"name": "update_topic",
"input": {
"summary": "Completed the initial exploration and benchmarking of the production baseline. Established the baseline validation NDCG@10 at 0.02128. Confirmed that the model suffers from severe popularity bias where a small fraction of catalog items dominate recommendations for almost all users. Verified that using discovery-graded target weighting improves NDCG@10 to 0.02352. Now focusing on diagnosing and fixing the negative sampling and loss objective under the CPU compute budget.",
"strategic_intent": "Investigate negative sampling limitations and test improved negative sampling strategies.",
"title": "Negative Sampling Investigation and Objective Optimization"
},
"recorded": {
"id": "update_topic__call_4844570",
"name": "update_topic",
"args": {
"summary": "Completed the initial exploration and benchmarking of the production baseline. Established the baseline validation NDCG@10 at 0.02128. Confirmed that the model suffers from severe popularity bias where a small fraction of catalog items dominate recommendations for almost all users. Verified that using discovery-graded target weighting improves NDCG@10 to 0.02352. Now focusing on diagnosing and fixing the negative sampling and loss objective under the CPU compute budget.",
"strategic_intent": "Investigate negative sampling limitations and test improved negative sampling strategies.",
"title": "Negative Sampling Investigation and Objective Optimization"
},
"result": [
{
"functionResponse": {
"id": "update_topic__call_4844570",
"name": "update_topic",
"response": {
"output": "Current topic: \"Negative Sampling Investigation and Objective Optimization\"\nTopic summary: Completed the initial exploration and benchmarking of the production baseline. Established the baseline validation NDCG@10 at 0.02128. Confirmed that the model suffers from severe popularity bias where a small fraction of catalog items dominate recommendations for almost all users. Verified that using discovery-graded target weighting improves NDCG@10 to 0.02352. Now focusing on diagnosing and fixing the negative sampling and loss objective under the CPU compute budget.\n\nStrategic Intent: Investigate negative sampling limitations and test improved negative sampling strategies."
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:55:32.371Z",
"resultDisplay": "## 📂 Topic: **Negative Sampling Investigation and Objective Optimization**\n\n**Summary:**\nCompleted the initial exploration and benchmarking of the production baseline. Established the baseline validation NDCG@10 at 0.02128. Confirmed that the model suffers from severe popularity bias where a small fraction of catalog items dominate recommendations for almost all users. Verified that using discovery-graded target weighting improves NDCG@10 to 0.02352. Now focusing on diagnosing and fixing the negative sampling and loss objective under the CPU compute budget.\n\n> [!STRATEGY]\n> **Intent:** Investigate negative sampling limitations and test improved negative sampling strategies.",
"description": "Update topic to: \"Negative Sampling Investigation and Objective Optimization\"",
"displayName": "Update Topic Context",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T05:55:32.372Z",
"content": "Current topic: \"Negative Sampling Investigation and Objective Optimization\"\nTopic summary: Completed the initial exploration and benchmarking of the production baseline. Established the baseline validation NDCG@10 at 0.02128. Confirmed that the model suffers from severe popularity bias where a small fraction of catalog items dominate recommendations for almost all users. Verified that using discovery-graded target weighting improves NDCG@10 to 0.02352. Now focusing on diagnosing and fixing the negative sampling and loss objective under the CPU compute budget.\n\nStrategic Intent: Investigate negative sampling limitations and test improved negative sampling strategies.",
"isError": false
},
"responseMessageId": "326f85f2-798e-40d6-b471-b4dc74ff8415"
}
]
},
{
"id": 35,
"role": "agent",
"timestamp": "2026-09-29T05:56:12.119Z",
"blocks": [
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:56:12.119Z",
"text": "Investigating Negative Sampling\n\nI'm currently scrutinizing how the training objective selects and utilizes negative examples, aiming to pinpoint any limitations within the current sampling approach."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:56:12.119Z",
"text": "Refining Sampling Techniques\n\nI'm now focusing on optimizing the negative sampling strategy to operate efficiently within the existing CPU constraints, ensuring all changes are rigorously validated against the public backtest while respecting established data and serving boundaries."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:56:12.119Z",
"text": "Optimizing In-Batch Sampling\n\nI'm refining the in-batch negative sampling mechanism. My focus is on intelligently selecting these negatives, particularly when unique target items are scarce, to maximize their impact within the current batch."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:56:12.119Z",
"text": "Analyzing Sampling Flaws\n\nI've identified a critical flaw in my in-batch negative sampling: `torch.unique` normalizes item frequencies within a batch to one. This undermines the empirical frequency distribution, causing rare and popular items to be sampled equally. I'm now investigating how the subsequent `correction` mechanism attempts to address this."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:56:12.119Z",
"text": "Investigating Correction Mechanism\n\nI'm now scrutinizing how the correction mechanism operates within the loss calculation, specifically examining its interaction with the weights assigned to popular versus rare items."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:56:12.119Z",
"text": "Revisiting Loss Calculations\n\nI'm now revisiting the loss calculations, specifically comparing InfoNCE and BCE with in-batch negatives. I'm analyzing how the log-Q correction and temperature parameter interact with sampling probabilities for both popular and rare items, seeking to understand gradient flow."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:56:12.119Z",
"text": "Evaluating Temperature Scaling\n\nI'm now analyzing the impact of the temperature parameter on logit scaling within my current loss formulation. I'm comparing how different temperature values, particularly when vectors are L2-normalized, influence the resulting gradients and their effectiveness for ranking."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:56:12.119Z",
"text": "Analyzing Logit Interactions\n\nI'm now examining how L2 normalization of item and user vectors impacts the dot product, effectively turning it into cosine similarity. I'm investigating if this affects the subsequent logit interactions within the loss calculation."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:56:12.119Z",
"text": "Re-evaluating Score Calculation\n\nI'm verifying if the serving path truly requires a raw dot product. My current analysis confirms that the standard two-tower score calculation, `queries @ items.T`, is indeed a dot product, and I'm now moving to analyze the limitations within the current negative sampling strategy in `train_current_model.py`."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:56:12.119Z",
"text": "Investigating Weighting Imbalance\n\nI've discovered a significant issue where popular items receive negligible penalty in the negative loss, preventing the model from learning to avoid recommending them inappropriately. This imbalance needs immediate correction."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:56:12.119Z",
"text": "Correcting Sampling Bias\n\nI've identified a critical sampling flaw: `torch.unique` normalizes item frequencies, leading to uniform sampling of negatives regardless of their actual occurrence. This is further compounded by an incorrect 0.25 downweighting, effectively preventing popular items from serving as effective negatives. I'm now focused on rectifying this bias to ensure a more balanced learning signal."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:56:12.119Z",
"text": "Correcting Negative Sample Loss\n\nI'm now addressing a significant issue where popular items receive minimal penalty in the negative loss, hindering the model's ability to learn appropriate avoidance. This bias must be rectified."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:56:12.119Z",
"text": "Analyzing Gradient Imbalance\n\nI'm currently analyzing a critical imbalance where positive gradients are significantly stronger than negative ones, up to 32x. This prevents sharp discrimination. I'm exploring InfoNCE as a solution that forces competition and directs gradients towards harder negatives."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:56:12.119Z",
"text": "Testing Sampling Fixes\n\nI'm now preparing to empirically test specific fixes for the negative sampling strategy. My focus is on removing the inverse propensity downweighting of popular negatives to observe its impact on NDCG@10."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:56:12.119Z",
"text": "Testing Loss Variations\n\nI am now exploring the impact of different loss functions and sampling strategies. Specifically, I am investigating whether sampling negatives from `targets` or catalog popularity offers improvements, and testing `inbatch_correction_power = 0.0` and InfoNCE loss to analyze gradient behavior and potential for sharper discrimination."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:56:12.119Z",
"text": "Analyzing Loss Function Interactions\n\nI'm currently re-evaluating how the chosen loss functions interact with sampling strategies. I'm focusing on the implications of negative sampling probabilities and potential corrections for popular items to ensure effective discrimination."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T05:56:12.119Z",
"text": "Testing Temperature and InfoNCE\n\nI'm investigating the impact of InfoNCE loss and temperature scaling ($\\tau$). Initial checks suggest a low $\\tau$ (like 0.2) with L2-normalized vectors could drastically improve pair accuracy by making the difference between positive and negative scores much larger, potentially enabling 90%+ accuracy. I'm now assessing the feasibility and speed of running validation experiments."
},
{
"kind": "tool",
"timestamp": "2026-09-29T05:56:12.119Z",
"callId": "run_shell_command__call_42331",
"name": "run_shell_command",
"input": {
"command": "python3 -c \"\nimport time\nimport numpy as np\nimport torch\n\nB = 4000\nK = 32\nL = 48\nnegatives = np.random.randint(0, 30000, (B, K))\ntargets = np.random.randint(0, 30000, B)\nliked_items = np.random.randint(0, 30000, (B, L))\nuser_rows = np.random.randint(0, 9000, B)\nknown_packed = np.zeros((9000, 3750), dtype=np.uint8)\n\nt0 = time.time()\nfor _ in range(10):\n byte_cols = negatives >> 3\n bit_cols = negatives & 7\n known = known_packed[user_rows[:, None], byte_cols]\n invalid = ((known >> bit_cols) & 1).astype(bool)\n invalid |= negatives == targets[:, None]\n invalid |= (negatives[:, :, None] == liked_items[:, None, :]).any(axis=2)\n for col in range(1, K):\n invalid[:, col] |= (negatives[:, col, None] == negatives[:, :col]).any(axis=1)\nt1 = time.time()\nprint('Time for 10 checks:', t1 - t0, 'seconds (per call:', (t1 - t0)/10, ')')\n\"",
"description": "Benchmark invalid_full_history_negatives"
},
"recorded": {
"id": "run_shell_command__call_42331",
"name": "run_shell_command",
"args": {
"command": "python3 -c \"\nimport time\nimport numpy as np\nimport torch\n\nB = 4000\nK = 32\nL = 48\nnegatives = np.random.randint(0, 30000, (B, K))\ntargets = np.random.randint(0, 30000, B)\nliked_items = np.random.randint(0, 30000, (B, L))\nuser_rows = np.random.randint(0, 9000, B)\nknown_packed = np.zeros((9000, 3750), dtype=np.uint8)\n\nt0 = time.time()\nfor _ in range(10):\n byte_cols = negatives >> 3\n bit_cols = negatives & 7\n known = known_packed[user_rows[:, None], byte_cols]\n invalid = ((known >> bit_cols) & 1).astype(bool)\n invalid |= negatives == targets[:, None]\n invalid |= (negatives[:, :, None] == liked_items[:, None, :]).any(axis=2)\n for col in range(1, K):\n invalid[:, col] |= (negatives[:, col, None] == negatives[:, :col]).any(axis=1)\nt1 = time.time()\nprint('Time for 10 checks:', t1 - t0, 'seconds (per call:', (t1 - t0)/10, ')')\n\"",
"description": "Benchmark invalid_full_history_negatives"
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_42331",
"name": "run_shell_command",
"response": {
"output": "\nOutput: Time for 10 checks: 0.14188003540039062 seconds (per call: 0.014188003540039063 )\nProcess Group PGID: 1969\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T05:56:15.218Z",
"resultDisplay": "Time for 10 checks: 0.14188003540039062 seconds (per call: 0.014188003540039063 )",
"description": "Benchmark invalid_full_history_negatives",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T05:56:15.220Z",
"content": "\nOutput: Time for 10 checks: 0.14188003540039062 seconds (per call: 0.014188003540039063 )\nProcess Group PGID: 1969\n",
"isError": false
},
"responseMessageId": "6869ebad-8a9d-4e17-98f5-6d0098f8d325"
}
]
},
{
"id": 36,
"role": "agent",
"timestamp": "2026-09-29T06:01:51.560Z",
"blocks": [
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:01:51.560Z",
"text": "Analyzing Batch Performance\n\nMy recent focus has been on understanding the performance bottleneck within batch processing. I've observed that individual calls are quite fast, around 14ms. However, I'm now investigating the overhead associated with `candidate_vectors`, specifically how the concatenation, unique item identification, and subsequent encoding steps contribute to the overall ~13-second processing time for 266 batches."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:01:51.560Z",
"text": "Revising Loss Function\n\nI'm now re-evaluating the training objective. The current sample-weighted Binary Cross-Entropy with softplus on negative logits, averaging across negatives, is problematic for retrieval tasks that prioritize relative ranking. I'm exploring alternatives like InfoNCE or sampled softmax to better address the lack of competition among candidates and improve gradient flow."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:01:51.560Z",
"text": "Refining Negative Sampling\n\nMy thoughts are now intensely focused on the severe flaws discovered in the in-batch negative sampling strategy. The uniform sampling of unique targets, rather than popularity-weighted sampling, and the subsequent inverse-propensity weighting are actively hindering performance by disproportionately downweighting popular items, leading to near-zero negative penalties for them."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:01:51.560Z",
"text": "Investigating Negative Examples\n\nI'm now scrutinizing how explicit dislikes are being handled. It appears they are entirely overlooked as negative training examples, which is a significant oversight given their high confidence. Furthermore, the current scope of negative filtering, while good at excluding positives, doesn't seem to account for these explicitly disliked items."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:01:51.560Z",
"text": "Fixing Negative Sampling\n\nMy current focus is on rectifying the in-batch negative sampling. I'm investigating frequency-based sampling from `targets`, or popularity-weighted sampling, alongside evaluating InfoNCE loss and removing the downweighting of popular items in the negative loss."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:01:51.560Z",
"text": "Revising Negative Loss Calculation\n\nI'm digging into the `corrected_negative_loss` function, specifically how the division by `weights.sum(dim=1)` might be problematic. This calculation appears to be a key area for potential optimization and refinement in how negative examples are weighted."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:01:51.560Z",
"text": "Revisiting Sampling Logic\n\nI'm reassessing the core mathematical objective of in-batch negative sampling, particularly how log-Q correction impacts logits. My current hypothesis is that the correction should be a subtraction from the logit before exponentiation, not an external weight applied later, which seems to be where my implementation diverges."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:01:51.560Z",
"text": "Adjusting Logit Corrections\n\nI'm reassessing how log-Q corrections are applied. My thinking now is that subtracting $-\\log Q(j)$ from logits before exponentiation in the softmax can better adjust for popularity bias. This approach seems to align with correcting for biased sampling in logistic regression, ensuring learned preferences reflect true relevance."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:01:51.560Z",
"text": "Correcting Logit Bias\n\nI've realized my previous approach to log-Q correction was backward. Adding $\\log Q(i)$ to negative logits actually increases the penalty for popular negatives, which is the desired effect. The baseline code's multiplicative approach actually diminished this penalty. I am now investigating if InfoNCE or sampled softmax offers further advantages over BCE for this ranking task."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:01:51.560Z",
"text": "Testing InfoNCE Implementation\n\nI'm now focused on verifying the practical impact of replacing Binary Cross-Entropy with InfoNCE loss within `score_target_microbatches`. My current thinking involves directly implementing this change and conducting validation runs to observe its performance characteristics, particularly regarding hard negative focus and gradient behavior."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:01:51.560Z",
"text": "Validating Gradient Flow\n\nI'm now carefully examining the gradient propagation during the `backward` pass in `score_target_microbatches`. My focus is on confirming that gradients are correctly flowing back to `sequence_input` and the model parameters after the loss calculation, ensuring effective learning from the entire process."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:01:51.560Z",
"text": "Verifying Gradient Propagation\n\nI'm now confirming that the `backward` call correctly propagates gradients through the GRU for any scalar loss sum, including when using InfoNCE. This also applies to `validation_loss` calculation in `evaluate_validation_bce`, ensuring early stopping based on validation InfoNCE loss is accurate. I'm also investigating the sampling logic for negatives in `draw_full_history_filtered_negatives`."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:01:51.560Z",
"text": "Refining Sampling Strategy\n\nI'm now questioning the necessity of using `unique_targets` for sampling in-batch negatives. My current hypothesis is that sampling directly from `targets_cpu` might naturally increase the probability of selecting more frequent items, thereby addressing popularity bias without explicit weighting. I'm also considering how to handle cases where a sampled negative might inadvertently be the actual target."
},
{
"kind": "tool",
"timestamp": "2026-09-29T06:01:51.560Z",
"callId": "run_shell_command__call_5973172",
"name": "run_shell_command",
"input": {
"description": "Benchmark torch.multinomial",
"command": "python3 -c \"\nimport time\nimport torch\n\nweights = torch.rand(30000)\nt0 = time.time()\nfor _ in range(10):\n samples = torch.multinomial(weights, 6400 * 16, replacement=True)\nt1 = time.time()\nprint('Time for 10 multinomial draws of 6400*16:', t1 - t0, 's (per draw:', (t1 - t0)/10, ')')\n\""
},
"recorded": {
"id": "run_shell_command__call_5973172",
"name": "run_shell_command",
"args": {
"description": "Benchmark torch.multinomial",
"command": "python3 -c \"\nimport time\nimport torch\n\nweights = torch.rand(30000)\nt0 = time.time()\nfor _ in range(10):\n samples = torch.multinomial(weights, 6400 * 16, replacement=True)\nt1 = time.time()\nprint('Time for 10 multinomial draws of 6400*16:', t1 - t0, 's (per draw:', (t1 - t0)/10, ')')\n\""
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_5973172",
"name": "run_shell_command",
"response": {
"output": "\nOutput: Time for 10 multinomial draws of 6400*16: 0.11437463760375977 s (per draw: 0.011437463760375976 )\nProcess Group PGID: 2009\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T06:01:54.553Z",
"resultDisplay": "Time for 10 multinomial draws of 6400*16: 0.11437463760375977 s (per draw: 0.011437463760375976 )",
"description": "Benchmark torch.multinomial",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T06:01:54.555Z",
"content": "\nOutput: Time for 10 multinomial draws of 6400*16: 0.11437463760375977 s (per draw: 0.011437463760375976 )\nProcess Group PGID: 2009\n",
"isError": false
},
"responseMessageId": "bf45d9ea-cb9c-489b-8300-1a9b30114660"
}
]
},
{
"id": 37,
"role": "agent",
"timestamp": "2026-09-29T06:05:05.286Z",
"blocks": [
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:05:05.285Z",
"text": "Exploring Catalog Sampling\n\nI'm focusing on how sampling negatives from catalog popularity, specifically with an exponent $\\alpha$ like 0.5 or 0.75, impacts batch quality. This approach seems to consistently yield high-quality negatives drawn directly from the catalog distribution."
},
{
"kind": "tool",
"timestamp": "2026-09-29T06:05:05.286Z",
"callId": "run_shell_command__call_234661",
"name": "run_shell_command",
"input": {
"description": "Inspect train_dislikes statistics",
"command": "python3 -c \"\nimport pandas as pd\nimport numpy as np\n\ndislikes = pd.read_parquet('/task/data/train_dislikes.parquet')\nprint('Dislikes users:', dislikes.uid.nunique())\nprint('Total dislike events:', len(dislikes))\nprint('Dislikes per user min/mean/max:', dislikes.groupby('uid').size().min(), dislikes.groupby('uid').size().mean(), dislikes.groupby('uid').size().max())\n\""
},
"recorded": {
"id": "run_shell_command__call_234661",
"name": "run_shell_command",
"args": {
"description": "Inspect train_dislikes statistics",
"command": "python3 -c \"\nimport pandas as pd\nimport numpy as np\n\ndislikes = pd.read_parquet('/task/data/train_dislikes.parquet')\nprint('Dislikes users:', dislikes.uid.nunique())\nprint('Total dislike events:', len(dislikes))\nprint('Dislikes per user min/mean/max:', dislikes.groupby('uid').size().min(), dislikes.groupby('uid').size().mean(), dislikes.groupby('uid').size().max())\n\""
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_234661",
"name": "run_shell_command",
"response": {
"output": "\nOutput: Dislikes users: 5732\nTotal dislike events: 101075\nDislikes per user min/mean/max: 1 17.633461270062806 2985\nProcess Group PGID: 2037\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T06:05:06.832Z",
"resultDisplay": "Dislikes users: 5732\nTotal dislike events: 101075\nDislikes per user min/mean/max: 1 17.633461270062806 2985",
"description": "Inspect train_dislikes statistics",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T06:05:06.834Z",
"content": "\nOutput: Dislikes users: 5732\nTotal dislike events: 101075\nDislikes per user min/mean/max: 1 17.633461270062806 2985\nProcess Group PGID: 2037\n",
"isError": false
},
"responseMessageId": "601701f1-31be-4948-8d9b-84da3a99e39d"
}
]
},
{
"id": 38,
"role": "agent",
"timestamp": "2026-09-29T06:05:24.884Z",
"blocks": [
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:05:24.884Z",
"text": "Analyzing Dislike Data\n\nMy current focus is on understanding the \"dislike\" data. I've noted that a significant number of users have dislikes, averaging around 17.6 per user. I'm also confirming that `prepared.disliked_states` is already constructed within `train_current_model.py` and is being populated at each timestep with the relevant state information."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:05:24.884Z",
"text": "Exploring Loss Functions\n\nI've been investigating the impact of different loss functions on model performance. Specifically, I'm now focused on comparing Softmax/InfoNCE loss against BCE loss and exploring the efficacy of various negative sampling strategies, including popularity-based, uniform, and in-batch negatives, along with hard negatives derived from user dislikes. I am planning an experiment to test InfoNCE loss."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:05:24.884Z",
"text": "Refining Loss Calculations\n\nI'm now diving into the specifics of the InfoNCE loss implementation. I'm carefully examining how the `corrected_negative_loss` interacts with the `F.cross_entropy` function, especially when incorporating proposal distribution corrections (`log_Q`) and temperature scaling. My next step is to ensure this correction is applied correctly within the sampled softmax framework."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:05:24.884Z",
"text": "Testing InfoNCE Objective\n\nI'm now focused on empirically validating the InfoNCE loss. I plan to introduce an `--objective` flag to `train_current_model.py` to easily switch between BCE and InfoNCE, and will then run an immediate validation test. My initial thoughts suggest that `F.cross_entropy` may offer a performance advantage over a manual InfoNCE implementation due to its optimized C++ kernel."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:05:24.884Z",
"text": "Optimizing InfoNCE Training\n\nI'm now investigating how `cross_entropy`'s fused C++ kernel could accelerate InfoNCE, potentially outperforming manual implementations. I'm exploring adding an `--objective` flag to `train_current_model.py` for easy switching. My current focus is on determining an optimal `tau` value by examining logit differences and standard two-tower retrieval temperature values, considering that my model's vectors are not unit-normalized."
},
{
"kind": "tool",
"timestamp": "2026-09-29T06:05:24.884Z",
"callId": "read_file__call_129618",
"name": "read_file",
"input": {
"start_line": 1120,
"end_line": 1200,
"file_path": "/app/solution/train_current_model.py"
},
"recorded": {
"id": "read_file__call_129618",
"name": "read_file",
"args": {
"start_line": 1120,
"end_line": 1200,
"file_path": "/app/solution/train_current_model.py"
},
"result": [
{
"functionResponse": {
"id": "read_file__call_129618",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1120-1200 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 1201.\n\n--- FILE CONTENT (truncated) ---\n negative_weights: torch.Tensor,\n sample_weights: np.ndarray,\n device: torch.device,\n backward: bool,\n) -> tuple[dict[str, float], torch.Tensor | None]:\n \"\"\"Compute one logical block's exact BCE in bounded-memory slices.\"\"\"\n target_count = len(positive_items)\n total_weight = float(sample_weights.sum())\n if total_weight <= 0:\n raise RuntimeError(\"target block has no positive sample weight\")\n sequence_gradient = torch.zeros_like(sequence_vectors) if backward else None\n statistics = {\n \"loss_sum\": 0.0,\n \"positive_sum\": 0.0,\n \"negative_sum\": 0.0,\n \"pair_correct\": 0.0,\n \"weight_sum\": 0.0,\n }\n for begin in range(0, target_count, config.loss_microbatch_size):\n end = min(begin + config.loss_microbatch_size, target_count)\n selected_states = flat_states[begin:end]\n if backward:\n sequence_input = (\n sequence_vectors[begin:end].detach().requires_grad_(True)\n )\n else:\n sequence_input = sequence_vectors[begin:end]\n liked_vectors, disliked_vectors, _ = pool_explicit_states(\n model, prepared, selected_states, device\n )\n counts = prepared.state_counts[selected_states]\n numeric = np.zeros((end - begin, 6), dtype=np.float32)\n numeric[:, :4] = flat_activity[begin:end]\n numeric[:, 4] = np.log1p(counts[:, 0]) / 6.0\n numeric[:, 5] = np.log1p(counts[:, 1]) / 4.0\n query_vectors = user_queries(\n model,\n sequence_input,\n liked_vectors,\n disliked_vectors,\n torch.from_numpy(numeric).to(device),\n )\n positive_vectors, negative_vectors = candidate_vectors(\n model,\n positive_items[begin:end],\n negatives[begin:end],\n )\n positive_logits = (query_vectors * positive_vectors).sum(dim=1)\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n positive_loss = F.softplus(-positive_logits)\n negative_loss = corrected_negative_loss(\n negative_logits, negative_weights[begin:end], config\n )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n loss_sum = (\n selected_weights * (positive_loss + negative_loss) / 2.0\n ).sum()\n if not bool(torch.isfinite(loss_sum)):\n raise RuntimeError(\"dense next-item training produced non-finite loss\")\n if backward:\n (loss_sum / total_weight).backward()\n if sequence_input.grad is None:\n raise RuntimeError(\"sequence microbatch did not produce a gradient\")\n sequence_gradient[begin:end] = sequence_input.grad\n statistics[\"loss_sum\"] += float(loss_sum.detach().cpu())\n statistics[\"positive_sum\"] += float(positive_logits.detach().sum().cpu())\n statistics[\"negative_sum\"] += float(negative_logits.detach().sum().cpu())\n statistics[\"pair_correct\"] += float(\n (positive_logits.detach()[:, None] > negative_logits.detach()).sum().cpu()\n )\n statistics[\"weight_sum\"] += float(selected_weights.sum().detach().cpu())\n return statistics, sequence_gradient\n\n\ndef release_accelerator_cache(device: torch.device) -> None:\n if device.type == \"mps\":\n torch.mps.empty_cache()"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T06:05:24.891Z",
"resultDisplay": "Read lines 1120-1200 of 2165 from solution/train_current_model.py",
"description": "solution/train_current_model.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T06:05:24.893Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1120-1200 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 1201.\n\n--- FILE CONTENT (truncated) ---\n negative_weights: torch.Tensor,\n sample_weights: np.ndarray,\n device: torch.device,\n backward: bool,\n) -> tuple[dict[str, float], torch.Tensor | None]:\n \"\"\"Compute one logical block's exact BCE in bounded-memory slices.\"\"\"\n target_count = len(positive_items)\n total_weight = float(sample_weights.sum())\n if total_weight <= 0:\n raise RuntimeError(\"target block has no positive sample weight\")\n sequence_gradient = torch.zeros_like(sequence_vectors) if backward else None\n statistics = {\n \"loss_sum\": 0.0,\n \"positive_sum\": 0.0,\n \"negative_sum\": 0.0,\n \"pair_correct\": 0.0,\n \"weight_sum\": 0.0,\n }\n for begin in range(0, target_count, config.loss_microbatch_size):\n end = min(begin + config.loss_microbatch_size, target_count)\n selected_states = flat_states[begin:end]\n if backward:\n sequence_input = (\n sequence_vectors[begin:end].detach().requires_grad_(True)\n )\n else:\n sequence_input = sequence_vectors[begin:end]\n liked_vectors, disliked_vectors, _ = pool_explicit_states(\n model, prepared, selected_states, device\n )\n counts = prepared.state_counts[selected_states]\n numeric = np.zeros((end - begin, 6), dtype=np.float32)\n numeric[:, :4] = flat_activity[begin:end]\n numeric[:, 4] = np.log1p(counts[:, 0]) / 6.0\n numeric[:, 5] = np.log1p(counts[:, 1]) / 4.0\n query_vectors = user_queries(\n model,\n sequence_input,\n liked_vectors,\n disliked_vectors,\n torch.from_numpy(numeric).to(device),\n )\n positive_vectors, negative_vectors = candidate_vectors(\n model,\n positive_items[begin:end],\n negatives[begin:end],\n )\n positive_logits = (query_vectors * positive_vectors).sum(dim=1)\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n positive_loss = F.softplus(-positive_logits)\n negative_loss = corrected_negative_loss(\n negative_logits, negative_weights[begin:end], config\n )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n loss_sum = (\n selected_weights * (positive_loss + negative_loss) / 2.0\n ).sum()\n if not bool(torch.isfinite(loss_sum)):\n raise RuntimeError(\"dense next-item training produced non-finite loss\")\n if backward:\n (loss_sum / total_weight).backward()\n if sequence_input.grad is None:\n raise RuntimeError(\"sequence microbatch did not produce a gradient\")\n sequence_gradient[begin:end] = sequence_input.grad\n statistics[\"loss_sum\"] += float(loss_sum.detach().cpu())\n statistics[\"positive_sum\"] += float(positive_logits.detach().sum().cpu())\n statistics[\"negative_sum\"] += float(negative_logits.detach().sum().cpu())\n statistics[\"pair_correct\"] += float(\n (positive_logits.detach()[:, None] > negative_logits.detach()).sum().cpu()\n )\n statistics[\"weight_sum\"] += float(selected_weights.sum().detach().cpu())\n return statistics, sequence_gradient\n\n\ndef release_accelerator_cache(device: torch.device) -> None:\n if device.type == \"mps\":\n torch.mps.empty_cache()",
"isError": false
},
"responseMessageId": "0e7375e5-ebbc-4c9d-902e-4ead9a435e59"
}
]
},
{
"id": 39,
"role": "agent",
"timestamp": "2026-09-29T06:05:28.416Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T06:05:28.416Z",
"callId": "read_file__call_49604",
"name": "read_file",
"input": {
"start_line": 35,
"end_line": 75,
"file_path": "/app/solution/train_current_model.py"
},
"recorded": {
"id": "read_file__call_49604",
"name": "read_file",
"args": {
"start_line": 35,
"end_line": 75,
"file_path": "/app/solution/train_current_model.py"
},
"result": [
{
"functionResponse": {
"id": "read_file__call_49604",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 35-75 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 76.\n\n--- FILE CONTENT (truncated) ---\n\n\nCACHE_VERSION = 2\n\n\n@dataclass(frozen=True)\nclass Config:\n bptt_steps: int = 100\n user_batch_size: int = 64\n embedding_dim: int = 192\n hidden_dim: int = 192\n user_hidden_1: int = 512\n user_hidden_2: int = 256\n item_hidden_1: int = 512\n item_hidden_2: int = 256\n dropout: float = 0.05\n epochs: int = 1\n learning_rate: float = 1.0e-3\n weight_decay: float = 1.0e-5\n gradient_clip: float = 2.0\n loss_microbatch_size: int = 65_536\n uniform_negatives: int = 24\n inbatch_negatives: int = 8\n inbatch_correction_power: float = 0.5\n inbatch_correction_min: float = 0.25\n inbatch_correction_max: float = 4.0\n negative_filter_scope: str = \"full-training-history\"\n target_weighting: str = \"organic-strong\"\n ordinary_target_weight: float = 0.25\n organic_strong_target_weight: float = 1.0\n organic_like_target_weight: float = 3.0\n discovery_horizon_days: int = 7\n validation_fraction: float = 0.10\n max_train_targets_per_user: int = 128\n max_validation_targets_per_user: int = 16\n early_stopping_patience: int = 1\n early_stopping_min_delta: float = 1.0e-4\n seed: int = 29\n\n\n@dataclass(frozen=True)"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T06:05:28.421Z",
"resultDisplay": "Read lines 35-75 of 2165 from solution/train_current_model.py",
"description": "solution/train_current_model.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T06:05:28.422Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 35-75 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 76.\n\n--- FILE CONTENT (truncated) ---\n\n\nCACHE_VERSION = 2\n\n\n@dataclass(frozen=True)\nclass Config:\n bptt_steps: int = 100\n user_batch_size: int = 64\n embedding_dim: int = 192\n hidden_dim: int = 192\n user_hidden_1: int = 512\n user_hidden_2: int = 256\n item_hidden_1: int = 512\n item_hidden_2: int = 256\n dropout: float = 0.05\n epochs: int = 1\n learning_rate: float = 1.0e-3\n weight_decay: float = 1.0e-5\n gradient_clip: float = 2.0\n loss_microbatch_size: int = 65_536\n uniform_negatives: int = 24\n inbatch_negatives: int = 8\n inbatch_correction_power: float = 0.5\n inbatch_correction_min: float = 0.25\n inbatch_correction_max: float = 4.0\n negative_filter_scope: str = \"full-training-history\"\n target_weighting: str = \"organic-strong\"\n ordinary_target_weight: float = 0.25\n organic_strong_target_weight: float = 1.0\n organic_like_target_weight: float = 3.0\n discovery_horizon_days: int = 7\n validation_fraction: float = 0.10\n max_train_targets_per_user: int = 128\n max_validation_targets_per_user: int = 16\n early_stopping_patience: int = 1\n early_stopping_min_delta: float = 1.0e-4\n seed: int = 29\n\n\n@dataclass(frozen=True)",
"isError": false
},
"responseMessageId": "aa2c9fba-5de9-4da1-8e4e-7293b5cf2b3a"
}
]
},
{
"id": 40,
"role": "agent",
"timestamp": "2026-09-29T06:05:38.857Z",
"blocks": [
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:05:38.857Z",
"text": "Refining Training Parameters\n\nI'm currently exploring additions to the `Config` for training, specifically introducing options for objectives like \"bce\" or \"infonce,\" alongside a temperature setting and parameters for catalog popularity negatives, including a power adjustment."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:05:38.857Z",
"text": "Analyzing Objective Logic\n\nI'm currently re-examining the interplay between different objectives and their associated loss calculations. My focus is on precisely tracing how the `score_target_microbatches` function handles the \"infonce\" objective and its reliance on temperature and negative sampling strategies."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:05:38.857Z",
"text": "Incorporating Log-Q Correction\n\nI'm digging into the specifics of the InfoNCE objective, realizing I need to account for the log-Q correction. I'm currently analyzing how to properly implement this, especially considering how sampling probabilities ($Q(j)$) differ for uniform versus popularity-based negatives and how that affects the logit calculation."
},
{
"kind": "tool",
"timestamp": "2026-09-29T06:05:38.857Z",
"callId": "read_file__call_54054",
"name": "read_file",
"input": {
"file_path": "/app/solution/train_current_model.py",
"end_line": 1990,
"start_line": 1930
},
"recorded": {
"id": "read_file__call_54054",
"name": "read_file",
"args": {
"file_path": "/app/solution/train_current_model.py",
"end_line": 1990,
"start_line": 1930
},
"result": [
{
"functionResponse": {
"id": "read_file__call_54054",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1930-1990 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 1991.\n\n--- FILE CONTENT (truncated) ---\n parser.add_argument(\"--device\", choices=(\"cpu\", \"mps\"), default=\"cpu\")\n parser.add_argument(\"--threads\", type=int, default=8)\n parser.add_argument(\"--bptt-steps\", type=int, default=Config.bptt_steps)\n parser.add_argument(\n \"--user-batch-size\", type=int, default=Config.user_batch_size\n )\n parser.add_argument(\"--epochs\", type=int, default=Config.epochs)\n parser.add_argument(\"--learning-rate\", type=float, default=Config.learning_rate)\n parser.add_argument(\n \"--loss-microbatch-size\",\n type=int,\n default=Config.loss_microbatch_size,\n )\n parser.add_argument(\n \"--negative-filter-scope\",\n choices=(\"full-training-history\", \"last-100\"),\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n )\n parser.add_argument(\n \"--ordinary-target-weight\",\n type=float,\n default=Config.ordinary_target_weight,\n )\n parser.add_argument(\n \"--organic-strong-target-weight\",\n type=float,\n default=Config.organic_strong_target_weight,\n )\n parser.add_argument(\n \"--organic-like-target-weight\",\n type=float,\n default=Config.organic_like_target_weight,\n )\n parser.add_argument(\n \"--discovery-horizon-days\",\n type=int,\n default=Config.discovery_horizon_days,\n )\n parser.add_argument(\n \"--validation-fraction\",\n type=float,\n default=Config.validation_fraction,\n )\n parser.add_argument(\n \"--max-train-targets-per-user\",\n type=int,\n default=Config.max_train_targets_per_user,\n help=\"retain this many recent pre-validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--max-validation-targets-per-user\",\n type=int,\n default=Config.max_validation_targets_per_user,\n help=\"retain this many final validation targets per user; 0 is uncapped\",\n )\n parser.add_argument("
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T06:05:38.863Z",
"resultDisplay": "Read lines 1930-1990 of 2165 from solution/train_current_model.py",
"description": "solution/train_current_model.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T06:05:38.865Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1930-1990 of 2165 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 1991.\n\n--- FILE CONTENT (truncated) ---\n parser.add_argument(\"--device\", choices=(\"cpu\", \"mps\"), default=\"cpu\")\n parser.add_argument(\"--threads\", type=int, default=8)\n parser.add_argument(\"--bptt-steps\", type=int, default=Config.bptt_steps)\n parser.add_argument(\n \"--user-batch-size\", type=int, default=Config.user_batch_size\n )\n parser.add_argument(\"--epochs\", type=int, default=Config.epochs)\n parser.add_argument(\"--learning-rate\", type=float, default=Config.learning_rate)\n parser.add_argument(\n \"--loss-microbatch-size\",\n type=int,\n default=Config.loss_microbatch_size,\n )\n parser.add_argument(\n \"--negative-filter-scope\",\n choices=(\"full-training-history\", \"last-100\"),\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n )\n parser.add_argument(\n \"--ordinary-target-weight\",\n type=float,\n default=Config.ordinary_target_weight,\n )\n parser.add_argument(\n \"--organic-strong-target-weight\",\n type=float,\n default=Config.organic_strong_target_weight,\n )\n parser.add_argument(\n \"--organic-like-target-weight\",\n type=float,\n default=Config.organic_like_target_weight,\n )\n parser.add_argument(\n \"--discovery-horizon-days\",\n type=int,\n default=Config.discovery_horizon_days,\n )\n parser.add_argument(\n \"--validation-fraction\",\n type=float,\n default=Config.validation_fraction,\n )\n parser.add_argument(\n \"--max-train-targets-per-user\",\n type=int,\n default=Config.max_train_targets_per_user,\n help=\"retain this many recent pre-validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--max-validation-targets-per-user\",\n type=int,\n default=Config.max_validation_targets_per_user,\n help=\"retain this many final validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(",
"isError": false
},
"responseMessageId": "f2b6c1d9-c9b1-4048-a48b-c9c10fc66f12"
}
]
},
{
"id": 41,
"role": "agent",
"timestamp": "2026-09-29T06:05:45.641Z",
"blocks": [
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:05:45.641Z",
"text": "Exploring Negative Sampling\n\nI'm currently investigating how to enhance the negative sampling strategy. My focus is on exploring alternative methods beyond the current uniform negative sampling and in-batch negatives, aiming to optimize its effectiveness."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:05:45.641Z",
"text": "Analyzing In-Batch Negatives\n\nI'm reviewing the logic for in-batch negatives. Specifically, I'm questioning why `unique_targets` is used for sampling and examining the implementation details in `/app/solution/train_current_model.py`, particularly the conditional masking."
},
{
"kind": "tool",
"timestamp": "2026-09-29T06:05:45.641Z",
"callId": "run_shell_command__call_91230",
"name": "run_shell_command",
"input": {
"command": "python3 -c \"\nimport time\nimport torch\n\nt0 = time.time()\nfor _ in range(10):\n p = torch.rand((6400, 3000))\n top = torch.topk(p, 8, dim=1)\nt1 = time.time()\nprint('Time for 10 topk(6400, 3000):', t1 - t0, 's (per call:', (t1 - t0)/10, 's)')\n\"",
"description": "Benchmark topk on 6400x3000"
},
"recorded": {
"id": "run_shell_command__call_91230",
"name": "run_shell_command",
"args": {
"command": "python3 -c \"\nimport time\nimport torch\n\nt0 = time.time()\nfor _ in range(10):\n p = torch.rand((6400, 3000))\n top = torch.topk(p, 8, dim=1)\nt1 = time.time()\nprint('Time for 10 topk(6400, 3000):', t1 - t0, 's (per call:', (t1 - t0)/10, 's)')\n\"",
"description": "Benchmark topk on 6400x3000"
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_91230",
"name": "run_shell_command",
"response": {
"output": "\nOutput: Time for 10 topk(6400, 3000): 1.11387300491333 s (per call: 0.11138730049133301 s)\nProcess Group PGID: 2060\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T06:05:49.653Z",
"resultDisplay": "Time for 10 topk(6400, 3000): 1.11387300491333 s (per call: 0.11138730049133301 s)",
"description": "Benchmark topk on 6400x3000",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T06:05:49.655Z",
"content": "\nOutput: Time for 10 topk(6400, 3000): 1.11387300491333 s (per call: 0.11138730049133301 s)\nProcess Group PGID: 2060\n",
"isError": false
},
"responseMessageId": "874ac637-d32a-4d4f-bdb3-30672820122b"
}
]
},
{
"id": 42,
"role": "agent",
"timestamp": "2026-09-29T06:06:03.546Z",
"blocks": [
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:06:03.546Z",
"text": "Optimizing Data Sampling\n\nI'm realizing our current random sampling is incredibly inefficient, consuming significant CPU time and losing crucial frequency data. My focus is shifting towards finding a more performant and accurate method to select our data points."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:06:03.546Z",
"text": "Refining Sampling Strategy\n\nMy exploration into sampling methods has yielded significant insights. I've discovered a way to drastically accelerate index generation and preserve true item frequency by sampling 8 random indices per batch, achieving a 1000x speedup. I'm now investigating how incorporating catalog popularity, with exponents like 0.5 or 0.75, can further enhance this approach, drawing from established recommendation system literature."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:06:03.546Z",
"text": "Exploring Negative Sampling\n\nI'm considering how to best leverage negative sampling. By combining uniform negatives for global coverage, popularity-based negatives to prevent over-biasing towards popular items, and in-batch negatives for local context, I believe we can significantly improve the model's learning. I'm now investigating the potential of incorporating user-disliked tracks into this strategy, examining existing data structures for feasibility."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:06:03.546Z",
"text": "Incorporating Disliked Items\n\nI'm realizing the immense potential of using explicitly disliked items as strong negatives. My immediate plan is to incrementally test this by first experimenting with InfoNCE loss at different temperatures ($\\tau = 0.1, 0.2, 0.5$) before integrating this new negative sampling strategy."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:06:03.546Z",
"text": "Testing Temperature Sensitivity\n\nMy current focus is on the InfoNCE loss implementation. I'm systematically evaluating how different temperature values, specifically 0.1, 0.2, and 0.5, impact the loss calculation, aiming to find the optimal setting before proceeding with the negative sampling integration."
},
{
"kind": "reasoning",
"representation": "summary",
"timestamp": "2026-09-29T06:06:03.546Z",
"text": "Evaluating Loss Objective\n\nI'm now focused on integrating the InfoNCE loss and understanding its temperature sensitivity. My immediate plan is to modify the training script to accept `--objective` and `--temperature` arguments, then run initial tests with `--target-weighting discovery-graded` and `--temperature 0.2` to assess the impact."
},
{
"kind": "tool",
"timestamp": "2026-09-29T06:06:03.546Z",
"callId": "replace__call_120894",
"name": "replace",
"input": {
"instruction": "Add objective and temperature parameters to Config",
"old_string": " negative_filter_scope: str = \"full-training-history\"\n target_weighting: str = \"organic-strong\"",
"file_path": "/app/solution/train_current_model.py",
"new_string": " negative_filter_scope: str = \"full-training-history\"\n objective: str = \"bce\"\n temperature: float = 0.2\n target_weighting: str = \"organic-strong\"",
"allow_multiple": false
},
"recorded": {
"id": "replace__call_120894",
"name": "replace",
"args": {
"instruction": "Add objective and temperature parameters to Config",
"old_string": " negative_filter_scope: str = \"full-training-history\"\n target_weighting: str = \"organic-strong\"",
"file_path": "/app/solution/train_current_model.py",
"new_string": " negative_filter_scope: str = \"full-training-history\"\n objective: str = \"bce\"\n temperature: float = 0.2\n target_weighting: str = \"organic-strong\"",
"allow_multiple": false
},
"result": [
{
"functionResponse": {
"id": "replace__call_120894",
"name": "replace",
"response": {
"error": "EACCES: permission denied, open '/app/solution/train_current_model.py'"
}
}
}
],
"status": "error",
"timestamp": "2026-09-29T06:06:03.553Z",
"resultDisplay": "Error writing file: EACCES: permission denied, open '/app/solution/train_current_model.py'",
"description": "Replaces text within a file. By default, the tool expects to find and replace exactly ONE occurrence of `old_string`. If you want to replace multiple occurrences of the exact same string, set `allow_multiple` to true. This tool requires providing significant context around the change to ensure precise targeting. Always use the read_file tool to examine the file's current content before attempting a text replacement.\n \n The user has the ability to modify the `new_string` content. If modified, this will be stated in the response.\n \n Expectation for required parameters:\n 1. `old_string` MUST be the exact literal text to replace (including all whitespace, indentation, newlines, and surrounding code etc.).\n 2. `new_string` MUST be the exact literal text to replace `old_string` with (also including all whitespace, indentation, newlines, and surrounding code etc.). Ensure the resulting code is correct and idiomatic and that `old_string` and `new_string` are different.\n 3. `instruction` is the detailed instruction of what needs to be changed. It is important to Make it specific and detailed so developers or large language models can understand what needs to be changed and perform the changes on their own if necessary. \n 4. NEVER escape `old_string` or `new_string`, that would break the exact literal text requirement.\n **Important:** If ANY of the above are not satisfied, the tool will fail. CRITICAL for `old_string`: Must uniquely identify the instance(s) to change. Include at least 3 lines of context BEFORE and AFTER the target text, matching whitespace and indentation precisely. If this string matches multiple locations and `allow_multiple` is not true, the tool will fail.\n 5. Prefer to break down complex and long changes into multiple smaller atomic calls to this tool. Always check the content of the file after changes or not finding a string to match.\n **Multiple replacements:** Set `allow_multiple` to true if you want to replace ALL occurrences that match `old_string` exactly.",
"displayName": "Edit",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T06:06:03.554Z",
"content": "EACCES: permission denied, open '/app/solution/train_current_model.py'",
"isError": true
},
"responseMessageId": "6891a1ef-03ae-4429-98ab-57cb9fbcf96f"
}
]
},
{
"id": 43,
"role": "agent",
"timestamp": "2026-09-29T06:06:39.409Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T06:06:39.409Z",
"callId": "run_shell_command__call_131355",
"name": "run_shell_command",
"input": {
"command": "ls -l /app/solution/",
"description": "Check permissions in /app/solution"
},
"recorded": {
"id": "run_shell_command__call_131355",
"name": "run_shell_command",
"args": {
"command": "ls -l /app/solution/",
"description": "Check permissions in /app/solution"
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_131355",
"name": "run_shell_command",
"response": {
"output": "\nOutput: total 1064960\n-rw-r--r-- 1 agent agent 552686232 Sep 29 05:47 prepared_sub.npz\n-rw-r--r-- 1 agent agent 535225808 Sep 29 05:45 prepared_val.npz\n-rw-r--r-- 1 agent agent 1244230 Sep 29 05:43 production_validation.parquet\n-r--r--r-- 1 agent agent 80058 Sep 29 05:46 train_current_model.py\n-r--r--r-- 1 agent agent 15985 Sep 29 05:46 two_tower_model.py\n-rw-r--r-- 1 agent agent 1249069 Sep 29 05:55 val_discovery_graded.parquet\nProcess Group PGID: 2079\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T06:06:39.426Z",
"resultDisplay": "total 1064960\n-rw-r--r-- 1 agent agent 552686232 Sep 29 05:47 prepared_sub.npz\n-rw-r--r-- 1 agent agent 535225808 Sep 29 05:45 prepared_val.npz\n-rw-r--r-- 1 agent agent 1244230 Sep 29 05:43 production_validation.parquet\n-r--r--r-- 1 agent agent 80058 Sep 29 05:46 train_current_model.py\n-r--r--r-- 1 agent agent 15985 Sep 29 05:46 two_tower_model.py\n-rw-r--r-- 1 agent agent 1249069 Sep 29 05:55 val_discovery_graded.parquet",
"description": "ls -l /app/solution/",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T06:06:39.428Z",
"content": "\nOutput: total 1064960\n-rw-r--r-- 1 agent agent 552686232 Sep 29 05:47 prepared_sub.npz\n-rw-r--r-- 1 agent agent 535225808 Sep 29 05:45 prepared_val.npz\n-rw-r--r-- 1 agent agent 1244230 Sep 29 05:43 production_validation.parquet\n-r--r--r-- 1 agent agent 80058 Sep 29 05:46 train_current_model.py\n-r--r--r-- 1 agent agent 15985 Sep 29 05:46 two_tower_model.py\n-rw-r--r-- 1 agent agent 1249069 Sep 29 05:55 val_discovery_graded.parquet\nProcess Group PGID: 2079\n",
"isError": false
},
"responseMessageId": "2b926727-67ab-4a3e-811b-339ac41f6878"
}
]
},
{
"id": 44,
"role": "agent",
"timestamp": "2026-09-29T06:06:43.315Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T06:06:43.315Z",
"callId": "run_shell_command__call_5977581",
"name": "run_shell_command",
"input": {
"command": "chmod u+w /app/solution/*.py",
"description": "Make copied python files writable in /app/solution"
},
"recorded": {
"id": "run_shell_command__call_5977581",
"name": "run_shell_command",
"args": {
"command": "chmod u+w /app/solution/*.py",
"description": "Make copied python files writable in /app/solution"
},
"result": [
{
"functionResponse": {
"id": "run_shell_command__call_5977581",
"name": "run_shell_command",
"response": {
"output": "\nOutput: (empty)\nProcess Group PGID: 2082\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T06:06:43.331Z",
"resultDisplay": "",
"description": "chmod u+w /app/solution/*.py",
"displayName": "Shell",
"renderOutputAsMarkdown": false
},
"result": {
"timestamp": "2026-09-29T06:06:43.333Z",
"content": "\nOutput: (empty)\nProcess Group PGID: 2082\n",
"isError": false
},
"responseMessageId": "e2922981-aa81-435e-930a-77adf44bd2be"
}
]
},
{
"id": 45,
"role": "agent",
"timestamp": "2026-09-29T06:06:46.031Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T06:06:46.031Z",
"callId": "replace__call_54862",
"name": "replace",
"input": {
"allow_multiple": false,
"old_string": " negative_filter_scope: str = \"full-training-history\"\n target_weighting: str = \"organic-strong\"",
"file_path": "/app/solution/train_current_model.py",
"instruction": "Add objective and temperature parameters to Config",
"new_string": " negative_filter_scope: str = \"full-training-history\"\n objective: str = \"bce\"\n temperature: float = 0.2\n target_weighting: str = \"organic-strong\""
},
"recorded": {
"id": "replace__call_54862",
"name": "replace",
"args": {
"allow_multiple": false,
"old_string": " negative_filter_scope: str = \"full-training-history\"\n target_weighting: str = \"organic-strong\"",
"file_path": "/app/solution/train_current_model.py",
"instruction": "Add objective and temperature parameters to Config",
"new_string": " negative_filter_scope: str = \"full-training-history\"\n objective: str = \"bce\"\n temperature: float = 0.2\n target_weighting: str = \"organic-strong\""
},
"result": [
{
"functionResponse": {
"id": "replace__call_54862",
"name": "replace",
"response": {
"output": "Successfully modified file: /app/solution/train_current_model.py (1 replacements). Here is the updated code:\n...\n inbatch_negatives: int = 8\n inbatch_correction_power: float = 0.5\n inbatch_correction_min: float = 0.25\n inbatch_correction_max: float = 4.0\n negative_filter_scope: str = \"full-training-history\"\n objective: str = \"bce\"\n temperature: float = 0.2\n target_weighting: str = \"organic-strong\"\n ordinary_target_weight: float = 0.25\n organic_strong_target_weight: float = 1.0\n organic_like_target_weight: float = 3.0\n discovery_horizon_days: int = 7\n..."
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T06:06:46.050Z",
"resultDisplay": {
"fileDiff": "Index: train_current_model.py\n===================================================================\n--- train_current_model.py\tCurrent\n+++ train_current_model.py\tProposed\n@@ -59,6 +59,8 @@\n inbatch_correction_min: float = 0.25\n inbatch_correction_max: float = 4.0\n negative_filter_scope: str = \"full-training-history\"\n+ objective: str = \"bce\"\n+ temperature: float = 0.2\n target_weighting: str = \"organic-strong\"\n ordinary_target_weight: float = 0.25\n organic_strong_target_weight: float = 1.0\n",
"fileName": "train_current_model.py",
"filePath": "/app/solution/train_current_model.py",
"originalContent": "#!/usr/bin/env python3\n\"\"\"Train the Yambda windowed two-tower baseline.\n\nEvery catalog listen with at least 50 percent completion is a sequence token.\nWithin each selected chronological window, every adjacent target contributes a\nloss. The fast default retains each user's most recent 128 training targets and\n16 held-out validation targets; passing zero for either cap restores the\nuncapped behavior for that side of the split. A 100-step truncated-BPTT block\ncomputes all of its timestep losses and carries the detached GRU hidden state\ninto the next block.\n\nThe model is a pure two-tower retriever: no user-ID embedding, no output\nnormalization, raw dot-product retrieval, sample-weighted binary cross entropy,\nand 24 uniform plus eight propensity-corrected unique in-batch negatives. Every\nselected adjacent target retains a nonzero weight. Only files under the supplied\npublic directory are read.\n\"\"\"\nfrom __future__ import annotations\n\nimport argparse\nimport json\nimport random\nimport sys\nimport time\nfrom dataclasses import asdict, dataclass\nfrom pathlib import Path\n\nimport numpy as np\nimport pandas as pd\nimport torch\nfrom torch import nn\nfrom torch.nn import functional as F\n\nimport two_tower_model as base\n\n\nCACHE_VERSION = 2\n\n\n@dataclass(frozen=True)\nclass Config:\n bptt_steps: int = 100\n user_batch_size: int = 64\n embedding_dim: int = 192\n hidden_dim: int = 192\n user_hidden_1: int = 512\n user_hidden_2: int = 256\n item_hidden_1: int = 512\n item_hidden_2: int = 256\n dropout: float = 0.05\n epochs: int = 1\n learning_rate: float = 1.0e-3\n weight_decay: float = 1.0e-5\n gradient_clip: float = 2.0\n loss_microbatch_size: int = 65_536\n uniform_negatives: int = 24\n inbatch_negatives: int = 8\n inbatch_correction_power: float = 0.5\n inbatch_correction_min: float = 0.25\n inbatch_correction_max: float = 4.0\n negative_filter_scope: str = \"full-training-history\"\n target_weighting: str = \"organic-strong\"\n ordinary_target_weight: float = 0.25\n organic_strong_target_weight: float = 1.0\n organic_like_target_weight: float = 3.0\n discovery_horizon_days: int = 7\n validation_fraction: float = 0.10\n max_train_targets_per_user: int = 128\n max_validation_targets_per_user: int = 16\n early_stopping_patience: int = 1\n early_stopping_min_delta: float = 1.0e-4\n seed: int = 29\n\n\n@dataclass(frozen=True)\nclass CandidateSet:\n uids: np.ndarray\n item_ids: np.ndarray\n packed: np.ndarray\n\n @classmethod\n def read(cls, path: Path) -> \"CandidateSet\":\n with np.load(path, allow_pickle=False) as artifact:\n result = cls(\n uids=artifact[\"uids\"].astype(np.int64, copy=True),\n item_ids=artifact[\"item_ids\"].astype(np.int64, copy=True),\n packed=artifact[\"packed_eligible\"].astype(np.uint8, copy=True),\n )\n expected = (len(result.uids), (len(result.item_ids) + 7) // 8)\n if result.packed.shape != expected:\n raise ValueError(\n f\"invalid candidate artifact shape {result.packed.shape}; \"\n f\"expected {expected}\"\n )\n return result\n\n\n@dataclass(frozen=True)\nclass SequenceData:\n item_ids: np.ndarray\n item_features: np.ndarray\n user_ids: np.ndarray\n offsets: np.ndarray\n sequence_timestamps: np.ndarray\n sequence_items: np.ndarray\n event_features: np.ndarray\n activity_features: np.ndarray\n state_ids: np.ndarray\n liked_states: np.ndarray\n disliked_states: np.ndarray\n state_counts: np.ndarray\n inference_user_rows: np.ndarray\n inference_liked_items: np.ndarray\n inference_disliked_items: np.ndarray\n inference_user_features: np.ndarray\n\n @property\n def lengths(self) -> np.ndarray:\n return np.diff(self.offsets)\n\n @property\n def target_count(self) -> int:\n return int(np.maximum(self.lengths - 1, 0).sum())\n\n\n@dataclass(frozen=True)\nclass TargetPartition:\n \"\"\"A contiguous recent target window and its chronological BCE split.\"\"\"\n\n window_starts: np.ndarray\n training_counts: np.ndarray\n validation_counts: np.ndarray\n history_training_counts: np.ndarray\n\n @property\n def selected_count(self) -> int:\n return int(self.training_counts.sum() + self.validation_counts.sum())\n\n\ndef log(message: str) -> None:\n print(message, file=sys.stderr, flush=True)\n\n\ndef seed_everything(seed: int) -> None:\n random.seed(seed)\n np.random.seed(seed)\n torch.manual_seed(seed)\n\n\ndef model_config(config: Config) -> base.TowerConfig:\n \"\"\"Map the dense trainer configuration onto the unchanged tower.\"\"\"\n return base.TowerConfig(\n sequence_length=config.bptt_steps,\n embedding_dim=config.embedding_dim,\n hidden_dim=config.hidden_dim,\n user_hidden_1=config.user_hidden_1,\n user_hidden_2=config.user_hidden_2,\n item_hidden_1=config.item_hidden_1,\n item_hidden_2=config.item_hidden_2,\n dropout=config.dropout,\n epochs=config.epochs,\n learning_rate=config.learning_rate,\n weight_decay=config.weight_decay,\n gradient_clip=config.gradient_clip,\n uniform_negatives=config.uniform_negatives,\n inbatch_negatives=config.inbatch_negatives,\n inbatch_correction_power=config.inbatch_correction_power,\n inbatch_correction_min=config.inbatch_correction_min,\n inbatch_correction_max=config.inbatch_correction_max,\n seed=config.seed,\n )\n\n\ndef append_state_snapshot(\n active_likes: dict[int, int],\n active_dislikes: dict[int, int],\n n_items: int,\n liked_states: list[np.ndarray],\n disliked_states: list[np.ndarray],\n state_counts: list[tuple[int, int]],\n) -> int:\n liked_states.append(base.pack_active_items(active_likes, base.LIKE_ITEMS, n_items))\n disliked_states.append(\n base.pack_active_items(active_dislikes, base.DISLIKE_ITEMS, n_items)\n )\n state_counts.append((len(active_likes), len(active_dislikes)))\n return len(liked_states) - 1\n\n\ndef causal_activity_features(\n timestamps: np.ndarray,\n items: np.ndarray,\n organic: np.ndarray,\n completion: np.ndarray,\n) -> np.ndarray:\n \"\"\"Compute the first four reviewed user features at every event cutoff.\"\"\"\n length = len(items)\n result = np.zeros((length, 4), dtype=np.float16)\n if not length:\n return result\n left_rows = np.searchsorted(\n timestamps,\n timestamps - base.PROFILE_LOOKBACK_DAYS * base.DAY,\n side=\"left\",\n )\n rows = np.arange(length, dtype=np.int64)\n counts = rows - left_rows + 1\n organic_sum = np.cumsum(organic.astype(np.float64))\n completion_sum = np.cumsum(np.minimum(completion, 100).astype(np.float64))\n before_organic = np.where(\n left_rows > 0, organic_sum[np.maximum(left_rows - 1, 0)], 0.0\n )\n before_completion = np.where(\n left_rows > 0, completion_sum[np.maximum(left_rows - 1, 0)], 0.0\n )\n\n unique_counts = np.empty(length, dtype=np.int32)\n frequencies: dict[int, int] = {}\n left = 0\n for right, raw_item in enumerate(items):\n item = int(raw_item)\n frequencies[item] = frequencies.get(item, 0) + 1\n required_left = int(left_rows[right])\n while left < required_left:\n old_item = int(items[left])\n remaining = frequencies[old_item] - 1\n if remaining:\n frequencies[old_item] = remaining\n else:\n del frequencies[old_item]\n left += 1\n unique_counts[right] = len(frequencies)\n\n result[:, 0] = (np.log1p(counts) / 8.0).astype(np.float16)\n result[:, 1] = (np.log1p(unique_counts) / 6.0).astype(np.float16)\n result[:, 2] = (\n (organic_sum - before_organic) / np.maximum(counts, 1)\n ).astype(np.float16)\n result[:, 3] = (\n (completion_sum - before_completion)\n / np.maximum(counts, 1)\n / 100.0\n ).astype(np.float16)\n return result\n\n\ndef build_sequence_parts(\n listens: base.ListenEvents,\n explicit: base.ExplicitEvents,\n n_items: int,\n liked_states: list[np.ndarray],\n disliked_states: list[np.ndarray],\n state_counts: list[tuple[int, int]],\n) -> tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray, np.ndarray] | None:\n qualifying = listens.completion_pct >= base.LISTEN_POSITIVE_THRESHOLD\n timestamps = listens.timestamps[qualifying].astype(np.int64, copy=False)\n items = listens.items[qualifying].astype(np.uint16, copy=False)\n organic = listens.organic[qualifying].astype(np.float32, copy=False)\n completion = listens.completion_pct[qualifying].astype(np.float32, copy=False)\n if not len(items):\n return None\n\n event_features = np.zeros((len(items), 3), dtype=np.float16)\n event_features[:, 0] = organic.astype(np.float16)\n event_features[:, 1] = (\n np.minimum(completion, 100.0) / 100.0\n ).astype(np.float16)\n if len(items) > 1:\n gaps = np.maximum(0, timestamps[1:] - timestamps[:-1])\n event_features[1:, 2] = np.exp(\n -gaps.astype(np.float64) / (60.0 * base.DAY)\n ).astype(np.float16)\n\n activity = causal_activity_features(\n timestamps, items, organic, completion\n )\n states = np.zeros(len(items), dtype=np.uint32)\n active_likes: dict[int, int] = {}\n active_dislikes: dict[int, int] = {}\n explicit_position = 0\n current_state = 0\n for row, timestamp in enumerate(timestamps):\n changed = False\n while (\n explicit_position < len(explicit.timestamps)\n and int(explicit.timestamps[explicit_position]) < int(timestamp)\n ):\n item = int(explicit.items[explicit_position])\n event_timestamp = int(explicit.timestamps[explicit_position])\n destination = (\n active_likes\n if int(explicit.channel[explicit_position]) == 0\n else active_dislikes\n )\n if int(explicit.active[explicit_position]) == 1:\n destination[item] = event_timestamp\n else:\n destination.pop(item, None)\n explicit_position += 1\n changed = True\n if changed:\n current_state = append_state_snapshot(\n active_likes,\n active_dislikes,\n n_items,\n liked_states,\n disliked_states,\n state_counts,\n )\n states[row] = np.uint32(current_state)\n return timestamps, items.copy(), event_features, activity, states\n\n\ndef final_user_features(\n timestamps: np.ndarray,\n items: np.ndarray,\n event_features: np.ndarray,\n history_end: int,\n like_count: int,\n dislike_count: int,\n) -> np.ndarray:\n start = int(\n np.searchsorted(\n timestamps,\n history_end - base.PROFILE_LOOKBACK_DAYS * base.DAY,\n side=\"left\",\n )\n )\n result = np.zeros(6, dtype=np.float32)\n recent_items = items[start:]\n if len(recent_items):\n result[0] = np.log1p(len(recent_items)) / 8.0\n result[1] = np.log1p(len(np.unique(recent_items))) / 6.0\n result[2] = float(np.mean(event_features[start:, 0]))\n result[3] = float(np.mean(event_features[start:, 1]))\n result[4] = np.log1p(like_count) / 6.0\n result[5] = np.log1p(dislike_count) / 4.0\n return result\n\n\ndef prepare_data(\n public: Path, stage: str, config: Config\n) -> tuple[SequenceData, CandidateSet]:\n data = public / \"data\"\n catalog = pd.read_parquet(data / \"candidate_catalog.parquet\").sort_values(\n \"popularity_rank\"\n )\n item_ids = catalog.item_id.to_numpy(dtype=np.int64, copy=True)\n item_index = pd.Index(item_ids)\n n_items = len(item_ids)\n if n_items >= np.iinfo(np.uint16).max:\n raise RuntimeError(\"uint16 item storage is too small for this catalog\")\n\n if stage == \"validation\":\n prefixes = (\"train\",)\n history_end = base.TRAIN_END\n candidate_name = \"validation_eligible_candidates.npz\"\n state_name = \"validation_preference_state.parquet\"\n else:\n prefixes = (\"train\", \"val\")\n history_end = base.VALIDATION_END\n candidate_name = \"test_eligible_candidates.npz\"\n state_name = \"test_preference_state.parquet\"\n eligibility = CandidateSet.read(data / candidate_name)\n if not np.array_equal(eligibility.item_ids, item_ids):\n raise RuntimeError(\"catalog and candidate artifact item order differ\")\n\n explicit_by_user = base.load_explicit_events(data, prefixes, item_index)\n inference_likes, inference_dislikes, like_counts, dislike_counts = (\n base.load_inference_state(\n data / state_name,\n item_index,\n eligibility.uids,\n n_items,\n )\n )\n inference_rows = {int(uid): row for row, uid in enumerate(eligibility.uids)}\n inference_user_features = np.zeros((len(eligibility.uids), 6), dtype=np.float32)\n inference_sequence_rows = np.full(len(eligibility.uids), -1, dtype=np.int32)\n\n # Snapshot zero is the shared empty explicit state.\n liked_states = [np.full(base.LIKE_ITEMS, n_items, dtype=np.uint16)]\n disliked_states = [\n np.full(base.DISLIKE_ITEMS, n_items, dtype=np.uint16)\n ]\n state_counts: list[tuple[int, int]] = [(0, 0)]\n user_ids: list[int] = []\n offsets = [0]\n timestamp_parts: list[np.ndarray] = []\n item_parts: list[np.ndarray] = []\n event_parts: list[np.ndarray] = []\n activity_parts: list[np.ndarray] = []\n state_parts: list[np.ndarray] = []\n\n listen_paths = [data / f\"{prefix}_listens.parquet\" for prefix in prefixes]\n started = time.monotonic()\n for seen_users, (uid, listens) in enumerate(\n base.merge_listen_streams(listen_paths, item_index), start=1\n ):\n parts = build_sequence_parts(\n listens,\n explicit_by_user.get(uid, base.empty_explicit_events()),\n n_items,\n liked_states,\n disliked_states,\n state_counts,\n )\n if parts is None:\n continue\n sequence_row = len(user_ids)\n user_ids.append(uid)\n timestamp_parts.append(parts[0].astype(np.uint32, copy=False))\n item_parts.append(parts[1])\n event_parts.append(parts[2])\n activity_parts.append(parts[3])\n state_parts.append(parts[4])\n offsets.append(offsets[-1] + len(parts[1]))\n inference_row = inference_rows.get(uid)\n if inference_row is not None:\n inference_sequence_rows[inference_row] = sequence_row\n inference_user_features[inference_row] = final_user_features(\n parts[0],\n parts[1],\n parts[2],\n history_end,\n int(like_counts[inference_row]),\n int(dislike_counts[inference_row]),\n )\n if seen_users % 1_000 == 0:\n log(\n f\"prepared users={seen_users:,} tokens={offsets[-1]:,} \"\n f\"states={len(liked_states):,} \"\n f\"elapsed={time.monotonic() - started:.1f}s\"\n )\n\n missing = np.flatnonzero(inference_sequence_rows < 0)\n if len(missing):\n inference_user_features[missing, 4] = np.log1p(like_counts[missing]) / 6.0\n inference_user_features[missing, 5] = (\n np.log1p(dislike_counts[missing]) / 4.0\n )\n log(f\"target users without qualifying catalog sequence={len(missing):,}\")\n\n prepared = SequenceData(\n item_ids=item_ids,\n item_features=base.build_item_features(catalog),\n user_ids=np.asarray(user_ids, dtype=np.int64),\n offsets=np.asarray(offsets, dtype=np.int64),\n sequence_timestamps=np.concatenate(timestamp_parts),\n sequence_items=np.concatenate(item_parts),\n event_features=np.concatenate(event_parts),\n activity_features=np.concatenate(activity_parts),\n state_ids=np.concatenate(state_parts),\n liked_states=np.stack(liked_states),\n disliked_states=np.stack(disliked_states),\n state_counts=np.asarray(state_counts, dtype=np.int32),\n inference_user_rows=inference_sequence_rows,\n inference_liked_items=inference_likes,\n inference_disliked_items=inference_dislikes,\n inference_user_features=inference_user_features,\n )\n log(\n f\"prepared stage={stage} sequence_users={len(prepared.user_ids):,} \"\n f\"tokens={len(prepared.sequence_items):,} \"\n f\"next_item_targets={prepared.target_count:,} \"\n f\"states={len(prepared.liked_states):,}\"\n )\n return prepared, eligibility\n\n\ndef cache_metadata(stage: str, config: Config) -> dict[str, object]:\n return {\n \"cache_version\": CACHE_VERSION,\n \"stage\": stage,\n \"positive_listen_threshold\": base.LISTEN_POSITIVE_THRESHOLD,\n \"profile_lookback_days\": base.PROFILE_LOOKBACK_DAYS,\n \"event_time_feature\": \"exp-negative-gap-over-60-days\",\n }\n\n\ndef save_cache(\n path: Path, prepared: SequenceData, stage: str, config: Config\n) -> None:\n path.parent.mkdir(parents=True, exist_ok=True)\n np.savez(\n path,\n metadata=np.asarray(json.dumps(cache_metadata(stage, config), sort_keys=True)),\n **prepared.__dict__,\n )\n log(f\"saved public-derived full-sequence cache: {path}\")\n\n\ndef load_cache(path: Path, stage: str, config: Config) -> SequenceData:\n with np.load(path, allow_pickle=False) as artifact:\n actual = json.loads(str(artifact[\"metadata\"].item()))\n expected = cache_metadata(stage, config)\n if actual != expected:\n raise RuntimeError(f\"sequence cache differs: {actual} != {expected}\")\n values = {\n field: artifact[field].copy()\n for field in SequenceData.__dataclass_fields__\n }\n prepared = SequenceData(**values)\n log(\n f\"loaded full-sequence cache: tokens={len(prepared.sequence_items):,} \"\n f\"targets={prepared.target_count:,}\"\n )\n return prepared\n\n\ndef load_or_prepare(\n public: Path,\n stage: str,\n config: Config,\n cache: Path | None,\n) -> tuple[SequenceData, CandidateSet]:\n candidate_name = (\n \"validation_eligible_candidates.npz\"\n if stage == \"validation\"\n else \"test_eligible_candidates.npz\"\n )\n eligibility = CandidateSet.read(public / \"data\" / candidate_name)\n if cache is not None and cache.is_file():\n prepared = load_cache(cache, stage, config)\n if not np.array_equal(prepared.item_ids, eligibility.item_ids):\n raise RuntimeError(\"cached catalog differs from candidate artifact\")\n return prepared, eligibility\n prepared, built_eligibility = prepare_data(public, stage, config)\n if cache is not None:\n save_cache(cache, prepared, stage, config)\n return prepared, built_eligibility\n\n\ndef build_item_correction(\n prepared: SequenceData,\n config: Config,\n target_starts: np.ndarray,\n target_counts: np.ndarray,\n) -> torch.Tensor:\n \"\"\"Estimate in-batch propensities from the selected training targets.\"\"\"\n counts = np.zeros(len(prepared.item_ids), dtype=np.float64)\n for user_row, count in enumerate(target_counts):\n if not count:\n continue\n offset = int(prepared.offsets[user_row])\n target_start = int(target_starts[user_row])\n counts += np.bincount(\n prepared.sequence_items[\n offset\n + target_start\n + 1 : offset\n + target_start\n + int(count)\n + 1\n ].astype(np.int64),\n minlength=len(prepared.item_ids),\n )\n total = float(counts.sum())\n probabilities = (counts + 1.0) / (total + len(counts))\n raw = np.power(\n (1.0 / len(counts)) / probabilities,\n config.inbatch_correction_power,\n )\n normalization = float(np.sum(counts * raw) / max(total, 1.0))\n correction = np.clip(\n raw / max(normalization, 1.0e-12),\n config.inbatch_correction_min,\n config.inbatch_correction_max,\n ).astype(np.float32)\n mean_on_targets = float(np.sum(counts * correction) / max(total, 1.0))\n log(\n f\"inbatch correction: min={correction.min():.3f} \"\n f\"mean_on_targets={mean_on_targets:.3f} max={correction.max():.3f}\"\n )\n return torch.from_numpy(correction)\n\n\ndef build_known_positive_filter(\n prepared: SequenceData, training_counts: np.ndarray\n) -> np.ndarray:\n \"\"\"Pack each user's complete public-training positive-item set.\"\"\"\n width = (len(prepared.item_ids) + 7) // 8\n packed = np.zeros((len(prepared.user_ids), width), dtype=np.uint8)\n retained = 0\n for user_row, target_count in enumerate(training_counts):\n if not target_count:\n continue\n offset = int(prepared.offsets[user_row])\n # T adjacent targets consume sequence positions 0 through T. The\n # chronological BCE suffix starts at T + 1 and is intentionally absent.\n items = np.unique(\n prepared.sequence_items[\n offset : offset + int(target_count) + 1\n ].astype(np.int64)\n )\n byte_columns = items >> 3\n bit_values = np.left_shift(\n np.uint8(1), (items & 7).astype(np.uint8)\n )\n np.bitwise_or.at(packed[user_row], byte_columns, bit_values)\n retained += len(items)\n log(\n f\"full-history false-negative filter: users={len(prepared.user_ids):,} \"\n f\"unique_user_items={retained:,} bytes={packed.nbytes:,}\"\n )\n return packed\n\n\ndef invalid_full_history_negatives(\n negatives: np.ndarray,\n targets: np.ndarray,\n user_rows: np.ndarray,\n known_positive_packed: np.ndarray,\n liked_items: np.ndarray,\n) -> np.ndarray:\n \"\"\"Return row-level invalid negatives using the complete training prefix.\"\"\"\n byte_columns = negatives >> 3\n bit_columns = negatives & 7\n known = known_positive_packed[user_rows[:, None], byte_columns]\n invalid = ((known >> bit_columns) & 1).astype(bool)\n invalid |= negatives == targets[:, None]\n invalid |= (negatives[:, :, None] == liked_items[:, None, :]).any(axis=2)\n for column in range(1, negatives.shape[1]):\n invalid[:, column] |= (\n negatives[:, column, None] == negatives[:, :column]\n ).any(axis=1)\n return invalid\n\n\ndef draw_full_history_filtered_negatives(\n targets: torch.Tensor,\n user_rows: np.ndarray,\n liked_items: torch.Tensor,\n known_positive_packed: np.ndarray,\n n_items: int,\n config: Config,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n \"\"\"Draw mixed negatives and reject all known public-training positives.\"\"\"\n targets_cpu = targets.detach().cpu().numpy().astype(np.int64, copy=False)\n liked_cpu = liked_items.detach().cpu().numpy().astype(np.int64, copy=False)\n batch_size = len(targets_cpu)\n uniform = torch.randint(\n n_items,\n (batch_size, config.uniform_negatives),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.empty((batch_size, 0), dtype=torch.long)\n corrected_count = 0\n if config.inbatch_negatives:\n unique_targets = torch.unique(targets.detach().cpu())\n corrected_count = min(\n config.inbatch_negatives, max(len(unique_targets) - 1, 0)\n )\n if corrected_count:\n priorities = torch.rand(\n (batch_size, len(unique_targets)), generator=generator\n )\n priorities.masked_fill_(\n unique_targets[None, :].eq(\n torch.from_numpy(targets_cpu)[:, None]\n ),\n -1.0,\n )\n columns = torch.topk(\n priorities, k=corrected_count, dim=1\n ).indices\n inbatch = unique_targets[columns]\n if corrected_count < config.inbatch_negatives:\n filler = torch.randint(\n n_items,\n (batch_size, config.inbatch_negatives - corrected_count),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.cat([inbatch, filler], dim=1)\n\n negatives = torch.cat([uniform, inbatch], dim=1).numpy()\n correction = np.ones_like(negatives, dtype=np.float32)\n if corrected_count:\n correction[\n :, config.uniform_negatives : config.uniform_negatives + corrected_count\n ] = item_correction[inbatch[:, :corrected_count]].numpy()\n for _ in range(32):\n invalid = invalid_full_history_negatives(\n negatives,\n targets_cpu,\n user_rows,\n known_positive_packed,\n liked_cpu,\n )\n if not np.any(invalid):\n return (\n torch.from_numpy(negatives).to(device),\n torch.from_numpy(correction).to(device),\n )\n replacements = torch.randint(\n n_items,\n (int(invalid.sum()),),\n generator=generator,\n device=\"cpu\",\n ).numpy()\n negatives[invalid] = replacements\n # Rejected in-batch items become ordinary uniform draws.\n correction[invalid] = 1.0\n invalid = invalid_full_history_negatives(\n negatives,\n targets_cpu,\n user_rows,\n known_positive_packed,\n liked_cpu,\n )\n if np.any(invalid):\n raise RuntimeError(\"full-history negative resampling failed\")\n return (\n torch.from_numpy(negatives).to(device),\n torch.from_numpy(correction).to(device),\n )\n\n\ndef draw_training_negatives(\n positive_items: torch.Tensor,\n known_history: torch.Tensor,\n liked_items: torch.Tensor,\n user_rows: np.ndarray,\n known_positive_packed: np.ndarray | None,\n prepared: SequenceData,\n config: Config,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n if config.negative_filter_scope == \"full-training-history\":\n if known_positive_packed is None:\n raise RuntimeError(\"full-history negative filter was not built\")\n return draw_full_history_filtered_negatives(\n positive_items,\n user_rows,\n liked_items,\n known_positive_packed,\n len(prepared.item_ids),\n config,\n item_correction,\n generator,\n device,\n )\n return base.draw_filtered_mixed_negatives(\n positive_items,\n known_history,\n liked_items,\n len(prepared.item_ids),\n config.uniform_negatives,\n config.inbatch_negatives,\n item_correction,\n generator,\n device,\n )\n\n\ndef event_sequence(\n model: base.SequentialTwoTower,\n items: torch.Tensor,\n features: torch.Tensor,\n hidden: torch.Tensor,\n) -> tuple[torch.Tensor, torch.Tensor]:\n inputs = torch.cat([model.context_embedding(items), features], dim=-1)\n projected = model.event_projection(inputs)\n return model.sequence_encoder(projected, hidden)\n\n\ndef pool_explicit_states(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n state_ids: np.ndarray,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor, torch.Tensor]:\n unique_states, inverse = np.unique(state_ids, return_inverse=True)\n liked_items = torch.from_numpy(\n prepared.liked_states[unique_states].astype(np.int64)\n ).to(device)\n disliked_items = torch.from_numpy(\n prepared.disliked_states[unique_states].astype(np.int64)\n ).to(device)\n liked_vectors = model._mean_context_channel(liked_items)\n disliked_vectors = model._mean_context_channel(disliked_items)\n inverse_rows = torch.from_numpy(inverse.astype(np.int64)).to(device)\n counts = prepared.state_counts[state_ids]\n return (\n liked_vectors[inverse_rows],\n disliked_vectors[inverse_rows],\n torch.from_numpy(counts.astype(np.float32)).to(device),\n )\n\n\ndef user_queries(\n model: base.SequentialTwoTower,\n sequence_vectors: torch.Tensor,\n liked_vectors: torch.Tensor,\n disliked_vectors: torch.Tensor,\n user_features: torch.Tensor,\n) -> torch.Tensor:\n return model.user_tower(\n torch.cat(\n [sequence_vectors, liked_vectors, disliked_vectors, user_features],\n dim=1,\n )\n )\n\n\ndef candidate_vectors(\n model: base.SequentialTwoTower,\n positives: torch.Tensor,\n negatives: torch.Tensor,\n) -> tuple[torch.Tensor, torch.Tensor]:\n all_items = torch.cat([positives, negatives.reshape(-1)])\n unique_items, inverse = torch.unique(all_items, return_inverse=True)\n encoded = model.encode_items(unique_items)\n positive_vectors = encoded[inverse[: len(positives)]]\n negative_vectors = encoded[inverse[len(positives) :]].reshape(\n len(positives), negatives.shape[1], -1\n )\n return positive_vectors, negative_vectors\n\n\ndef history_windows(\n sequence: np.ndarray,\n start: int,\n count: int,\n width: int,\n padding_item: int,\n) -> np.ndarray:\n base_start = max(0, start - width + 1)\n segment = sequence[base_start : start + count]\n padded = np.concatenate(\n [np.full(width - 1, padding_item, dtype=np.uint16), segment]\n )\n windows = np.lib.stride_tricks.sliding_window_view(padded, width)\n offset = start - base_start\n return windows[offset : offset + count].copy()\n\n\ndef corrected_negative_loss(\n logits: torch.Tensor,\n correction: torch.Tensor,\n config: Config,\n) -> torch.Tensor:\n raw = F.softplus(logits)\n total = config.uniform_negatives + config.inbatch_negatives\n result = torch.zeros(len(logits), device=logits.device)\n if config.uniform_negatives:\n result += config.uniform_negatives * raw[\n :, : config.uniform_negatives\n ].mean(dim=1)\n if config.inbatch_negatives:\n selected = slice(config.uniform_negatives, total)\n weights = correction[:, selected]\n result += config.inbatch_negatives * (\n (raw[:, selected] * weights).sum(dim=1)\n / weights.sum(dim=1).clamp_min(1.0e-6)\n )\n return result / total\n\n\ndef retained_target_window_starts(\n total_targets: np.ndarray,\n validation_counts: np.ndarray,\n training_counts: np.ndarray,\n) -> np.ndarray:\n \"\"\"Return the discarded prefix before the retained chronological suffix.\"\"\"\n training_region_ends = total_targets - validation_counts\n window_starts = training_region_ends - training_counts\n if (\n np.any(window_starts < 0)\n or np.any(training_counts < 0)\n or np.any(validation_counts < 0)\n ):\n raise RuntimeError(\"chronological target partition produced negative counts\")\n if not np.array_equal(\n window_starts + training_counts + validation_counts, total_targets\n ):\n raise RuntimeError(\"chronological target window is not a final contiguous suffix\")\n return window_starts\n\n\ndef chronological_target_partition(\n prepared: SequenceData,\n validation_fraction: float,\n max_train_targets_per_user: int,\n max_validation_targets_per_user: int,\n) -> TargetPartition:\n \"\"\"Select a recent contiguous window with a final out-of-time BCE suffix.\"\"\"\n totals = np.maximum(prepared.lengths - 1, 0).astype(np.int64)\n validation = np.floor(totals * validation_fraction).astype(np.int64)\n validation[(totals >= 2) & (validation == 0)] = 1\n if max_validation_targets_per_user:\n validation = np.minimum(validation, max_validation_targets_per_user)\n\n history_training = totals - validation\n training = history_training.copy()\n if max_train_targets_per_user:\n training = np.minimum(training, max_train_targets_per_user)\n window_starts = retained_target_window_starts(totals, validation, training)\n return TargetPartition(\n window_starts=window_starts,\n training_counts=training,\n validation_counts=validation,\n history_training_counts=history_training,\n )\n\n\ndef load_organic_like_outcomes(\n public: Path, stage: str, item_ids: np.ndarray\n) -> dict[int, dict[int, np.ndarray]]:\n \"\"\"Load mapped organic-like timestamps from public fitting history.\"\"\"\n prefixes = (\"train\",) if stage == \"validation\" else (\"train\", \"val\")\n frames = [\n pd.read_parquet(\n public / \"data\" / f\"{prefix}_likes.parquet\",\n columns=[\"uid\", \"timestamp\", \"item_id\", \"is_organic\"],\n )\n for prefix in prefixes\n ]\n likes = pd.concat(frames, ignore_index=True)\n likes = likes[likes.is_organic == 1].copy()\n mapped = pd.Index(item_ids).get_indexer(\n likes.item_id.to_numpy(dtype=np.int64)\n )\n likes = likes.loc[mapped >= 0, [\"uid\", \"timestamp\"]].copy()\n likes[\"item\"] = mapped[mapped >= 0].astype(np.int32, copy=False)\n likes = likes.sort_values([\"uid\", \"item\", \"timestamp\"], kind=\"stable\")\n nested: dict[int, dict[int, list[int]]] = {}\n for raw_uid, raw_timestamp, raw_item in likes.itertuples(\n index=False, name=None\n ):\n nested.setdefault(int(raw_uid), {}).setdefault(int(raw_item), []).append(\n int(raw_timestamp)\n )\n result = {\n uid: {\n item: np.asarray(timestamps, dtype=np.int64)\n for item, timestamps in items.items()\n }\n for uid, items in nested.items()\n }\n log(\n f\"loaded organic-like outcomes: rows={len(likes):,} \"\n f\"users={len(result):,}\"\n )\n return result\n\n\ndef build_target_weights(\n public: Path,\n stage: str,\n prepared: SequenceData,\n training_counts: np.ndarray,\n config: Config,\n) -> np.ndarray:\n \"\"\"Build nonzero per-token weights without dropping adjacent targets.\"\"\"\n if config.target_weighting == \"uniform\":\n weights = np.ones(len(prepared.sequence_items), dtype=np.float32)\n else:\n weights = np.full(\n len(prepared.sequence_items),\n config.ordinary_target_weight,\n dtype=np.float32,\n )\n organic_strong = (\n (prepared.event_features[:, 0] > 0.5)\n & (prepared.event_features[:, 1] >= 0.8)\n )\n weights[organic_strong] = config.organic_strong_target_weight\n if config.target_weighting == \"discovery-graded\":\n like_outcomes = load_organic_like_outcomes(\n public, stage, prepared.item_ids\n )\n history_end = (\n base.TRAIN_END if stage == \"validation\" else base.VALIDATION_END\n )\n horizon = config.discovery_horizon_days * base.DAY\n grade_three = 0\n for user_row, uid in enumerate(prepared.user_ids):\n start = int(prepared.offsets[user_row])\n end = int(prepared.offsets[user_row + 1])\n if end - start < 2:\n continue\n last_training_target = start + int(training_counts[user_row])\n first_validation_target = last_training_target + 1\n split_timestamp = (\n int(prepared.sequence_timestamps[first_validation_target])\n if first_validation_target < end\n else history_end\n )\n user_likes = like_outcomes.get(int(uid), {})\n seen = {int(prepared.sequence_items[start])}\n for position in range(start + 1, end):\n item = int(prepared.sequence_items[position])\n novel = item not in seen\n seen.add(item)\n if not novel:\n # Repeats remain training examples, but they cannot be\n # eligible discovery outcomes at the corresponding\n # cutoff and therefore keep only the ordinary weight.\n weights[position] = config.ordinary_target_weight\n continue\n timestamps = user_likes.get(item)\n if timestamps is None:\n continue\n timestamp = int(prepared.sequence_timestamps[position])\n upper = min(\n timestamp + horizon,\n split_timestamp\n if position <= last_training_target\n else history_end,\n )\n outcome = int(np.searchsorted(timestamps, timestamp, side=\"left\"))\n if outcome < len(timestamps) and int(timestamps[outcome]) < upper:\n weights[position] = config.organic_like_target_weight\n grade_three += 1\n log(\n f\"grade-3 first-encounter targets={grade_three:,} \"\n f\"horizon_days={config.discovery_horizon_days}\"\n )\n log(\n f\"target weighting={config.target_weighting} \"\n f\"min={weights.min():.3f} mean={weights.mean():.3f} \"\n f\"max={weights.max():.3f}\"\n )\n return weights\n\n\ndef build_dense_block(\n prepared: SequenceData,\n target_weights: np.ndarray,\n users: np.ndarray,\n window_starts: np.ndarray,\n process_counts: np.ndarray,\n score_starts: np.ndarray,\n start: int,\n steps: int,\n config: Config,\n) -> tuple[np.ndarray, ...]:\n \"\"\"Build one causal sequence block and mark the target rows to score.\"\"\"\n batch_size = len(users)\n padding_item = len(prepared.item_ids)\n input_items = np.full((batch_size, steps), padding_item, dtype=np.uint16)\n event_features = np.zeros((batch_size, steps, 3), dtype=np.float16)\n activity_features = np.zeros((batch_size, steps, 4), dtype=np.float16)\n state_ids = np.zeros((batch_size, steps), dtype=np.uint32)\n targets = np.zeros((batch_size, steps), dtype=np.uint16)\n weights = np.ones((batch_size, steps), dtype=np.float32)\n scored = np.zeros((batch_size, steps), dtype=bool)\n histories = np.full(\n (batch_size, steps, config.bptt_steps),\n padding_item,\n dtype=np.uint16,\n )\n processed = np.zeros(batch_size, dtype=np.int32)\n\n lengths = prepared.lengths\n for row, user_row in enumerate(users):\n count = min(steps, max(int(process_counts[row]) - start, 0))\n if not count:\n continue\n offset = int(prepared.offsets[user_row])\n window_start = int(window_starts[row])\n absolute_start = window_start + start\n positions = slice(\n offset + absolute_start,\n offset + absolute_start + count,\n )\n target_positions = slice(\n offset + absolute_start + 1,\n offset + absolute_start + count + 1,\n )\n input_items[row, :count] = prepared.sequence_items[positions]\n event_features[row, :count] = prepared.event_features[positions]\n activity_features[row, :count] = prepared.activity_features[positions]\n state_ids[row, :count] = prepared.state_ids[positions]\n targets[row, :count] = prepared.sequence_items[target_positions]\n weights[row, :count] = target_weights[target_positions]\n score_from = min(count, max(int(score_starts[row]) - start, 0))\n scored[row, score_from:count] = True\n processed[row] = count\n user_sequence = prepared.sequence_items[\n offset : offset + int(lengths[user_row])\n ]\n histories[row, :count] = history_windows(\n user_sequence,\n absolute_start,\n count,\n config.bptt_steps,\n padding_item,\n )\n return (\n input_items,\n event_features,\n activity_features,\n state_ids,\n targets,\n scored,\n histories,\n processed,\n weights,\n )\n\n\ndef score_target_microbatches(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n sequence_vectors: torch.Tensor,\n flat_states: np.ndarray,\n flat_activity: np.ndarray,\n positive_items: torch.Tensor,\n negatives: torch.Tensor,\n negative_weights: torch.Tensor,\n sample_weights: np.ndarray,\n device: torch.device,\n backward: bool,\n) -> tuple[dict[str, float], torch.Tensor | None]:\n \"\"\"Compute one logical block's exact BCE in bounded-memory slices.\"\"\"\n target_count = len(positive_items)\n total_weight = float(sample_weights.sum())\n if total_weight <= 0:\n raise RuntimeError(\"target block has no positive sample weight\")\n sequence_gradient = torch.zeros_like(sequence_vectors) if backward else None\n statistics = {\n \"loss_sum\": 0.0,\n \"positive_sum\": 0.0,\n \"negative_sum\": 0.0,\n \"pair_correct\": 0.0,\n \"weight_sum\": 0.0,\n }\n for begin in range(0, target_count, config.loss_microbatch_size):\n end = min(begin + config.loss_microbatch_size, target_count)\n selected_states = flat_states[begin:end]\n if backward:\n sequence_input = (\n sequence_vectors[begin:end].detach().requires_grad_(True)\n )\n else:\n sequence_input = sequence_vectors[begin:end]\n liked_vectors, disliked_vectors, _ = pool_explicit_states(\n model, prepared, selected_states, device\n )\n counts = prepared.state_counts[selected_states]\n numeric = np.zeros((end - begin, 6), dtype=np.float32)\n numeric[:, :4] = flat_activity[begin:end]\n numeric[:, 4] = np.log1p(counts[:, 0]) / 6.0\n numeric[:, 5] = np.log1p(counts[:, 1]) / 4.0\n query_vectors = user_queries(\n model,\n sequence_input,\n liked_vectors,\n disliked_vectors,\n torch.from_numpy(numeric).to(device),\n )\n positive_vectors, negative_vectors = candidate_vectors(\n model,\n positive_items[begin:end],\n negatives[begin:end],\n )\n positive_logits = (query_vectors * positive_vectors).sum(dim=1)\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n positive_loss = F.softplus(-positive_logits)\n negative_loss = corrected_negative_loss(\n negative_logits, negative_weights[begin:end], config\n )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n loss_sum = (\n selected_weights * (positive_loss + negative_loss) / 2.0\n ).sum()\n if not bool(torch.isfinite(loss_sum)):\n raise RuntimeError(\"dense next-item training produced non-finite loss\")\n if backward:\n (loss_sum / total_weight).backward()\n if sequence_input.grad is None:\n raise RuntimeError(\"sequence microbatch did not produce a gradient\")\n sequence_gradient[begin:end] = sequence_input.grad\n statistics[\"loss_sum\"] += float(loss_sum.detach().cpu())\n statistics[\"positive_sum\"] += float(positive_logits.detach().sum().cpu())\n statistics[\"negative_sum\"] += float(negative_logits.detach().sum().cpu())\n statistics[\"pair_correct\"] += float(\n (positive_logits.detach()[:, None] > negative_logits.detach()).sum().cpu()\n )\n statistics[\"weight_sum\"] += float(selected_weights.sum().detach().cpu())\n return statistics, sequence_gradient\n\n\ndef release_accelerator_cache(device: torch.device) -> None:\n if device.type == \"mps\":\n torch.mps.empty_cache()\n\n\ndef run_training_epoch(\n model: base.SequentialTwoTower,\n optimizer: torch.optim.Optimizer,\n prepared: SequenceData,\n config: Config,\n target_weights: np.ndarray,\n window_starts: np.ndarray,\n target_counts: np.ndarray,\n item_correction: torch.Tensor,\n known_positive_packed: np.ndarray | None,\n generator: torch.Generator,\n device: torch.device,\n epoch: int,\n) -> dict[str, float]:\n model.train()\n train_users = np.flatnonzero(target_counts > 0)\n sorted_users = train_users[\n np.argsort(target_counts[train_users], kind=\"stable\")\n ]\n groups = [\n sorted_users[start : start + config.user_batch_size]\n for start in range(0, len(sorted_users), config.user_batch_size)\n ]\n group_order = torch.randperm(len(groups), generator=generator).numpy()\n negative_count = config.uniform_negatives + config.inbatch_negatives\n total_loss = 0.0\n total_targets = 0\n total_weight = 0.0\n total_positive = 0.0\n total_negative = 0.0\n total_pair_correct = 0.0\n batches = 0\n started = time.monotonic()\n\n for group_number in group_order:\n users = groups[int(group_number)]\n group_counts = target_counts[users]\n hidden = torch.zeros(\n 1, len(users), config.embedding_dim, device=device\n )\n max_targets = int(group_counts.max())\n for start in range(0, max_targets, config.bptt_steps):\n steps = min(config.bptt_steps, max_targets - start)\n block = build_dense_block(\n prepared,\n target_weights,\n users,\n window_starts[users],\n group_counts,\n np.zeros(len(users), dtype=np.int64),\n start,\n steps,\n config,\n )\n sequence_outputs, _ = event_sequence(\n model,\n torch.from_numpy(block[0].astype(np.int64)).to(device),\n torch.from_numpy(block[1].astype(np.float32)).to(device),\n hidden,\n )\n hidden = torch.stack(\n [\n sequence_outputs[row, count - 1]\n if count\n else hidden[0, row]\n for row, count in enumerate(block[7])\n ]\n ).unsqueeze(0).detach()\n flat_scored = block[5].reshape(-1)\n if not np.any(flat_scored):\n continue\n mask = torch.from_numpy(flat_scored).to(device)\n sequence_vectors = sequence_outputs.reshape(\n -1, config.embedding_dim\n )[mask]\n flat_states = block[3].reshape(-1)[flat_scored]\n positive_items = torch.from_numpy(\n block[4].reshape(-1)[flat_scored].astype(np.int64)\n ).to(device)\n known_history = torch.from_numpy(\n block[6].reshape(-1, config.bptt_steps)[flat_scored].astype(\n np.int64\n )\n ).to(device)\n liked_items = torch.from_numpy(\n prepared.liked_states[flat_states].astype(np.int64)\n ).to(device)\n flat_user_rows = np.broadcast_to(\n users[:, None], block[5].shape\n ).reshape(-1)[flat_scored]\n negatives, negative_weights = draw_training_negatives(\n positive_items,\n known_history,\n liked_items,\n flat_user_rows,\n known_positive_packed,\n prepared,\n config,\n item_correction,\n generator,\n device,\n )\n optimizer.zero_grad(set_to_none=True)\n statistics, sequence_gradient = score_target_microbatches(\n model,\n prepared,\n config,\n sequence_vectors,\n flat_states,\n block[2].reshape(-1, 4)[flat_scored].astype(np.float32),\n positive_items,\n negatives,\n negative_weights,\n block[8].reshape(-1)[flat_scored],\n device,\n backward=True,\n )\n if sequence_gradient is None:\n raise RuntimeError(\"training block omitted its sequence gradient\")\n sequence_vectors.backward(sequence_gradient)\n nn.utils.clip_grad_norm_(model.parameters(), config.gradient_clip)\n optimizer.step()\n\n count = len(positive_items)\n total_loss += statistics[\"loss_sum\"]\n total_targets += count\n total_weight += statistics[\"weight_sum\"]\n total_positive += statistics[\"positive_sum\"]\n total_negative += statistics[\"negative_sum\"]\n total_pair_correct += statistics[\"pair_correct\"]\n batches += 1\n if batches % 25 == 0:\n release_accelerator_cache(device)\n if batches % 250 == 0:\n batch_bce = statistics[\"loss_sum\"] / max(\n statistics[\"weight_sum\"], 1.0e-12\n )\n running_bce = total_loss / max(total_weight, 1.0e-12)\n log(\n f\"epoch={epoch}/{config.epochs} batches={batches:,} \"\n f\"targets={total_targets:,} \"\n f\"batch_bce={batch_bce:.6f} \"\n f\"running_bce={running_bce:.6f} \"\n f\"elapsed={time.monotonic() - started:.1f}s\"\n )\n\n return {\n \"epoch\": float(epoch),\n \"loss\": total_loss / max(total_weight, 1.0e-12),\n \"positive_logit\": total_positive / max(total_targets, 1),\n \"negative_logit\": total_negative / max(total_targets * negative_count, 1),\n \"pair_accuracy\": total_pair_correct\n / max(total_targets * negative_count, 1),\n \"targets\": float(total_targets),\n \"target_weight_sum\": total_weight,\n \"batches\": float(batches),\n \"seconds\": time.monotonic() - started,\n }\n\n\n@torch.no_grad()\ndef evaluate_validation_bce(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n target_weights: np.ndarray,\n window_starts: np.ndarray,\n training_counts: np.ndarray,\n validation_counts: np.ndarray,\n item_correction: torch.Tensor,\n known_positive_packed: np.ndarray | None,\n device: torch.device,\n) -> dict[str, float]:\n \"\"\"Measure deterministic BCE on each user's chronological target suffix.\"\"\"\n model.eval()\n totals = training_counts + validation_counts\n users_with_validation = np.flatnonzero(validation_counts > 0)\n ordered = users_with_validation[\n np.argsort(totals[users_with_validation], kind=\"stable\")\n ]\n groups = [\n ordered[start : start + config.user_batch_size]\n for start in range(0, len(ordered), config.user_batch_size)\n ]\n generator = torch.Generator(device=\"cpu\").manual_seed(config.seed + 20_000)\n total_loss = 0.0\n total_targets = 0\n total_weight = 0.0\n started = time.monotonic()\n\n for group_number, users in enumerate(groups, start=1):\n group_totals = totals[users]\n group_starts = training_counts[users]\n hidden = torch.zeros(\n 1, len(users), config.embedding_dim, device=device\n )\n max_targets = int(group_totals.max())\n for start in range(0, max_targets, config.bptt_steps):\n steps = min(config.bptt_steps, max_targets - start)\n block = build_dense_block(\n prepared,\n target_weights,\n users,\n window_starts[users],\n group_totals,\n group_starts,\n start,\n steps,\n config,\n )\n sequence_outputs, _ = event_sequence(\n model,\n torch.from_numpy(block[0].astype(np.int64)).to(device),\n torch.from_numpy(block[1].astype(np.float32)).to(device),\n hidden,\n )\n hidden = torch.stack(\n [\n sequence_outputs[row, count - 1]\n if count\n else hidden[0, row]\n for row, count in enumerate(block[7])\n ]\n ).unsqueeze(0)\n flat_scored = block[5].reshape(-1)\n if not np.any(flat_scored):\n continue\n mask = torch.from_numpy(flat_scored).to(device)\n sequence_vectors = sequence_outputs.reshape(\n -1, config.embedding_dim\n )[mask]\n flat_states = block[3].reshape(-1)[flat_scored]\n positive_items = torch.from_numpy(\n block[4].reshape(-1)[flat_scored].astype(np.int64)\n ).to(device)\n known_history = torch.from_numpy(\n block[6].reshape(-1, config.bptt_steps)[flat_scored].astype(\n np.int64\n )\n ).to(device)\n liked_items = torch.from_numpy(\n prepared.liked_states[flat_states].astype(np.int64)\n ).to(device)\n flat_user_rows = np.broadcast_to(\n users[:, None], block[5].shape\n ).reshape(-1)[flat_scored]\n negatives, negative_weights = draw_training_negatives(\n positive_items,\n known_history,\n liked_items,\n flat_user_rows,\n known_positive_packed,\n prepared,\n config,\n item_correction,\n generator,\n device,\n )\n statistics, _ = score_target_microbatches(\n model,\n prepared,\n config,\n sequence_vectors,\n flat_states,\n block[2].reshape(-1, 4)[flat_scored].astype(np.float32),\n positive_items,\n negatives,\n negative_weights,\n block[8].reshape(-1)[flat_scored],\n device,\n backward=False,\n )\n total_loss += statistics[\"loss_sum\"]\n total_targets += len(positive_items)\n total_weight += statistics[\"weight_sum\"]\n if group_number % 25 == 0:\n release_accelerator_cache(device)\n return {\n \"validation_loss\": total_loss / max(total_weight, 1.0e-12),\n \"validation_targets\": float(total_targets),\n \"validation_target_weight_sum\": total_weight,\n \"validation_seconds\": time.monotonic() - started,\n }\n\n\ndef snapshot_model(model: nn.Module) -> dict[str, torch.Tensor]:\n return {\n name: value.detach().cpu().clone()\n for name, value in model.state_dict().items()\n }\n\n\ndef save_epoch_checkpoint(\n checkpoint_dir: Path,\n epoch: int,\n model_state: dict[str, torch.Tensor],\n config: Config,\n diagnostics: dict[str, float],\n optimizer: torch.optim.Optimizer,\n generator: torch.Generator,\n) -> Path:\n checkpoint_dir.mkdir(parents=True, exist_ok=True)\n path = checkpoint_dir / f\"epoch-{epoch:02d}.pt\"\n torch.save(\n {\n \"epoch\": epoch,\n \"config\": asdict(config),\n \"diagnostics\": diagnostics,\n \"model_state\": model_state,\n \"optimizer_state\": optimizer.state_dict(),\n \"generator_state\": generator.get_state(),\n },\n path,\n )\n log(f\"saved epoch checkpoint: {path}\")\n return path\n\n\ndef train_model(\n public: Path,\n stage: str,\n prepared: SequenceData,\n config: Config,\n device: torch.device,\n checkpoint_dir: Path | None,\n resume_checkpoint: Path | None,\n) -> tuple[base.SequentialTwoTower, list[dict[str, float]]]:\n partition = chronological_target_partition(\n prepared,\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n training_counts = partition.training_counts\n validation_counts = partition.validation_counts\n log(\n f\"chronological BCE window: train={int(training_counts.sum()):,} \"\n f\"validation={int(validation_counts.sum()):,} \"\n f\"selected={partition.selected_count:,} \"\n f\"discarded_old_targets={int(partition.window_starts.sum()):,} \"\n f\"all={prepared.target_count:,}\"\n )\n model = base.SequentialTwoTower(prepared.item_features, model_config(config)).to(\n device\n )\n optimizer = torch.optim.AdamW(\n model.parameters(), lr=config.learning_rate, weight_decay=config.weight_decay\n )\n generator = torch.Generator(device=\"cpu\").manual_seed(config.seed + 10_000)\n item_correction = build_item_correction(\n prepared,\n config,\n partition.window_starts,\n training_counts,\n )\n target_weights = build_target_weights(\n public,\n stage,\n prepared,\n partition.history_training_counts,\n config,\n )\n known_positive_packed = (\n build_known_positive_filter(prepared, partition.history_training_counts)\n if config.negative_filter_scope == \"full-training-history\"\n else None\n )\n history: list[dict[str, float]] = []\n best_loss = float(\"inf\")\n best_epoch = 0\n best_state: dict[str, torch.Tensor] | None = None\n epochs_without_improvement = 0\n start_epoch = 0\n\n if resume_checkpoint is not None:\n payload = torch.load(\n resume_checkpoint, map_location=device, weights_only=False\n )\n stored_config = payload.get(\"config\", {})\n architecture_fields = (\n \"bptt_steps\",\n \"embedding_dim\",\n \"hidden_dim\",\n \"user_hidden_1\",\n \"user_hidden_2\",\n \"item_hidden_1\",\n \"item_hidden_2\",\n )\n mismatches = {\n name: (stored_config.get(name), getattr(config, name))\n for name in architecture_fields\n if stored_config.get(name) != getattr(config, name)\n }\n if mismatches:\n raise RuntimeError(\n f\"resume checkpoint architecture differs: {mismatches}\"\n )\n stored_filter_scope = stored_config.get(\n \"negative_filter_scope\", \"last-100\"\n )\n stored_target_weighting = stored_config.get(\n \"target_weighting\", \"uniform\"\n )\n stored_weight_values = (\n stored_config.get(\"ordinary_target_weight\", 1.0),\n stored_config.get(\"organic_strong_target_weight\", 1.0),\n stored_config.get(\"organic_like_target_weight\", 3.0),\n stored_config.get(\"discovery_horizon_days\", 7),\n )\n current_weight_values = (\n config.ordinary_target_weight,\n config.organic_strong_target_weight,\n config.organic_like_target_weight,\n config.discovery_horizon_days,\n )\n stored_window_values = (\n stored_config.get(\"validation_fraction\", 0.10),\n stored_config.get(\"max_train_targets_per_user\", 0),\n stored_config.get(\"max_validation_targets_per_user\", 0),\n )\n current_window_values = (\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n objective_changed = (\n stored_filter_scope != config.negative_filter_scope\n or stored_target_weighting != config.target_weighting\n or stored_weight_values != current_weight_values\n or stored_window_values != current_window_values\n )\n model.load_state_dict(payload[\"model_state\"])\n if \"optimizer_state\" in payload and not objective_changed:\n optimizer.load_state_dict(payload[\"optimizer_state\"])\n for group in optimizer.param_groups:\n group[\"lr\"] = config.learning_rate\n group[\"weight_decay\"] = config.weight_decay\n log(f\"restored optimizer state from {resume_checkpoint}\")\n else:\n log(\n f\"using a fresh AdamW optimizer after loading {resume_checkpoint}\"\n )\n if \"generator_state\" in payload and not objective_changed:\n generator.set_state(payload[\"generator_state\"])\n start_epoch = int(payload[\"epoch\"])\n if start_epoch >= config.epochs:\n raise RuntimeError(\n f\"resume epoch {start_epoch} must be below --epochs {config.epochs}\"\n )\n prior = {\n str(name): float(value)\n for name, value in payload.get(\"diagnostics\", {}).items()\n }\n if objective_changed:\n prior.pop(\"validation_loss\", None)\n log(\n \"BCE objective changed: filter \"\n f\"{stored_filter_scope}->{config.negative_filter_scope}, \"\n \"target weighting \"\n f\"{stored_target_weighting}->{config.target_weighting}, \"\n f\"target window {stored_window_values}->{current_window_values}; \"\n \"recomputing the resume checkpoint's validation BCE\"\n )\n if \"validation_loss\" not in prior:\n prior.update(\n evaluate_validation_bce(\n model,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n validation_counts,\n item_correction,\n known_positive_packed,\n device,\n )\n )\n history.append(prior)\n best_loss = float(prior[\"validation_loss\"])\n best_epoch = start_epoch\n best_state = snapshot_model(model)\n log(\n f\"resumed validation-BCE training: epoch={start_epoch} \"\n f\"loss={best_loss:.6f}\"\n )\n\n for epoch in range(start_epoch + 1, config.epochs + 1):\n diagnostics = run_training_epoch(\n model,\n optimizer,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n item_correction,\n known_positive_packed,\n generator,\n device,\n epoch,\n )\n validation = evaluate_validation_bce(\n model,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n validation_counts,\n item_correction,\n known_positive_packed,\n device,\n )\n diagnostics.update(validation)\n history.append(diagnostics)\n current_state = snapshot_model(model)\n if checkpoint_dir is not None:\n save_epoch_checkpoint(\n checkpoint_dir,\n epoch,\n current_state,\n config,\n diagnostics,\n optimizer,\n generator,\n )\n improved = (\n diagnostics[\"validation_loss\"]\n < best_loss - config.early_stopping_min_delta\n )\n if improved:\n best_loss = diagnostics[\"validation_loss\"]\n best_epoch = epoch\n best_state = current_state\n epochs_without_improvement = 0\n if checkpoint_dir is not None:\n torch.save(\n {\n \"epoch\": epoch,\n \"config\": asdict(config),\n \"diagnostics\": diagnostics,\n \"model_state\": current_state,\n \"optimizer_state\": optimizer.state_dict(),\n \"generator_state\": generator.get_state(),\n },\n checkpoint_dir / \"best.pt\",\n )\n else:\n epochs_without_improvement += 1\n log(\n f\"epoch={epoch}/{config.epochs} train_loss={diagnostics['loss']:.5f} \"\n f\"validation_loss={diagnostics['validation_loss']:.5f} \"\n f\"pair_accuracy={diagnostics['pair_accuracy']:.3f} \"\n f\"train_targets={int(diagnostics['targets']):,} \"\n f\"validation_targets={int(diagnostics['validation_targets']):,} \"\n f\"train_seconds={diagnostics['seconds']:.1f} \"\n f\"validation_seconds={diagnostics['validation_seconds']:.1f}\"\n )\n release_accelerator_cache(device)\n if epochs_without_improvement >= config.early_stopping_patience:\n log(\n f\"early stopping after epoch={epoch}; \"\n f\"best_epoch={best_epoch} best_validation_loss={best_loss:.6f}\"\n )\n break\n\n if best_state is None:\n raise RuntimeError(\"training did not produce a valid checkpoint\")\n model.load_state_dict(best_state)\n model.to(device).eval()\n log(\n f\"restored best validation-BCE checkpoint: epoch={best_epoch} \"\n f\"loss={best_loss:.6f}\"\n )\n return model, history\n\n\n@torch.inference_mode()\ndef inference_queries(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n device: torch.device,\n) -> torch.Tensor:\n sequence_rows = prepared.inference_user_rows\n lengths = np.where(\n sequence_rows >= 0, prepared.lengths[np.maximum(sequence_rows, 0)], 0\n )\n ordered = np.argsort(lengths, kind=\"stable\")\n result = torch.zeros(\n len(sequence_rows), config.embedding_dim, dtype=torch.float32\n )\n for group_start in range(0, len(ordered), config.user_batch_size):\n output_rows = ordered[group_start : group_start + config.user_batch_size]\n rows = sequence_rows[output_rows]\n group_lengths = lengths[output_rows]\n hidden = torch.zeros(\n 1, len(output_rows), config.embedding_dim, device=device\n )\n for start in range(0, int(group_lengths.max(initial=0)), config.bptt_steps):\n steps = min(\n config.bptt_steps,\n int(group_lengths.max(initial=0)) - start,\n )\n items = np.full(\n (len(output_rows), steps),\n len(prepared.item_ids),\n dtype=np.uint16,\n )\n features = np.zeros((len(output_rows), steps, 3), dtype=np.float16)\n counts = np.zeros(len(output_rows), dtype=np.int32)\n for local_row, sequence_row in enumerate(rows):\n if sequence_row < 0:\n continue\n count = min(steps, max(int(group_lengths[local_row]) - start, 0))\n if not count:\n continue\n offset = int(prepared.offsets[sequence_row])\n selected = slice(offset + start, offset + start + count)\n items[local_row, :count] = prepared.sequence_items[selected]\n features[local_row, :count] = prepared.event_features[selected]\n counts[local_row] = count\n outputs, _ = event_sequence(\n model,\n torch.from_numpy(items.astype(np.int64)).to(device),\n torch.from_numpy(features.astype(np.float32)).to(device),\n hidden,\n )\n next_rows: list[torch.Tensor] = []\n for local_row, count in enumerate(counts):\n next_rows.append(\n outputs[local_row, count - 1]\n if count\n else hidden[0, local_row]\n )\n hidden = torch.stack(next_rows).unsqueeze(0)\n\n liked_items = torch.from_numpy(\n prepared.inference_liked_items[output_rows].astype(np.int64)\n ).to(device)\n disliked_items = torch.from_numpy(\n prepared.inference_disliked_items[output_rows].astype(np.int64)\n ).to(device)\n queries = user_queries(\n model,\n hidden.squeeze(0),\n model._mean_context_channel(liked_items),\n model._mean_context_channel(disliked_items),\n torch.from_numpy(\n prepared.inference_user_features[output_rows].astype(np.float32)\n ).to(device),\n )\n result[torch.from_numpy(output_rows.astype(np.int64))] = queries.cpu()\n return result\n\n\n@torch.inference_mode()\ndef rank_candidates(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n eligibility: CandidateSet,\n config: Config,\n device: torch.device,\n ranking_batch_size: int,\n) -> tuple[pd.DataFrame, dict[str, float]]:\n model.eval()\n queries = inference_queries(model, prepared, config, device)\n item_parts: list[torch.Tensor] = []\n for start in range(0, len(prepared.item_ids), 2_048):\n rows = torch.arange(\n start,\n min(start + 2_048, len(prepared.item_ids)),\n dtype=torch.long,\n device=device,\n )\n item_parts.append(model.encode_items(rows).cpu())\n items = torch.cat(item_parts)\n rows: list[tuple[int, int, int]] = []\n for start in range(0, len(eligibility.uids), ranking_batch_size):\n end = min(start + ranking_batch_size, len(eligibility.uids))\n scores = queries[start:end] @ items.T\n eligible = np.unpackbits(\n eligibility.packed[start:end],\n axis=1,\n count=len(prepared.item_ids),\n bitorder=\"little\",\n ).astype(bool, copy=False)\n scores.masked_fill_(~torch.from_numpy(eligible), -torch.inf)\n selected = torch.topk(scores, k=base.TOP_N, dim=1).indices.numpy()\n for local_row, columns in enumerate(selected):\n uid = int(eligibility.uids[start + local_row])\n rows.extend(\n (uid, int(prepared.item_ids[column]), rank)\n for rank, column in enumerate(columns, start=1)\n )\n if end == len(eligibility.uids) or end % 1_024 == 0:\n log(f\"ranked users={end:,}/{len(eligibility.uids):,}\")\n query_norm = torch.linalg.vector_norm(queries, dim=1).numpy()\n item_norm = torch.linalg.vector_norm(items, dim=1).numpy()\n diagnostics = {\n \"query_norm_mean\": float(query_norm.mean()),\n \"query_norm_std\": float(query_norm.std()),\n \"item_norm_mean\": float(item_norm.mean()),\n \"item_norm_std\": float(item_norm.std()),\n }\n return pd.DataFrame(rows, columns=[\"uid\", \"item_id\", \"rank\"]), diagnostics\n\n\ndef write_ranking(frame: pd.DataFrame, output: Path) -> None:\n output.parent.mkdir(parents=True, exist_ok=True)\n if output.suffix == \".parquet\":\n frame.to_parquet(output, index=False, compression=\"zstd\")\n elif output.suffix == \".csv\":\n frame.to_csv(output, index=False)\n else:\n raise ValueError(\"--output must end in .parquet or .csv\")\n\n\ndef main() -> int:\n parser = argparse.ArgumentParser(description=__doc__)\n parser.add_argument(\n \"--stage\", choices=(\"validation\", \"submission\"), default=\"validation\"\n )\n parser.add_argument(\"--output\", type=Path, required=True)\n parser.add_argument(\"--public-dir\", type=Path, required=True)\n parser.add_argument(\"--prepared-cache\", type=Path)\n parser.add_argument(\"--device\", choices=(\"cpu\", \"mps\"), default=\"cpu\")\n parser.add_argument(\"--threads\", type=int, default=8)\n parser.add_argument(\"--bptt-steps\", type=int, default=Config.bptt_steps)\n parser.add_argument(\n \"--user-batch-size\", type=int, default=Config.user_batch_size\n )\n parser.add_argument(\"--epochs\", type=int, default=Config.epochs)\n parser.add_argument(\"--learning-rate\", type=float, default=Config.learning_rate)\n parser.add_argument(\n \"--loss-microbatch-size\",\n type=int,\n default=Config.loss_microbatch_size,\n )\n parser.add_argument(\n \"--negative-filter-scope\",\n choices=(\"full-training-history\", \"last-100\"),\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n )\n parser.add_argument(\n \"--ordinary-target-weight\",\n type=float,\n default=Config.ordinary_target_weight,\n )\n parser.add_argument(\n \"--organic-strong-target-weight\",\n type=float,\n default=Config.organic_strong_target_weight,\n )\n parser.add_argument(\n \"--organic-like-target-weight\",\n type=float,\n default=Config.organic_like_target_weight,\n )\n parser.add_argument(\n \"--discovery-horizon-days\",\n type=int,\n default=Config.discovery_horizon_days,\n )\n parser.add_argument(\n \"--validation-fraction\",\n type=float,\n default=Config.validation_fraction,\n )\n parser.add_argument(\n \"--max-train-targets-per-user\",\n type=int,\n default=Config.max_train_targets_per_user,\n help=\"retain this many recent pre-validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--max-validation-targets-per-user\",\n type=int,\n default=Config.max_validation_targets_per_user,\n help=\"retain this many final validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--early-stopping-patience\",\n type=int,\n default=Config.early_stopping_patience,\n )\n parser.add_argument(\n \"--early-stopping-min-delta\",\n type=float,\n default=Config.early_stopping_min_delta,\n )\n parser.add_argument(\"--checkpoint-dir\", type=Path)\n parser.add_argument(\"--resume-checkpoint\", type=Path)\n parser.add_argument(\"--seed\", type=int, default=Config.seed)\n parser.add_argument(\"--ranking-batch-size\", type=int, default=64)\n parser.add_argument(\"--prepare-only\", action=\"store_true\")\n args = parser.parse_args()\n\n if args.threads < 1 or args.bptt_steps < 2 or args.user_batch_size < 1:\n parser.error(\"threads, BPTT steps, and user batch size must be positive\")\n if (\n args.epochs < 1\n or args.learning_rate <= 0\n or args.loss_microbatch_size < 1\n or args.early_stopping_patience < 1\n or args.early_stopping_min_delta < 0\n or args.ordinary_target_weight <= 0\n or args.organic_strong_target_weight <= 0\n or args.organic_like_target_weight <= 0\n or args.discovery_horizon_days < 1\n or args.max_train_targets_per_user < 0\n or args.max_validation_targets_per_user < 0\n ):\n parser.error(\"training counts, learning rate, and patience are invalid\")\n if not 0.0 < args.validation_fraction < 1.0:\n parser.error(\"validation fraction must be strictly between zero and one\")\n public = args.public_dir.resolve()\n if not (public / \"data\" / \"candidate_catalog.parquet\").is_file():\n parser.error(f\"public task data not found under {public}\")\n config = Config(\n bptt_steps=args.bptt_steps,\n user_batch_size=args.user_batch_size,\n epochs=args.epochs,\n learning_rate=args.learning_rate,\n loss_microbatch_size=args.loss_microbatch_size,\n negative_filter_scope=args.negative_filter_scope,\n target_weighting=args.target_weighting,\n ordinary_target_weight=args.ordinary_target_weight,\n organic_strong_target_weight=args.organic_strong_target_weight,\n organic_like_target_weight=args.organic_like_target_weight,\n discovery_horizon_days=args.discovery_horizon_days,\n validation_fraction=args.validation_fraction,\n max_train_targets_per_user=args.max_train_targets_per_user,\n max_validation_targets_per_user=args.max_validation_targets_per_user,\n early_stopping_patience=args.early_stopping_patience,\n early_stopping_min_delta=args.early_stopping_min_delta,\n seed=args.seed,\n )\n torch.set_num_threads(args.threads)\n torch.set_num_interop_threads(1)\n seed_everything(config.seed)\n device = torch.device(args.device)\n if device.type == \"mps\" and not torch.backends.mps.is_available():\n parser.error(\"MPS was requested but is unavailable\")\n\n started = time.monotonic()\n prepared, eligibility = load_or_prepare(\n public,\n args.stage,\n config,\n args.prepared_cache.resolve() if args.prepared_cache else None,\n )\n target_partition = chronological_target_partition(\n prepared,\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n if args.prepare_only:\n print(\n json.dumps(\n {\n \"stage\": args.stage,\n \"sequence_users\": len(prepared.user_ids),\n \"sequence_tokens\": len(prepared.sequence_items),\n \"next_item_targets\": prepared.target_count,\n \"selected_training_targets\": int(\n target_partition.training_counts.sum()\n ),\n \"selected_validation_targets\": int(\n target_partition.validation_counts.sum()\n ),\n \"discarded_old_targets\": int(\n target_partition.window_starts.sum()\n ),\n \"target_users\": len(eligibility.uids),\n \"runtime_seconds\": round(time.monotonic() - started, 3),\n \"config\": asdict(config),\n },\n indent=2,\n )\n )\n return 0\n\n model, training_history = train_model(\n public,\n args.stage,\n prepared,\n config,\n device,\n args.checkpoint_dir.resolve() if args.checkpoint_dir else None,\n args.resume_checkpoint.resolve() if args.resume_checkpoint else None,\n )\n ranking, embedding_diagnostics = rank_candidates(\n model,\n prepared,\n eligibility,\n config,\n device,\n args.ranking_batch_size,\n )\n write_ranking(ranking, args.output.resolve())\n print(\n json.dumps(\n {\n \"baseline\": \"yambda-windowed-sequence-two-tower-base\",\n \"stage\": args.stage,\n \"output\": str(args.output.resolve()),\n \"rows\": len(ranking),\n \"users\": int(ranking.uid.nunique()),\n \"runtime_seconds\": round(time.monotonic() - started, 3),\n \"public_inputs_only\": True,\n \"architecture\": {\n \"family\": \"pure-windowed-sequence-two-tower\",\n \"sequence_training\": (\n \"every-adjacent-target-within-recent-chronological-window\"\n ),\n \"sequence_encoder\": \"one-layer-GRU-dense-timestep-loss\",\n \"bptt_hidden_state\": \"carried-and-detached\",\n \"model_selection\": \"chronological-validation-BCE-early-stopping\",\n \"loss_execution\": \"gradient-accumulated-microbatches\",\n \"user_id_embedding\": False,\n \"user_mlp\": \"582-512-256-192\",\n \"item_mlp\": \"196-512-256-192\",\n \"normalization\": \"none\",\n \"score\": \"raw-dot-product\",\n \"objective\": \"sample-weighted-BCE-at-every-timestep\",\n \"negative_sampling\": (\n \"24-uniform-plus-8-propensity-corrected-inbatch\"\n ),\n \"false_negative_filter\": config.negative_filter_scope,\n \"target_weighting\": config.target_weighting,\n },\n \"config\": asdict(config),\n \"sequence_tokens\": len(prepared.sequence_items),\n \"all_next_item_targets\": prepared.target_count,\n \"selected_training_targets\": int(\n target_partition.training_counts.sum()\n ),\n \"selected_validation_targets\": int(\n target_partition.validation_counts.sum()\n ),\n \"discarded_old_targets\": int(target_partition.window_starts.sum()),\n \"training_history\": training_history,\n \"embedding_diagnostics\": embedding_diagnostics,\n \"torch_version\": torch.__version__,\n \"device\": str(device),\n },\n indent=2,\n )\n )\n return 0\n\n\nif __name__ == \"__main__\":\n raise SystemExit(main())\n",
"newContent": "#!/usr/bin/env python3\n\"\"\"Train the Yambda windowed two-tower baseline.\n\nEvery catalog listen with at least 50 percent completion is a sequence token.\nWithin each selected chronological window, every adjacent target contributes a\nloss. The fast default retains each user's most recent 128 training targets and\n16 held-out validation targets; passing zero for either cap restores the\nuncapped behavior for that side of the split. A 100-step truncated-BPTT block\ncomputes all of its timestep losses and carries the detached GRU hidden state\ninto the next block.\n\nThe model is a pure two-tower retriever: no user-ID embedding, no output\nnormalization, raw dot-product retrieval, sample-weighted binary cross entropy,\nand 24 uniform plus eight propensity-corrected unique in-batch negatives. Every\nselected adjacent target retains a nonzero weight. Only files under the supplied\npublic directory are read.\n\"\"\"\nfrom __future__ import annotations\n\nimport argparse\nimport json\nimport random\nimport sys\nimport time\nfrom dataclasses import asdict, dataclass\nfrom pathlib import Path\n\nimport numpy as np\nimport pandas as pd\nimport torch\nfrom torch import nn\nfrom torch.nn import functional as F\n\nimport two_tower_model as base\n\n\nCACHE_VERSION = 2\n\n\n@dataclass(frozen=True)\nclass Config:\n bptt_steps: int = 100\n user_batch_size: int = 64\n embedding_dim: int = 192\n hidden_dim: int = 192\n user_hidden_1: int = 512\n user_hidden_2: int = 256\n item_hidden_1: int = 512\n item_hidden_2: int = 256\n dropout: float = 0.05\n epochs: int = 1\n learning_rate: float = 1.0e-3\n weight_decay: float = 1.0e-5\n gradient_clip: float = 2.0\n loss_microbatch_size: int = 65_536\n uniform_negatives: int = 24\n inbatch_negatives: int = 8\n inbatch_correction_power: float = 0.5\n inbatch_correction_min: float = 0.25\n inbatch_correction_max: float = 4.0\n negative_filter_scope: str = \"full-training-history\"\n objective: str = \"bce\"\n temperature: float = 0.2\n target_weighting: str = \"organic-strong\"\n ordinary_target_weight: float = 0.25\n organic_strong_target_weight: float = 1.0\n organic_like_target_weight: float = 3.0\n discovery_horizon_days: int = 7\n validation_fraction: float = 0.10\n max_train_targets_per_user: int = 128\n max_validation_targets_per_user: int = 16\n early_stopping_patience: int = 1\n early_stopping_min_delta: float = 1.0e-4\n seed: int = 29\n\n\n@dataclass(frozen=True)\nclass CandidateSet:\n uids: np.ndarray\n item_ids: np.ndarray\n packed: np.ndarray\n\n @classmethod\n def read(cls, path: Path) -> \"CandidateSet\":\n with np.load(path, allow_pickle=False) as artifact:\n result = cls(\n uids=artifact[\"uids\"].astype(np.int64, copy=True),\n item_ids=artifact[\"item_ids\"].astype(np.int64, copy=True),\n packed=artifact[\"packed_eligible\"].astype(np.uint8, copy=True),\n )\n expected = (len(result.uids), (len(result.item_ids) + 7) // 8)\n if result.packed.shape != expected:\n raise ValueError(\n f\"invalid candidate artifact shape {result.packed.shape}; \"\n f\"expected {expected}\"\n )\n return result\n\n\n@dataclass(frozen=True)\nclass SequenceData:\n item_ids: np.ndarray\n item_features: np.ndarray\n user_ids: np.ndarray\n offsets: np.ndarray\n sequence_timestamps: np.ndarray\n sequence_items: np.ndarray\n event_features: np.ndarray\n activity_features: np.ndarray\n state_ids: np.ndarray\n liked_states: np.ndarray\n disliked_states: np.ndarray\n state_counts: np.ndarray\n inference_user_rows: np.ndarray\n inference_liked_items: np.ndarray\n inference_disliked_items: np.ndarray\n inference_user_features: np.ndarray\n\n @property\n def lengths(self) -> np.ndarray:\n return np.diff(self.offsets)\n\n @property\n def target_count(self) -> int:\n return int(np.maximum(self.lengths - 1, 0).sum())\n\n\n@dataclass(frozen=True)\nclass TargetPartition:\n \"\"\"A contiguous recent target window and its chronological BCE split.\"\"\"\n\n window_starts: np.ndarray\n training_counts: np.ndarray\n validation_counts: np.ndarray\n history_training_counts: np.ndarray\n\n @property\n def selected_count(self) -> int:\n return int(self.training_counts.sum() + self.validation_counts.sum())\n\n\ndef log(message: str) -> None:\n print(message, file=sys.stderr, flush=True)\n\n\ndef seed_everything(seed: int) -> None:\n random.seed(seed)\n np.random.seed(seed)\n torch.manual_seed(seed)\n\n\ndef model_config(config: Config) -> base.TowerConfig:\n \"\"\"Map the dense trainer configuration onto the unchanged tower.\"\"\"\n return base.TowerConfig(\n sequence_length=config.bptt_steps,\n embedding_dim=config.embedding_dim,\n hidden_dim=config.hidden_dim,\n user_hidden_1=config.user_hidden_1,\n user_hidden_2=config.user_hidden_2,\n item_hidden_1=config.item_hidden_1,\n item_hidden_2=config.item_hidden_2,\n dropout=config.dropout,\n epochs=config.epochs,\n learning_rate=config.learning_rate,\n weight_decay=config.weight_decay,\n gradient_clip=config.gradient_clip,\n uniform_negatives=config.uniform_negatives,\n inbatch_negatives=config.inbatch_negatives,\n inbatch_correction_power=config.inbatch_correction_power,\n inbatch_correction_min=config.inbatch_correction_min,\n inbatch_correction_max=config.inbatch_correction_max,\n seed=config.seed,\n )\n\n\ndef append_state_snapshot(\n active_likes: dict[int, int],\n active_dislikes: dict[int, int],\n n_items: int,\n liked_states: list[np.ndarray],\n disliked_states: list[np.ndarray],\n state_counts: list[tuple[int, int]],\n) -> int:\n liked_states.append(base.pack_active_items(active_likes, base.LIKE_ITEMS, n_items))\n disliked_states.append(\n base.pack_active_items(active_dislikes, base.DISLIKE_ITEMS, n_items)\n )\n state_counts.append((len(active_likes), len(active_dislikes)))\n return len(liked_states) - 1\n\n\ndef causal_activity_features(\n timestamps: np.ndarray,\n items: np.ndarray,\n organic: np.ndarray,\n completion: np.ndarray,\n) -> np.ndarray:\n \"\"\"Compute the first four reviewed user features at every event cutoff.\"\"\"\n length = len(items)\n result = np.zeros((length, 4), dtype=np.float16)\n if not length:\n return result\n left_rows = np.searchsorted(\n timestamps,\n timestamps - base.PROFILE_LOOKBACK_DAYS * base.DAY,\n side=\"left\",\n )\n rows = np.arange(length, dtype=np.int64)\n counts = rows - left_rows + 1\n organic_sum = np.cumsum(organic.astype(np.float64))\n completion_sum = np.cumsum(np.minimum(completion, 100).astype(np.float64))\n before_organic = np.where(\n left_rows > 0, organic_sum[np.maximum(left_rows - 1, 0)], 0.0\n )\n before_completion = np.where(\n left_rows > 0, completion_sum[np.maximum(left_rows - 1, 0)], 0.0\n )\n\n unique_counts = np.empty(length, dtype=np.int32)\n frequencies: dict[int, int] = {}\n left = 0\n for right, raw_item in enumerate(items):\n item = int(raw_item)\n frequencies[item] = frequencies.get(item, 0) + 1\n required_left = int(left_rows[right])\n while left < required_left:\n old_item = int(items[left])\n remaining = frequencies[old_item] - 1\n if remaining:\n frequencies[old_item] = remaining\n else:\n del frequencies[old_item]\n left += 1\n unique_counts[right] = len(frequencies)\n\n result[:, 0] = (np.log1p(counts) / 8.0).astype(np.float16)\n result[:, 1] = (np.log1p(unique_counts) / 6.0).astype(np.float16)\n result[:, 2] = (\n (organic_sum - before_organic) / np.maximum(counts, 1)\n ).astype(np.float16)\n result[:, 3] = (\n (completion_sum - before_completion)\n / np.maximum(counts, 1)\n / 100.0\n ).astype(np.float16)\n return result\n\n\ndef build_sequence_parts(\n listens: base.ListenEvents,\n explicit: base.ExplicitEvents,\n n_items: int,\n liked_states: list[np.ndarray],\n disliked_states: list[np.ndarray],\n state_counts: list[tuple[int, int]],\n) -> tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray, np.ndarray] | None:\n qualifying = listens.completion_pct >= base.LISTEN_POSITIVE_THRESHOLD\n timestamps = listens.timestamps[qualifying].astype(np.int64, copy=False)\n items = listens.items[qualifying].astype(np.uint16, copy=False)\n organic = listens.organic[qualifying].astype(np.float32, copy=False)\n completion = listens.completion_pct[qualifying].astype(np.float32, copy=False)\n if not len(items):\n return None\n\n event_features = np.zeros((len(items), 3), dtype=np.float16)\n event_features[:, 0] = organic.astype(np.float16)\n event_features[:, 1] = (\n np.minimum(completion, 100.0) / 100.0\n ).astype(np.float16)\n if len(items) > 1:\n gaps = np.maximum(0, timestamps[1:] - timestamps[:-1])\n event_features[1:, 2] = np.exp(\n -gaps.astype(np.float64) / (60.0 * base.DAY)\n ).astype(np.float16)\n\n activity = causal_activity_features(\n timestamps, items, organic, completion\n )\n states = np.zeros(len(items), dtype=np.uint32)\n active_likes: dict[int, int] = {}\n active_dislikes: dict[int, int] = {}\n explicit_position = 0\n current_state = 0\n for row, timestamp in enumerate(timestamps):\n changed = False\n while (\n explicit_position < len(explicit.timestamps)\n and int(explicit.timestamps[explicit_position]) < int(timestamp)\n ):\n item = int(explicit.items[explicit_position])\n event_timestamp = int(explicit.timestamps[explicit_position])\n destination = (\n active_likes\n if int(explicit.channel[explicit_position]) == 0\n else active_dislikes\n )\n if int(explicit.active[explicit_position]) == 1:\n destination[item] = event_timestamp\n else:\n destination.pop(item, None)\n explicit_position += 1\n changed = True\n if changed:\n current_state = append_state_snapshot(\n active_likes,\n active_dislikes,\n n_items,\n liked_states,\n disliked_states,\n state_counts,\n )\n states[row] = np.uint32(current_state)\n return timestamps, items.copy(), event_features, activity, states\n\n\ndef final_user_features(\n timestamps: np.ndarray,\n items: np.ndarray,\n event_features: np.ndarray,\n history_end: int,\n like_count: int,\n dislike_count: int,\n) -> np.ndarray:\n start = int(\n np.searchsorted(\n timestamps,\n history_end - base.PROFILE_LOOKBACK_DAYS * base.DAY,\n side=\"left\",\n )\n )\n result = np.zeros(6, dtype=np.float32)\n recent_items = items[start:]\n if len(recent_items):\n result[0] = np.log1p(len(recent_items)) / 8.0\n result[1] = np.log1p(len(np.unique(recent_items))) / 6.0\n result[2] = float(np.mean(event_features[start:, 0]))\n result[3] = float(np.mean(event_features[start:, 1]))\n result[4] = np.log1p(like_count) / 6.0\n result[5] = np.log1p(dislike_count) / 4.0\n return result\n\n\ndef prepare_data(\n public: Path, stage: str, config: Config\n) -> tuple[SequenceData, CandidateSet]:\n data = public / \"data\"\n catalog = pd.read_parquet(data / \"candidate_catalog.parquet\").sort_values(\n \"popularity_rank\"\n )\n item_ids = catalog.item_id.to_numpy(dtype=np.int64, copy=True)\n item_index = pd.Index(item_ids)\n n_items = len(item_ids)\n if n_items >= np.iinfo(np.uint16).max:\n raise RuntimeError(\"uint16 item storage is too small for this catalog\")\n\n if stage == \"validation\":\n prefixes = (\"train\",)\n history_end = base.TRAIN_END\n candidate_name = \"validation_eligible_candidates.npz\"\n state_name = \"validation_preference_state.parquet\"\n else:\n prefixes = (\"train\", \"val\")\n history_end = base.VALIDATION_END\n candidate_name = \"test_eligible_candidates.npz\"\n state_name = \"test_preference_state.parquet\"\n eligibility = CandidateSet.read(data / candidate_name)\n if not np.array_equal(eligibility.item_ids, item_ids):\n raise RuntimeError(\"catalog and candidate artifact item order differ\")\n\n explicit_by_user = base.load_explicit_events(data, prefixes, item_index)\n inference_likes, inference_dislikes, like_counts, dislike_counts = (\n base.load_inference_state(\n data / state_name,\n item_index,\n eligibility.uids,\n n_items,\n )\n )\n inference_rows = {int(uid): row for row, uid in enumerate(eligibility.uids)}\n inference_user_features = np.zeros((len(eligibility.uids), 6), dtype=np.float32)\n inference_sequence_rows = np.full(len(eligibility.uids), -1, dtype=np.int32)\n\n # Snapshot zero is the shared empty explicit state.\n liked_states = [np.full(base.LIKE_ITEMS, n_items, dtype=np.uint16)]\n disliked_states = [\n np.full(base.DISLIKE_ITEMS, n_items, dtype=np.uint16)\n ]\n state_counts: list[tuple[int, int]] = [(0, 0)]\n user_ids: list[int] = []\n offsets = [0]\n timestamp_parts: list[np.ndarray] = []\n item_parts: list[np.ndarray] = []\n event_parts: list[np.ndarray] = []\n activity_parts: list[np.ndarray] = []\n state_parts: list[np.ndarray] = []\n\n listen_paths = [data / f\"{prefix}_listens.parquet\" for prefix in prefixes]\n started = time.monotonic()\n for seen_users, (uid, listens) in enumerate(\n base.merge_listen_streams(listen_paths, item_index), start=1\n ):\n parts = build_sequence_parts(\n listens,\n explicit_by_user.get(uid, base.empty_explicit_events()),\n n_items,\n liked_states,\n disliked_states,\n state_counts,\n )\n if parts is None:\n continue\n sequence_row = len(user_ids)\n user_ids.append(uid)\n timestamp_parts.append(parts[0].astype(np.uint32, copy=False))\n item_parts.append(parts[1])\n event_parts.append(parts[2])\n activity_parts.append(parts[3])\n state_parts.append(parts[4])\n offsets.append(offsets[-1] + len(parts[1]))\n inference_row = inference_rows.get(uid)\n if inference_row is not None:\n inference_sequence_rows[inference_row] = sequence_row\n inference_user_features[inference_row] = final_user_features(\n parts[0],\n parts[1],\n parts[2],\n history_end,\n int(like_counts[inference_row]),\n int(dislike_counts[inference_row]),\n )\n if seen_users % 1_000 == 0:\n log(\n f\"prepared users={seen_users:,} tokens={offsets[-1]:,} \"\n f\"states={len(liked_states):,} \"\n f\"elapsed={time.monotonic() - started:.1f}s\"\n )\n\n missing = np.flatnonzero(inference_sequence_rows < 0)\n if len(missing):\n inference_user_features[missing, 4] = np.log1p(like_counts[missing]) / 6.0\n inference_user_features[missing, 5] = (\n np.log1p(dislike_counts[missing]) / 4.0\n )\n log(f\"target users without qualifying catalog sequence={len(missing):,}\")\n\n prepared = SequenceData(\n item_ids=item_ids,\n item_features=base.build_item_features(catalog),\n user_ids=np.asarray(user_ids, dtype=np.int64),\n offsets=np.asarray(offsets, dtype=np.int64),\n sequence_timestamps=np.concatenate(timestamp_parts),\n sequence_items=np.concatenate(item_parts),\n event_features=np.concatenate(event_parts),\n activity_features=np.concatenate(activity_parts),\n state_ids=np.concatenate(state_parts),\n liked_states=np.stack(liked_states),\n disliked_states=np.stack(disliked_states),\n state_counts=np.asarray(state_counts, dtype=np.int32),\n inference_user_rows=inference_sequence_rows,\n inference_liked_items=inference_likes,\n inference_disliked_items=inference_dislikes,\n inference_user_features=inference_user_features,\n )\n log(\n f\"prepared stage={stage} sequence_users={len(prepared.user_ids):,} \"\n f\"tokens={len(prepared.sequence_items):,} \"\n f\"next_item_targets={prepared.target_count:,} \"\n f\"states={len(prepared.liked_states):,}\"\n )\n return prepared, eligibility\n\n\ndef cache_metadata(stage: str, config: Config) -> dict[str, object]:\n return {\n \"cache_version\": CACHE_VERSION,\n \"stage\": stage,\n \"positive_listen_threshold\": base.LISTEN_POSITIVE_THRESHOLD,\n \"profile_lookback_days\": base.PROFILE_LOOKBACK_DAYS,\n \"event_time_feature\": \"exp-negative-gap-over-60-days\",\n }\n\n\ndef save_cache(\n path: Path, prepared: SequenceData, stage: str, config: Config\n) -> None:\n path.parent.mkdir(parents=True, exist_ok=True)\n np.savez(\n path,\n metadata=np.asarray(json.dumps(cache_metadata(stage, config), sort_keys=True)),\n **prepared.__dict__,\n )\n log(f\"saved public-derived full-sequence cache: {path}\")\n\n\ndef load_cache(path: Path, stage: str, config: Config) -> SequenceData:\n with np.load(path, allow_pickle=False) as artifact:\n actual = json.loads(str(artifact[\"metadata\"].item()))\n expected = cache_metadata(stage, config)\n if actual != expected:\n raise RuntimeError(f\"sequence cache differs: {actual} != {expected}\")\n values = {\n field: artifact[field].copy()\n for field in SequenceData.__dataclass_fields__\n }\n prepared = SequenceData(**values)\n log(\n f\"loaded full-sequence cache: tokens={len(prepared.sequence_items):,} \"\n f\"targets={prepared.target_count:,}\"\n )\n return prepared\n\n\ndef load_or_prepare(\n public: Path,\n stage: str,\n config: Config,\n cache: Path | None,\n) -> tuple[SequenceData, CandidateSet]:\n candidate_name = (\n \"validation_eligible_candidates.npz\"\n if stage == \"validation\"\n else \"test_eligible_candidates.npz\"\n )\n eligibility = CandidateSet.read(public / \"data\" / candidate_name)\n if cache is not None and cache.is_file():\n prepared = load_cache(cache, stage, config)\n if not np.array_equal(prepared.item_ids, eligibility.item_ids):\n raise RuntimeError(\"cached catalog differs from candidate artifact\")\n return prepared, eligibility\n prepared, built_eligibility = prepare_data(public, stage, config)\n if cache is not None:\n save_cache(cache, prepared, stage, config)\n return prepared, built_eligibility\n\n\ndef build_item_correction(\n prepared: SequenceData,\n config: Config,\n target_starts: np.ndarray,\n target_counts: np.ndarray,\n) -> torch.Tensor:\n \"\"\"Estimate in-batch propensities from the selected training targets.\"\"\"\n counts = np.zeros(len(prepared.item_ids), dtype=np.float64)\n for user_row, count in enumerate(target_counts):\n if not count:\n continue\n offset = int(prepared.offsets[user_row])\n target_start = int(target_starts[user_row])\n counts += np.bincount(\n prepared.sequence_items[\n offset\n + target_start\n + 1 : offset\n + target_start\n + int(count)\n + 1\n ].astype(np.int64),\n minlength=len(prepared.item_ids),\n )\n total = float(counts.sum())\n probabilities = (counts + 1.0) / (total + len(counts))\n raw = np.power(\n (1.0 / len(counts)) / probabilities,\n config.inbatch_correction_power,\n )\n normalization = float(np.sum(counts * raw) / max(total, 1.0))\n correction = np.clip(\n raw / max(normalization, 1.0e-12),\n config.inbatch_correction_min,\n config.inbatch_correction_max,\n ).astype(np.float32)\n mean_on_targets = float(np.sum(counts * correction) / max(total, 1.0))\n log(\n f\"inbatch correction: min={correction.min():.3f} \"\n f\"mean_on_targets={mean_on_targets:.3f} max={correction.max():.3f}\"\n )\n return torch.from_numpy(correction)\n\n\ndef build_known_positive_filter(\n prepared: SequenceData, training_counts: np.ndarray\n) -> np.ndarray:\n \"\"\"Pack each user's complete public-training positive-item set.\"\"\"\n width = (len(prepared.item_ids) + 7) // 8\n packed = np.zeros((len(prepared.user_ids), width), dtype=np.uint8)\n retained = 0\n for user_row, target_count in enumerate(training_counts):\n if not target_count:\n continue\n offset = int(prepared.offsets[user_row])\n # T adjacent targets consume sequence positions 0 through T. The\n # chronological BCE suffix starts at T + 1 and is intentionally absent.\n items = np.unique(\n prepared.sequence_items[\n offset : offset + int(target_count) + 1\n ].astype(np.int64)\n )\n byte_columns = items >> 3\n bit_values = np.left_shift(\n np.uint8(1), (items & 7).astype(np.uint8)\n )\n np.bitwise_or.at(packed[user_row], byte_columns, bit_values)\n retained += len(items)\n log(\n f\"full-history false-negative filter: users={len(prepared.user_ids):,} \"\n f\"unique_user_items={retained:,} bytes={packed.nbytes:,}\"\n )\n return packed\n\n\ndef invalid_full_history_negatives(\n negatives: np.ndarray,\n targets: np.ndarray,\n user_rows: np.ndarray,\n known_positive_packed: np.ndarray,\n liked_items: np.ndarray,\n) -> np.ndarray:\n \"\"\"Return row-level invalid negatives using the complete training prefix.\"\"\"\n byte_columns = negatives >> 3\n bit_columns = negatives & 7\n known = known_positive_packed[user_rows[:, None], byte_columns]\n invalid = ((known >> bit_columns) & 1).astype(bool)\n invalid |= negatives == targets[:, None]\n invalid |= (negatives[:, :, None] == liked_items[:, None, :]).any(axis=2)\n for column in range(1, negatives.shape[1]):\n invalid[:, column] |= (\n negatives[:, column, None] == negatives[:, :column]\n ).any(axis=1)\n return invalid\n\n\ndef draw_full_history_filtered_negatives(\n targets: torch.Tensor,\n user_rows: np.ndarray,\n liked_items: torch.Tensor,\n known_positive_packed: np.ndarray,\n n_items: int,\n config: Config,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n \"\"\"Draw mixed negatives and reject all known public-training positives.\"\"\"\n targets_cpu = targets.detach().cpu().numpy().astype(np.int64, copy=False)\n liked_cpu = liked_items.detach().cpu().numpy().astype(np.int64, copy=False)\n batch_size = len(targets_cpu)\n uniform = torch.randint(\n n_items,\n (batch_size, config.uniform_negatives),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.empty((batch_size, 0), dtype=torch.long)\n corrected_count = 0\n if config.inbatch_negatives:\n unique_targets = torch.unique(targets.detach().cpu())\n corrected_count = min(\n config.inbatch_negatives, max(len(unique_targets) - 1, 0)\n )\n if corrected_count:\n priorities = torch.rand(\n (batch_size, len(unique_targets)), generator=generator\n )\n priorities.masked_fill_(\n unique_targets[None, :].eq(\n torch.from_numpy(targets_cpu)[:, None]\n ),\n -1.0,\n )\n columns = torch.topk(\n priorities, k=corrected_count, dim=1\n ).indices\n inbatch = unique_targets[columns]\n if corrected_count < config.inbatch_negatives:\n filler = torch.randint(\n n_items,\n (batch_size, config.inbatch_negatives - corrected_count),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.cat([inbatch, filler], dim=1)\n\n negatives = torch.cat([uniform, inbatch], dim=1).numpy()\n correction = np.ones_like(negatives, dtype=np.float32)\n if corrected_count:\n correction[\n :, config.uniform_negatives : config.uniform_negatives + corrected_count\n ] = item_correction[inbatch[:, :corrected_count]].numpy()\n for _ in range(32):\n invalid = invalid_full_history_negatives(\n negatives,\n targets_cpu,\n user_rows,\n known_positive_packed,\n liked_cpu,\n )\n if not np.any(invalid):\n return (\n torch.from_numpy(negatives).to(device),\n torch.from_numpy(correction).to(device),\n )\n replacements = torch.randint(\n n_items,\n (int(invalid.sum()),),\n generator=generator,\n device=\"cpu\",\n ).numpy()\n negatives[invalid] = replacements\n # Rejected in-batch items become ordinary uniform draws.\n correction[invalid] = 1.0\n invalid = invalid_full_history_negatives(\n negatives,\n targets_cpu,\n user_rows,\n known_positive_packed,\n liked_cpu,\n )\n if np.any(invalid):\n raise RuntimeError(\"full-history negative resampling failed\")\n return (\n torch.from_numpy(negatives).to(device),\n torch.from_numpy(correction).to(device),\n )\n\n\ndef draw_training_negatives(\n positive_items: torch.Tensor,\n known_history: torch.Tensor,\n liked_items: torch.Tensor,\n user_rows: np.ndarray,\n known_positive_packed: np.ndarray | None,\n prepared: SequenceData,\n config: Config,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n if config.negative_filter_scope == \"full-training-history\":\n if known_positive_packed is None:\n raise RuntimeError(\"full-history negative filter was not built\")\n return draw_full_history_filtered_negatives(\n positive_items,\n user_rows,\n liked_items,\n known_positive_packed,\n len(prepared.item_ids),\n config,\n item_correction,\n generator,\n device,\n )\n return base.draw_filtered_mixed_negatives(\n positive_items,\n known_history,\n liked_items,\n len(prepared.item_ids),\n config.uniform_negatives,\n config.inbatch_negatives,\n item_correction,\n generator,\n device,\n )\n\n\ndef event_sequence(\n model: base.SequentialTwoTower,\n items: torch.Tensor,\n features: torch.Tensor,\n hidden: torch.Tensor,\n) -> tuple[torch.Tensor, torch.Tensor]:\n inputs = torch.cat([model.context_embedding(items), features], dim=-1)\n projected = model.event_projection(inputs)\n return model.sequence_encoder(projected, hidden)\n\n\ndef pool_explicit_states(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n state_ids: np.ndarray,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor, torch.Tensor]:\n unique_states, inverse = np.unique(state_ids, return_inverse=True)\n liked_items = torch.from_numpy(\n prepared.liked_states[unique_states].astype(np.int64)\n ).to(device)\n disliked_items = torch.from_numpy(\n prepared.disliked_states[unique_states].astype(np.int64)\n ).to(device)\n liked_vectors = model._mean_context_channel(liked_items)\n disliked_vectors = model._mean_context_channel(disliked_items)\n inverse_rows = torch.from_numpy(inverse.astype(np.int64)).to(device)\n counts = prepared.state_counts[state_ids]\n return (\n liked_vectors[inverse_rows],\n disliked_vectors[inverse_rows],\n torch.from_numpy(counts.astype(np.float32)).to(device),\n )\n\n\ndef user_queries(\n model: base.SequentialTwoTower,\n sequence_vectors: torch.Tensor,\n liked_vectors: torch.Tensor,\n disliked_vectors: torch.Tensor,\n user_features: torch.Tensor,\n) -> torch.Tensor:\n return model.user_tower(\n torch.cat(\n [sequence_vectors, liked_vectors, disliked_vectors, user_features],\n dim=1,\n )\n )\n\n\ndef candidate_vectors(\n model: base.SequentialTwoTower,\n positives: torch.Tensor,\n negatives: torch.Tensor,\n) -> tuple[torch.Tensor, torch.Tensor]:\n all_items = torch.cat([positives, negatives.reshape(-1)])\n unique_items, inverse = torch.unique(all_items, return_inverse=True)\n encoded = model.encode_items(unique_items)\n positive_vectors = encoded[inverse[: len(positives)]]\n negative_vectors = encoded[inverse[len(positives) :]].reshape(\n len(positives), negatives.shape[1], -1\n )\n return positive_vectors, negative_vectors\n\n\ndef history_windows(\n sequence: np.ndarray,\n start: int,\n count: int,\n width: int,\n padding_item: int,\n) -> np.ndarray:\n base_start = max(0, start - width + 1)\n segment = sequence[base_start : start + count]\n padded = np.concatenate(\n [np.full(width - 1, padding_item, dtype=np.uint16), segment]\n )\n windows = np.lib.stride_tricks.sliding_window_view(padded, width)\n offset = start - base_start\n return windows[offset : offset + count].copy()\n\n\ndef corrected_negative_loss(\n logits: torch.Tensor,\n correction: torch.Tensor,\n config: Config,\n) -> torch.Tensor:\n raw = F.softplus(logits)\n total = config.uniform_negatives + config.inbatch_negatives\n result = torch.zeros(len(logits), device=logits.device)\n if config.uniform_negatives:\n result += config.uniform_negatives * raw[\n :, : config.uniform_negatives\n ].mean(dim=1)\n if config.inbatch_negatives:\n selected = slice(config.uniform_negatives, total)\n weights = correction[:, selected]\n result += config.inbatch_negatives * (\n (raw[:, selected] * weights).sum(dim=1)\n / weights.sum(dim=1).clamp_min(1.0e-6)\n )\n return result / total\n\n\ndef retained_target_window_starts(\n total_targets: np.ndarray,\n validation_counts: np.ndarray,\n training_counts: np.ndarray,\n) -> np.ndarray:\n \"\"\"Return the discarded prefix before the retained chronological suffix.\"\"\"\n training_region_ends = total_targets - validation_counts\n window_starts = training_region_ends - training_counts\n if (\n np.any(window_starts < 0)\n or np.any(training_counts < 0)\n or np.any(validation_counts < 0)\n ):\n raise RuntimeError(\"chronological target partition produced negative counts\")\n if not np.array_equal(\n window_starts + training_counts + validation_counts, total_targets\n ):\n raise RuntimeError(\"chronological target window is not a final contiguous suffix\")\n return window_starts\n\n\ndef chronological_target_partition(\n prepared: SequenceData,\n validation_fraction: float,\n max_train_targets_per_user: int,\n max_validation_targets_per_user: int,\n) -> TargetPartition:\n \"\"\"Select a recent contiguous window with a final out-of-time BCE suffix.\"\"\"\n totals = np.maximum(prepared.lengths - 1, 0).astype(np.int64)\n validation = np.floor(totals * validation_fraction).astype(np.int64)\n validation[(totals >= 2) & (validation == 0)] = 1\n if max_validation_targets_per_user:\n validation = np.minimum(validation, max_validation_targets_per_user)\n\n history_training = totals - validation\n training = history_training.copy()\n if max_train_targets_per_user:\n training = np.minimum(training, max_train_targets_per_user)\n window_starts = retained_target_window_starts(totals, validation, training)\n return TargetPartition(\n window_starts=window_starts,\n training_counts=training,\n validation_counts=validation,\n history_training_counts=history_training,\n )\n\n\ndef load_organic_like_outcomes(\n public: Path, stage: str, item_ids: np.ndarray\n) -> dict[int, dict[int, np.ndarray]]:\n \"\"\"Load mapped organic-like timestamps from public fitting history.\"\"\"\n prefixes = (\"train\",) if stage == \"validation\" else (\"train\", \"val\")\n frames = [\n pd.read_parquet(\n public / \"data\" / f\"{prefix}_likes.parquet\",\n columns=[\"uid\", \"timestamp\", \"item_id\", \"is_organic\"],\n )\n for prefix in prefixes\n ]\n likes = pd.concat(frames, ignore_index=True)\n likes = likes[likes.is_organic == 1].copy()\n mapped = pd.Index(item_ids).get_indexer(\n likes.item_id.to_numpy(dtype=np.int64)\n )\n likes = likes.loc[mapped >= 0, [\"uid\", \"timestamp\"]].copy()\n likes[\"item\"] = mapped[mapped >= 0].astype(np.int32, copy=False)\n likes = likes.sort_values([\"uid\", \"item\", \"timestamp\"], kind=\"stable\")\n nested: dict[int, dict[int, list[int]]] = {}\n for raw_uid, raw_timestamp, raw_item in likes.itertuples(\n index=False, name=None\n ):\n nested.setdefault(int(raw_uid), {}).setdefault(int(raw_item), []).append(\n int(raw_timestamp)\n )\n result = {\n uid: {\n item: np.asarray(timestamps, dtype=np.int64)\n for item, timestamps in items.items()\n }\n for uid, items in nested.items()\n }\n log(\n f\"loaded organic-like outcomes: rows={len(likes):,} \"\n f\"users={len(result):,}\"\n )\n return result\n\n\ndef build_target_weights(\n public: Path,\n stage: str,\n prepared: SequenceData,\n training_counts: np.ndarray,\n config: Config,\n) -> np.ndarray:\n \"\"\"Build nonzero per-token weights without dropping adjacent targets.\"\"\"\n if config.target_weighting == \"uniform\":\n weights = np.ones(len(prepared.sequence_items), dtype=np.float32)\n else:\n weights = np.full(\n len(prepared.sequence_items),\n config.ordinary_target_weight,\n dtype=np.float32,\n )\n organic_strong = (\n (prepared.event_features[:, 0] > 0.5)\n & (prepared.event_features[:, 1] >= 0.8)\n )\n weights[organic_strong] = config.organic_strong_target_weight\n if config.target_weighting == \"discovery-graded\":\n like_outcomes = load_organic_like_outcomes(\n public, stage, prepared.item_ids\n )\n history_end = (\n base.TRAIN_END if stage == \"validation\" else base.VALIDATION_END\n )\n horizon = config.discovery_horizon_days * base.DAY\n grade_three = 0\n for user_row, uid in enumerate(prepared.user_ids):\n start = int(prepared.offsets[user_row])\n end = int(prepared.offsets[user_row + 1])\n if end - start < 2:\n continue\n last_training_target = start + int(training_counts[user_row])\n first_validation_target = last_training_target + 1\n split_timestamp = (\n int(prepared.sequence_timestamps[first_validation_target])\n if first_validation_target < end\n else history_end\n )\n user_likes = like_outcomes.get(int(uid), {})\n seen = {int(prepared.sequence_items[start])}\n for position in range(start + 1, end):\n item = int(prepared.sequence_items[position])\n novel = item not in seen\n seen.add(item)\n if not novel:\n # Repeats remain training examples, but they cannot be\n # eligible discovery outcomes at the corresponding\n # cutoff and therefore keep only the ordinary weight.\n weights[position] = config.ordinary_target_weight\n continue\n timestamps = user_likes.get(item)\n if timestamps is None:\n continue\n timestamp = int(prepared.sequence_timestamps[position])\n upper = min(\n timestamp + horizon,\n split_timestamp\n if position <= last_training_target\n else history_end,\n )\n outcome = int(np.searchsorted(timestamps, timestamp, side=\"left\"))\n if outcome < len(timestamps) and int(timestamps[outcome]) < upper:\n weights[position] = config.organic_like_target_weight\n grade_three += 1\n log(\n f\"grade-3 first-encounter targets={grade_three:,} \"\n f\"horizon_days={config.discovery_horizon_days}\"\n )\n log(\n f\"target weighting={config.target_weighting} \"\n f\"min={weights.min():.3f} mean={weights.mean():.3f} \"\n f\"max={weights.max():.3f}\"\n )\n return weights\n\n\ndef build_dense_block(\n prepared: SequenceData,\n target_weights: np.ndarray,\n users: np.ndarray,\n window_starts: np.ndarray,\n process_counts: np.ndarray,\n score_starts: np.ndarray,\n start: int,\n steps: int,\n config: Config,\n) -> tuple[np.ndarray, ...]:\n \"\"\"Build one causal sequence block and mark the target rows to score.\"\"\"\n batch_size = len(users)\n padding_item = len(prepared.item_ids)\n input_items = np.full((batch_size, steps), padding_item, dtype=np.uint16)\n event_features = np.zeros((batch_size, steps, 3), dtype=np.float16)\n activity_features = np.zeros((batch_size, steps, 4), dtype=np.float16)\n state_ids = np.zeros((batch_size, steps), dtype=np.uint32)\n targets = np.zeros((batch_size, steps), dtype=np.uint16)\n weights = np.ones((batch_size, steps), dtype=np.float32)\n scored = np.zeros((batch_size, steps), dtype=bool)\n histories = np.full(\n (batch_size, steps, config.bptt_steps),\n padding_item,\n dtype=np.uint16,\n )\n processed = np.zeros(batch_size, dtype=np.int32)\n\n lengths = prepared.lengths\n for row, user_row in enumerate(users):\n count = min(steps, max(int(process_counts[row]) - start, 0))\n if not count:\n continue\n offset = int(prepared.offsets[user_row])\n window_start = int(window_starts[row])\n absolute_start = window_start + start\n positions = slice(\n offset + absolute_start,\n offset + absolute_start + count,\n )\n target_positions = slice(\n offset + absolute_start + 1,\n offset + absolute_start + count + 1,\n )\n input_items[row, :count] = prepared.sequence_items[positions]\n event_features[row, :count] = prepared.event_features[positions]\n activity_features[row, :count] = prepared.activity_features[positions]\n state_ids[row, :count] = prepared.state_ids[positions]\n targets[row, :count] = prepared.sequence_items[target_positions]\n weights[row, :count] = target_weights[target_positions]\n score_from = min(count, max(int(score_starts[row]) - start, 0))\n scored[row, score_from:count] = True\n processed[row] = count\n user_sequence = prepared.sequence_items[\n offset : offset + int(lengths[user_row])\n ]\n histories[row, :count] = history_windows(\n user_sequence,\n absolute_start,\n count,\n config.bptt_steps,\n padding_item,\n )\n return (\n input_items,\n event_features,\n activity_features,\n state_ids,\n targets,\n scored,\n histories,\n processed,\n weights,\n )\n\n\ndef score_target_microbatches(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n sequence_vectors: torch.Tensor,\n flat_states: np.ndarray,\n flat_activity: np.ndarray,\n positive_items: torch.Tensor,\n negatives: torch.Tensor,\n negative_weights: torch.Tensor,\n sample_weights: np.ndarray,\n device: torch.device,\n backward: bool,\n) -> tuple[dict[str, float], torch.Tensor | None]:\n \"\"\"Compute one logical block's exact BCE in bounded-memory slices.\"\"\"\n target_count = len(positive_items)\n total_weight = float(sample_weights.sum())\n if total_weight <= 0:\n raise RuntimeError(\"target block has no positive sample weight\")\n sequence_gradient = torch.zeros_like(sequence_vectors) if backward else None\n statistics = {\n \"loss_sum\": 0.0,\n \"positive_sum\": 0.0,\n \"negative_sum\": 0.0,\n \"pair_correct\": 0.0,\n \"weight_sum\": 0.0,\n }\n for begin in range(0, target_count, config.loss_microbatch_size):\n end = min(begin + config.loss_microbatch_size, target_count)\n selected_states = flat_states[begin:end]\n if backward:\n sequence_input = (\n sequence_vectors[begin:end].detach().requires_grad_(True)\n )\n else:\n sequence_input = sequence_vectors[begin:end]\n liked_vectors, disliked_vectors, _ = pool_explicit_states(\n model, prepared, selected_states, device\n )\n counts = prepared.state_counts[selected_states]\n numeric = np.zeros((end - begin, 6), dtype=np.float32)\n numeric[:, :4] = flat_activity[begin:end]\n numeric[:, 4] = np.log1p(counts[:, 0]) / 6.0\n numeric[:, 5] = np.log1p(counts[:, 1]) / 4.0\n query_vectors = user_queries(\n model,\n sequence_input,\n liked_vectors,\n disliked_vectors,\n torch.from_numpy(numeric).to(device),\n )\n positive_vectors, negative_vectors = candidate_vectors(\n model,\n positive_items[begin:end],\n negatives[begin:end],\n )\n positive_logits = (query_vectors * positive_vectors).sum(dim=1)\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n positive_loss = F.softplus(-positive_logits)\n negative_loss = corrected_negative_loss(\n negative_logits, negative_weights[begin:end], config\n )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n loss_sum = (\n selected_weights * (positive_loss + negative_loss) / 2.0\n ).sum()\n if not bool(torch.isfinite(loss_sum)):\n raise RuntimeError(\"dense next-item training produced non-finite loss\")\n if backward:\n (loss_sum / total_weight).backward()\n if sequence_input.grad is None:\n raise RuntimeError(\"sequence microbatch did not produce a gradient\")\n sequence_gradient[begin:end] = sequence_input.grad\n statistics[\"loss_sum\"] += float(loss_sum.detach().cpu())\n statistics[\"positive_sum\"] += float(positive_logits.detach().sum().cpu())\n statistics[\"negative_sum\"] += float(negative_logits.detach().sum().cpu())\n statistics[\"pair_correct\"] += float(\n (positive_logits.detach()[:, None] > negative_logits.detach()).sum().cpu()\n )\n statistics[\"weight_sum\"] += float(selected_weights.sum().detach().cpu())\n return statistics, sequence_gradient\n\n\ndef release_accelerator_cache(device: torch.device) -> None:\n if device.type == \"mps\":\n torch.mps.empty_cache()\n\n\ndef run_training_epoch(\n model: base.SequentialTwoTower,\n optimizer: torch.optim.Optimizer,\n prepared: SequenceData,\n config: Config,\n target_weights: np.ndarray,\n window_starts: np.ndarray,\n target_counts: np.ndarray,\n item_correction: torch.Tensor,\n known_positive_packed: np.ndarray | None,\n generator: torch.Generator,\n device: torch.device,\n epoch: int,\n) -> dict[str, float]:\n model.train()\n train_users = np.flatnonzero(target_counts > 0)\n sorted_users = train_users[\n np.argsort(target_counts[train_users], kind=\"stable\")\n ]\n groups = [\n sorted_users[start : start + config.user_batch_size]\n for start in range(0, len(sorted_users), config.user_batch_size)\n ]\n group_order = torch.randperm(len(groups), generator=generator).numpy()\n negative_count = config.uniform_negatives + config.inbatch_negatives\n total_loss = 0.0\n total_targets = 0\n total_weight = 0.0\n total_positive = 0.0\n total_negative = 0.0\n total_pair_correct = 0.0\n batches = 0\n started = time.monotonic()\n\n for group_number in group_order:\n users = groups[int(group_number)]\n group_counts = target_counts[users]\n hidden = torch.zeros(\n 1, len(users), config.embedding_dim, device=device\n )\n max_targets = int(group_counts.max())\n for start in range(0, max_targets, config.bptt_steps):\n steps = min(config.bptt_steps, max_targets - start)\n block = build_dense_block(\n prepared,\n target_weights,\n users,\n window_starts[users],\n group_counts,\n np.zeros(len(users), dtype=np.int64),\n start,\n steps,\n config,\n )\n sequence_outputs, _ = event_sequence(\n model,\n torch.from_numpy(block[0].astype(np.int64)).to(device),\n torch.from_numpy(block[1].astype(np.float32)).to(device),\n hidden,\n )\n hidden = torch.stack(\n [\n sequence_outputs[row, count - 1]\n if count\n else hidden[0, row]\n for row, count in enumerate(block[7])\n ]\n ).unsqueeze(0).detach()\n flat_scored = block[5].reshape(-1)\n if not np.any(flat_scored):\n continue\n mask = torch.from_numpy(flat_scored).to(device)\n sequence_vectors = sequence_outputs.reshape(\n -1, config.embedding_dim\n )[mask]\n flat_states = block[3].reshape(-1)[flat_scored]\n positive_items = torch.from_numpy(\n block[4].reshape(-1)[flat_scored].astype(np.int64)\n ).to(device)\n known_history = torch.from_numpy(\n block[6].reshape(-1, config.bptt_steps)[flat_scored].astype(\n np.int64\n )\n ).to(device)\n liked_items = torch.from_numpy(\n prepared.liked_states[flat_states].astype(np.int64)\n ).to(device)\n flat_user_rows = np.broadcast_to(\n users[:, None], block[5].shape\n ).reshape(-1)[flat_scored]\n negatives, negative_weights = draw_training_negatives(\n positive_items,\n known_history,\n liked_items,\n flat_user_rows,\n known_positive_packed,\n prepared,\n config,\n item_correction,\n generator,\n device,\n )\n optimizer.zero_grad(set_to_none=True)\n statistics, sequence_gradient = score_target_microbatches(\n model,\n prepared,\n config,\n sequence_vectors,\n flat_states,\n block[2].reshape(-1, 4)[flat_scored].astype(np.float32),\n positive_items,\n negatives,\n negative_weights,\n block[8].reshape(-1)[flat_scored],\n device,\n backward=True,\n )\n if sequence_gradient is None:\n raise RuntimeError(\"training block omitted its sequence gradient\")\n sequence_vectors.backward(sequence_gradient)\n nn.utils.clip_grad_norm_(model.parameters(), config.gradient_clip)\n optimizer.step()\n\n count = len(positive_items)\n total_loss += statistics[\"loss_sum\"]\n total_targets += count\n total_weight += statistics[\"weight_sum\"]\n total_positive += statistics[\"positive_sum\"]\n total_negative += statistics[\"negative_sum\"]\n total_pair_correct += statistics[\"pair_correct\"]\n batches += 1\n if batches % 25 == 0:\n release_accelerator_cache(device)\n if batches % 250 == 0:\n batch_bce = statistics[\"loss_sum\"] / max(\n statistics[\"weight_sum\"], 1.0e-12\n )\n running_bce = total_loss / max(total_weight, 1.0e-12)\n log(\n f\"epoch={epoch}/{config.epochs} batches={batches:,} \"\n f\"targets={total_targets:,} \"\n f\"batch_bce={batch_bce:.6f} \"\n f\"running_bce={running_bce:.6f} \"\n f\"elapsed={time.monotonic() - started:.1f}s\"\n )\n\n return {\n \"epoch\": float(epoch),\n \"loss\": total_loss / max(total_weight, 1.0e-12),\n \"positive_logit\": total_positive / max(total_targets, 1),\n \"negative_logit\": total_negative / max(total_targets * negative_count, 1),\n \"pair_accuracy\": total_pair_correct\n / max(total_targets * negative_count, 1),\n \"targets\": float(total_targets),\n \"target_weight_sum\": total_weight,\n \"batches\": float(batches),\n \"seconds\": time.monotonic() - started,\n }\n\n\n@torch.no_grad()\ndef evaluate_validation_bce(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n target_weights: np.ndarray,\n window_starts: np.ndarray,\n training_counts: np.ndarray,\n validation_counts: np.ndarray,\n item_correction: torch.Tensor,\n known_positive_packed: np.ndarray | None,\n device: torch.device,\n) -> dict[str, float]:\n \"\"\"Measure deterministic BCE on each user's chronological target suffix.\"\"\"\n model.eval()\n totals = training_counts + validation_counts\n users_with_validation = np.flatnonzero(validation_counts > 0)\n ordered = users_with_validation[\n np.argsort(totals[users_with_validation], kind=\"stable\")\n ]\n groups = [\n ordered[start : start + config.user_batch_size]\n for start in range(0, len(ordered), config.user_batch_size)\n ]\n generator = torch.Generator(device=\"cpu\").manual_seed(config.seed + 20_000)\n total_loss = 0.0\n total_targets = 0\n total_weight = 0.0\n started = time.monotonic()\n\n for group_number, users in enumerate(groups, start=1):\n group_totals = totals[users]\n group_starts = training_counts[users]\n hidden = torch.zeros(\n 1, len(users), config.embedding_dim, device=device\n )\n max_targets = int(group_totals.max())\n for start in range(0, max_targets, config.bptt_steps):\n steps = min(config.bptt_steps, max_targets - start)\n block = build_dense_block(\n prepared,\n target_weights,\n users,\n window_starts[users],\n group_totals,\n group_starts,\n start,\n steps,\n config,\n )\n sequence_outputs, _ = event_sequence(\n model,\n torch.from_numpy(block[0].astype(np.int64)).to(device),\n torch.from_numpy(block[1].astype(np.float32)).to(device),\n hidden,\n )\n hidden = torch.stack(\n [\n sequence_outputs[row, count - 1]\n if count\n else hidden[0, row]\n for row, count in enumerate(block[7])\n ]\n ).unsqueeze(0)\n flat_scored = block[5].reshape(-1)\n if not np.any(flat_scored):\n continue\n mask = torch.from_numpy(flat_scored).to(device)\n sequence_vectors = sequence_outputs.reshape(\n -1, config.embedding_dim\n )[mask]\n flat_states = block[3].reshape(-1)[flat_scored]\n positive_items = torch.from_numpy(\n block[4].reshape(-1)[flat_scored].astype(np.int64)\n ).to(device)\n known_history = torch.from_numpy(\n block[6].reshape(-1, config.bptt_steps)[flat_scored].astype(\n np.int64\n )\n ).to(device)\n liked_items = torch.from_numpy(\n prepared.liked_states[flat_states].astype(np.int64)\n ).to(device)\n flat_user_rows = np.broadcast_to(\n users[:, None], block[5].shape\n ).reshape(-1)[flat_scored]\n negatives, negative_weights = draw_training_negatives(\n positive_items,\n known_history,\n liked_items,\n flat_user_rows,\n known_positive_packed,\n prepared,\n config,\n item_correction,\n generator,\n device,\n )\n statistics, _ = score_target_microbatches(\n model,\n prepared,\n config,\n sequence_vectors,\n flat_states,\n block[2].reshape(-1, 4)[flat_scored].astype(np.float32),\n positive_items,\n negatives,\n negative_weights,\n block[8].reshape(-1)[flat_scored],\n device,\n backward=False,\n )\n total_loss += statistics[\"loss_sum\"]\n total_targets += len(positive_items)\n total_weight += statistics[\"weight_sum\"]\n if group_number % 25 == 0:\n release_accelerator_cache(device)\n return {\n \"validation_loss\": total_loss / max(total_weight, 1.0e-12),\n \"validation_targets\": float(total_targets),\n \"validation_target_weight_sum\": total_weight,\n \"validation_seconds\": time.monotonic() - started,\n }\n\n\ndef snapshot_model(model: nn.Module) -> dict[str, torch.Tensor]:\n return {\n name: value.detach().cpu().clone()\n for name, value in model.state_dict().items()\n }\n\n\ndef save_epoch_checkpoint(\n checkpoint_dir: Path,\n epoch: int,\n model_state: dict[str, torch.Tensor],\n config: Config,\n diagnostics: dict[str, float],\n optimizer: torch.optim.Optimizer,\n generator: torch.Generator,\n) -> Path:\n checkpoint_dir.mkdir(parents=True, exist_ok=True)\n path = checkpoint_dir / f\"epoch-{epoch:02d}.pt\"\n torch.save(\n {\n \"epoch\": epoch,\n \"config\": asdict(config),\n \"diagnostics\": diagnostics,\n \"model_state\": model_state,\n \"optimizer_state\": optimizer.state_dict(),\n \"generator_state\": generator.get_state(),\n },\n path,\n )\n log(f\"saved epoch checkpoint: {path}\")\n return path\n\n\ndef train_model(\n public: Path,\n stage: str,\n prepared: SequenceData,\n config: Config,\n device: torch.device,\n checkpoint_dir: Path | None,\n resume_checkpoint: Path | None,\n) -> tuple[base.SequentialTwoTower, list[dict[str, float]]]:\n partition = chronological_target_partition(\n prepared,\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n training_counts = partition.training_counts\n validation_counts = partition.validation_counts\n log(\n f\"chronological BCE window: train={int(training_counts.sum()):,} \"\n f\"validation={int(validation_counts.sum()):,} \"\n f\"selected={partition.selected_count:,} \"\n f\"discarded_old_targets={int(partition.window_starts.sum()):,} \"\n f\"all={prepared.target_count:,}\"\n )\n model = base.SequentialTwoTower(prepared.item_features, model_config(config)).to(\n device\n )\n optimizer = torch.optim.AdamW(\n model.parameters(), lr=config.learning_rate, weight_decay=config.weight_decay\n )\n generator = torch.Generator(device=\"cpu\").manual_seed(config.seed + 10_000)\n item_correction = build_item_correction(\n prepared,\n config,\n partition.window_starts,\n training_counts,\n )\n target_weights = build_target_weights(\n public,\n stage,\n prepared,\n partition.history_training_counts,\n config,\n )\n known_positive_packed = (\n build_known_positive_filter(prepared, partition.history_training_counts)\n if config.negative_filter_scope == \"full-training-history\"\n else None\n )\n history: list[dict[str, float]] = []\n best_loss = float(\"inf\")\n best_epoch = 0\n best_state: dict[str, torch.Tensor] | None = None\n epochs_without_improvement = 0\n start_epoch = 0\n\n if resume_checkpoint is not None:\n payload = torch.load(\n resume_checkpoint, map_location=device, weights_only=False\n )\n stored_config = payload.get(\"config\", {})\n architecture_fields = (\n \"bptt_steps\",\n \"embedding_dim\",\n \"hidden_dim\",\n \"user_hidden_1\",\n \"user_hidden_2\",\n \"item_hidden_1\",\n \"item_hidden_2\",\n )\n mismatches = {\n name: (stored_config.get(name), getattr(config, name))\n for name in architecture_fields\n if stored_config.get(name) != getattr(config, name)\n }\n if mismatches:\n raise RuntimeError(\n f\"resume checkpoint architecture differs: {mismatches}\"\n )\n stored_filter_scope = stored_config.get(\n \"negative_filter_scope\", \"last-100\"\n )\n stored_target_weighting = stored_config.get(\n \"target_weighting\", \"uniform\"\n )\n stored_weight_values = (\n stored_config.get(\"ordinary_target_weight\", 1.0),\n stored_config.get(\"organic_strong_target_weight\", 1.0),\n stored_config.get(\"organic_like_target_weight\", 3.0),\n stored_config.get(\"discovery_horizon_days\", 7),\n )\n current_weight_values = (\n config.ordinary_target_weight,\n config.organic_strong_target_weight,\n config.organic_like_target_weight,\n config.discovery_horizon_days,\n )\n stored_window_values = (\n stored_config.get(\"validation_fraction\", 0.10),\n stored_config.get(\"max_train_targets_per_user\", 0),\n stored_config.get(\"max_validation_targets_per_user\", 0),\n )\n current_window_values = (\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n objective_changed = (\n stored_filter_scope != config.negative_filter_scope\n or stored_target_weighting != config.target_weighting\n or stored_weight_values != current_weight_values\n or stored_window_values != current_window_values\n )\n model.load_state_dict(payload[\"model_state\"])\n if \"optimizer_state\" in payload and not objective_changed:\n optimizer.load_state_dict(payload[\"optimizer_state\"])\n for group in optimizer.param_groups:\n group[\"lr\"] = config.learning_rate\n group[\"weight_decay\"] = config.weight_decay\n log(f\"restored optimizer state from {resume_checkpoint}\")\n else:\n log(\n f\"using a fresh AdamW optimizer after loading {resume_checkpoint}\"\n )\n if \"generator_state\" in payload and not objective_changed:\n generator.set_state(payload[\"generator_state\"])\n start_epoch = int(payload[\"epoch\"])\n if start_epoch >= config.epochs:\n raise RuntimeError(\n f\"resume epoch {start_epoch} must be below --epochs {config.epochs}\"\n )\n prior = {\n str(name): float(value)\n for name, value in payload.get(\"diagnostics\", {}).items()\n }\n if objective_changed:\n prior.pop(\"validation_loss\", None)\n log(\n \"BCE objective changed: filter \"\n f\"{stored_filter_scope}->{config.negative_filter_scope}, \"\n \"target weighting \"\n f\"{stored_target_weighting}->{config.target_weighting}, \"\n f\"target window {stored_window_values}->{current_window_values}; \"\n \"recomputing the resume checkpoint's validation BCE\"\n )\n if \"validation_loss\" not in prior:\n prior.update(\n evaluate_validation_bce(\n model,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n validation_counts,\n item_correction,\n known_positive_packed,\n device,\n )\n )\n history.append(prior)\n best_loss = float(prior[\"validation_loss\"])\n best_epoch = start_epoch\n best_state = snapshot_model(model)\n log(\n f\"resumed validation-BCE training: epoch={start_epoch} \"\n f\"loss={best_loss:.6f}\"\n )\n\n for epoch in range(start_epoch + 1, config.epochs + 1):\n diagnostics = run_training_epoch(\n model,\n optimizer,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n item_correction,\n known_positive_packed,\n generator,\n device,\n epoch,\n )\n validation = evaluate_validation_bce(\n model,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n validation_counts,\n item_correction,\n known_positive_packed,\n device,\n )\n diagnostics.update(validation)\n history.append(diagnostics)\n current_state = snapshot_model(model)\n if checkpoint_dir is not None:\n save_epoch_checkpoint(\n checkpoint_dir,\n epoch,\n current_state,\n config,\n diagnostics,\n optimizer,\n generator,\n )\n improved = (\n diagnostics[\"validation_loss\"]\n < best_loss - config.early_stopping_min_delta\n )\n if improved:\n best_loss = diagnostics[\"validation_loss\"]\n best_epoch = epoch\n best_state = current_state\n epochs_without_improvement = 0\n if checkpoint_dir is not None:\n torch.save(\n {\n \"epoch\": epoch,\n \"config\": asdict(config),\n \"diagnostics\": diagnostics,\n \"model_state\": current_state,\n \"optimizer_state\": optimizer.state_dict(),\n \"generator_state\": generator.get_state(),\n },\n checkpoint_dir / \"best.pt\",\n )\n else:\n epochs_without_improvement += 1\n log(\n f\"epoch={epoch}/{config.epochs} train_loss={diagnostics['loss']:.5f} \"\n f\"validation_loss={diagnostics['validation_loss']:.5f} \"\n f\"pair_accuracy={diagnostics['pair_accuracy']:.3f} \"\n f\"train_targets={int(diagnostics['targets']):,} \"\n f\"validation_targets={int(diagnostics['validation_targets']):,} \"\n f\"train_seconds={diagnostics['seconds']:.1f} \"\n f\"validation_seconds={diagnostics['validation_seconds']:.1f}\"\n )\n release_accelerator_cache(device)\n if epochs_without_improvement >= config.early_stopping_patience:\n log(\n f\"early stopping after epoch={epoch}; \"\n f\"best_epoch={best_epoch} best_validation_loss={best_loss:.6f}\"\n )\n break\n\n if best_state is None:\n raise RuntimeError(\"training did not produce a valid checkpoint\")\n model.load_state_dict(best_state)\n model.to(device).eval()\n log(\n f\"restored best validation-BCE checkpoint: epoch={best_epoch} \"\n f\"loss={best_loss:.6f}\"\n )\n return model, history\n\n\n@torch.inference_mode()\ndef inference_queries(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n device: torch.device,\n) -> torch.Tensor:\n sequence_rows = prepared.inference_user_rows\n lengths = np.where(\n sequence_rows >= 0, prepared.lengths[np.maximum(sequence_rows, 0)], 0\n )\n ordered = np.argsort(lengths, kind=\"stable\")\n result = torch.zeros(\n len(sequence_rows), config.embedding_dim, dtype=torch.float32\n )\n for group_start in range(0, len(ordered), config.user_batch_size):\n output_rows = ordered[group_start : group_start + config.user_batch_size]\n rows = sequence_rows[output_rows]\n group_lengths = lengths[output_rows]\n hidden = torch.zeros(\n 1, len(output_rows), config.embedding_dim, device=device\n )\n for start in range(0, int(group_lengths.max(initial=0)), config.bptt_steps):\n steps = min(\n config.bptt_steps,\n int(group_lengths.max(initial=0)) - start,\n )\n items = np.full(\n (len(output_rows), steps),\n len(prepared.item_ids),\n dtype=np.uint16,\n )\n features = np.zeros((len(output_rows), steps, 3), dtype=np.float16)\n counts = np.zeros(len(output_rows), dtype=np.int32)\n for local_row, sequence_row in enumerate(rows):\n if sequence_row < 0:\n continue\n count = min(steps, max(int(group_lengths[local_row]) - start, 0))\n if not count:\n continue\n offset = int(prepared.offsets[sequence_row])\n selected = slice(offset + start, offset + start + count)\n items[local_row, :count] = prepared.sequence_items[selected]\n features[local_row, :count] = prepared.event_features[selected]\n counts[local_row] = count\n outputs, _ = event_sequence(\n model,\n torch.from_numpy(items.astype(np.int64)).to(device),\n torch.from_numpy(features.astype(np.float32)).to(device),\n hidden,\n )\n next_rows: list[torch.Tensor] = []\n for local_row, count in enumerate(counts):\n next_rows.append(\n outputs[local_row, count - 1]\n if count\n else hidden[0, local_row]\n )\n hidden = torch.stack(next_rows).unsqueeze(0)\n\n liked_items = torch.from_numpy(\n prepared.inference_liked_items[output_rows].astype(np.int64)\n ).to(device)\n disliked_items = torch.from_numpy(\n prepared.inference_disliked_items[output_rows].astype(np.int64)\n ).to(device)\n queries = user_queries(\n model,\n hidden.squeeze(0),\n model._mean_context_channel(liked_items),\n model._mean_context_channel(disliked_items),\n torch.from_numpy(\n prepared.inference_user_features[output_rows].astype(np.float32)\n ).to(device),\n )\n result[torch.from_numpy(output_rows.astype(np.int64))] = queries.cpu()\n return result\n\n\n@torch.inference_mode()\ndef rank_candidates(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n eligibility: CandidateSet,\n config: Config,\n device: torch.device,\n ranking_batch_size: int,\n) -> tuple[pd.DataFrame, dict[str, float]]:\n model.eval()\n queries = inference_queries(model, prepared, config, device)\n item_parts: list[torch.Tensor] = []\n for start in range(0, len(prepared.item_ids), 2_048):\n rows = torch.arange(\n start,\n min(start + 2_048, len(prepared.item_ids)),\n dtype=torch.long,\n device=device,\n )\n item_parts.append(model.encode_items(rows).cpu())\n items = torch.cat(item_parts)\n rows: list[tuple[int, int, int]] = []\n for start in range(0, len(eligibility.uids), ranking_batch_size):\n end = min(start + ranking_batch_size, len(eligibility.uids))\n scores = queries[start:end] @ items.T\n eligible = np.unpackbits(\n eligibility.packed[start:end],\n axis=1,\n count=len(prepared.item_ids),\n bitorder=\"little\",\n ).astype(bool, copy=False)\n scores.masked_fill_(~torch.from_numpy(eligible), -torch.inf)\n selected = torch.topk(scores, k=base.TOP_N, dim=1).indices.numpy()\n for local_row, columns in enumerate(selected):\n uid = int(eligibility.uids[start + local_row])\n rows.extend(\n (uid, int(prepared.item_ids[column]), rank)\n for rank, column in enumerate(columns, start=1)\n )\n if end == len(eligibility.uids) or end % 1_024 == 0:\n log(f\"ranked users={end:,}/{len(eligibility.uids):,}\")\n query_norm = torch.linalg.vector_norm(queries, dim=1).numpy()\n item_norm = torch.linalg.vector_norm(items, dim=1).numpy()\n diagnostics = {\n \"query_norm_mean\": float(query_norm.mean()),\n \"query_norm_std\": float(query_norm.std()),\n \"item_norm_mean\": float(item_norm.mean()),\n \"item_norm_std\": float(item_norm.std()),\n }\n return pd.DataFrame(rows, columns=[\"uid\", \"item_id\", \"rank\"]), diagnostics\n\n\ndef write_ranking(frame: pd.DataFrame, output: Path) -> None:\n output.parent.mkdir(parents=True, exist_ok=True)\n if output.suffix == \".parquet\":\n frame.to_parquet(output, index=False, compression=\"zstd\")\n elif output.suffix == \".csv\":\n frame.to_csv(output, index=False)\n else:\n raise ValueError(\"--output must end in .parquet or .csv\")\n\n\ndef main() -> int:\n parser = argparse.ArgumentParser(description=__doc__)\n parser.add_argument(\n \"--stage\", choices=(\"validation\", \"submission\"), default=\"validation\"\n )\n parser.add_argument(\"--output\", type=Path, required=True)\n parser.add_argument(\"--public-dir\", type=Path, required=True)\n parser.add_argument(\"--prepared-cache\", type=Path)\n parser.add_argument(\"--device\", choices=(\"cpu\", \"mps\"), default=\"cpu\")\n parser.add_argument(\"--threads\", type=int, default=8)\n parser.add_argument(\"--bptt-steps\", type=int, default=Config.bptt_steps)\n parser.add_argument(\n \"--user-batch-size\", type=int, default=Config.user_batch_size\n )\n parser.add_argument(\"--epochs\", type=int, default=Config.epochs)\n parser.add_argument(\"--learning-rate\", type=float, default=Config.learning_rate)\n parser.add_argument(\n \"--loss-microbatch-size\",\n type=int,\n default=Config.loss_microbatch_size,\n )\n parser.add_argument(\n \"--negative-filter-scope\",\n choices=(\"full-training-history\", \"last-100\"),\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n )\n parser.add_argument(\n \"--ordinary-target-weight\",\n type=float,\n default=Config.ordinary_target_weight,\n )\n parser.add_argument(\n \"--organic-strong-target-weight\",\n type=float,\n default=Config.organic_strong_target_weight,\n )\n parser.add_argument(\n \"--organic-like-target-weight\",\n type=float,\n default=Config.organic_like_target_weight,\n )\n parser.add_argument(\n \"--discovery-horizon-days\",\n type=int,\n default=Config.discovery_horizon_days,\n )\n parser.add_argument(\n \"--validation-fraction\",\n type=float,\n default=Config.validation_fraction,\n )\n parser.add_argument(\n \"--max-train-targets-per-user\",\n type=int,\n default=Config.max_train_targets_per_user,\n help=\"retain this many recent pre-validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--max-validation-targets-per-user\",\n type=int,\n default=Config.max_validation_targets_per_user,\n help=\"retain this many final validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--early-stopping-patience\",\n type=int,\n default=Config.early_stopping_patience,\n )\n parser.add_argument(\n \"--early-stopping-min-delta\",\n type=float,\n default=Config.early_stopping_min_delta,\n )\n parser.add_argument(\"--checkpoint-dir\", type=Path)\n parser.add_argument(\"--resume-checkpoint\", type=Path)\n parser.add_argument(\"--seed\", type=int, default=Config.seed)\n parser.add_argument(\"--ranking-batch-size\", type=int, default=64)\n parser.add_argument(\"--prepare-only\", action=\"store_true\")\n args = parser.parse_args()\n\n if args.threads < 1 or args.bptt_steps < 2 or args.user_batch_size < 1:\n parser.error(\"threads, BPTT steps, and user batch size must be positive\")\n if (\n args.epochs < 1\n or args.learning_rate <= 0\n or args.loss_microbatch_size < 1\n or args.early_stopping_patience < 1\n or args.early_stopping_min_delta < 0\n or args.ordinary_target_weight <= 0\n or args.organic_strong_target_weight <= 0\n or args.organic_like_target_weight <= 0\n or args.discovery_horizon_days < 1\n or args.max_train_targets_per_user < 0\n or args.max_validation_targets_per_user < 0\n ):\n parser.error(\"training counts, learning rate, and patience are invalid\")\n if not 0.0 < args.validation_fraction < 1.0:\n parser.error(\"validation fraction must be strictly between zero and one\")\n public = args.public_dir.resolve()\n if not (public / \"data\" / \"candidate_catalog.parquet\").is_file():\n parser.error(f\"public task data not found under {public}\")\n config = Config(\n bptt_steps=args.bptt_steps,\n user_batch_size=args.user_batch_size,\n epochs=args.epochs,\n learning_rate=args.learning_rate,\n loss_microbatch_size=args.loss_microbatch_size,\n negative_filter_scope=args.negative_filter_scope,\n target_weighting=args.target_weighting,\n ordinary_target_weight=args.ordinary_target_weight,\n organic_strong_target_weight=args.organic_strong_target_weight,\n organic_like_target_weight=args.organic_like_target_weight,\n discovery_horizon_days=args.discovery_horizon_days,\n validation_fraction=args.validation_fraction,\n max_train_targets_per_user=args.max_train_targets_per_user,\n max_validation_targets_per_user=args.max_validation_targets_per_user,\n early_stopping_patience=args.early_stopping_patience,\n early_stopping_min_delta=args.early_stopping_min_delta,\n seed=args.seed,\n )\n torch.set_num_threads(args.threads)\n torch.set_num_interop_threads(1)\n seed_everything(config.seed)\n device = torch.device(args.device)\n if device.type == \"mps\" and not torch.backends.mps.is_available():\n parser.error(\"MPS was requested but is unavailable\")\n\n started = time.monotonic()\n prepared, eligibility = load_or_prepare(\n public,\n args.stage,\n config,\n args.prepared_cache.resolve() if args.prepared_cache else None,\n )\n target_partition = chronological_target_partition(\n prepared,\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n if args.prepare_only:\n print(\n json.dumps(\n {\n \"stage\": args.stage,\n \"sequence_users\": len(prepared.user_ids),\n \"sequence_tokens\": len(prepared.sequence_items),\n \"next_item_targets\": prepared.target_count,\n \"selected_training_targets\": int(\n target_partition.training_counts.sum()\n ),\n \"selected_validation_targets\": int(\n target_partition.validation_counts.sum()\n ),\n \"discarded_old_targets\": int(\n target_partition.window_starts.sum()\n ),\n \"target_users\": len(eligibility.uids),\n \"runtime_seconds\": round(time.monotonic() - started, 3),\n \"config\": asdict(config),\n },\n indent=2,\n )\n )\n return 0\n\n model, training_history = train_model(\n public,\n args.stage,\n prepared,\n config,\n device,\n args.checkpoint_dir.resolve() if args.checkpoint_dir else None,\n args.resume_checkpoint.resolve() if args.resume_checkpoint else None,\n )\n ranking, embedding_diagnostics = rank_candidates(\n model,\n prepared,\n eligibility,\n config,\n device,\n args.ranking_batch_size,\n )\n write_ranking(ranking, args.output.resolve())\n print(\n json.dumps(\n {\n \"baseline\": \"yambda-windowed-sequence-two-tower-base\",\n \"stage\": args.stage,\n \"output\": str(args.output.resolve()),\n \"rows\": len(ranking),\n \"users\": int(ranking.uid.nunique()),\n \"runtime_seconds\": round(time.monotonic() - started, 3),\n \"public_inputs_only\": True,\n \"architecture\": {\n \"family\": \"pure-windowed-sequence-two-tower\",\n \"sequence_training\": (\n \"every-adjacent-target-within-recent-chronological-window\"\n ),\n \"sequence_encoder\": \"one-layer-GRU-dense-timestep-loss\",\n \"bptt_hidden_state\": \"carried-and-detached\",\n \"model_selection\": \"chronological-validation-BCE-early-stopping\",\n \"loss_execution\": \"gradient-accumulated-microbatches\",\n \"user_id_embedding\": False,\n \"user_mlp\": \"582-512-256-192\",\n \"item_mlp\": \"196-512-256-192\",\n \"normalization\": \"none\",\n \"score\": \"raw-dot-product\",\n \"objective\": \"sample-weighted-BCE-at-every-timestep\",\n \"negative_sampling\": (\n \"24-uniform-plus-8-propensity-corrected-inbatch\"\n ),\n \"false_negative_filter\": config.negative_filter_scope,\n \"target_weighting\": config.target_weighting,\n },\n \"config\": asdict(config),\n \"sequence_tokens\": len(prepared.sequence_items),\n \"all_next_item_targets\": prepared.target_count,\n \"selected_training_targets\": int(\n target_partition.training_counts.sum()\n ),\n \"selected_validation_targets\": int(\n target_partition.validation_counts.sum()\n ),\n \"discarded_old_targets\": int(target_partition.window_starts.sum()),\n \"training_history\": training_history,\n \"embedding_diagnostics\": embedding_diagnostics,\n \"torch_version\": torch.__version__,\n \"device\": str(device),\n },\n indent=2,\n )\n )\n return 0\n\n\nif __name__ == \"__main__\":\n raise SystemExit(main())\n",
"diffStat": {
"model_added_lines": 2,
"model_removed_lines": 0,
"model_added_chars": 54,
"model_removed_chars": 0,
"user_added_lines": 0,
"user_removed_lines": 0,
"user_added_chars": 0,
"user_removed_chars": 0
},
"isNewFile": false,
"isBuildFile": false
},
"description": "solution/train_current_model.py: negative_filter_scope: str... => negative_filter_scope: str...",
"displayName": "Edit",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T06:06:46.052Z",
"content": "Successfully modified file: /app/solution/train_current_model.py (1 replacements). Here is the updated code:\n...\n inbatch_negatives: int = 8\n inbatch_correction_power: float = 0.5\n inbatch_correction_min: float = 0.25\n inbatch_correction_max: float = 4.0\n negative_filter_scope: str = \"full-training-history\"\n objective: str = \"bce\"\n temperature: float = 0.2\n target_weighting: str = \"organic-strong\"\n ordinary_target_weight: float = 0.25\n organic_strong_target_weight: float = 1.0\n organic_like_target_weight: float = 3.0\n discovery_horizon_days: int = 7\n...",
"isError": false
},
"responseMessageId": "29590359-c4b4-4061-ae25-e6dce5183073"
}
]
},
{
"id": 46,
"role": "agent",
"timestamp": "2026-09-29T06:06:50.731Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T06:06:50.731Z",
"callId": "read_file__call_5977661",
"name": "read_file",
"input": {
"start_line": 1160,
"file_path": "/app/solution/train_current_model.py",
"end_line": 1190
},
"recorded": {
"id": "read_file__call_5977661",
"name": "read_file",
"args": {
"start_line": 1160,
"file_path": "/app/solution/train_current_model.py",
"end_line": 1190
},
"result": [
{
"functionResponse": {
"id": "read_file__call_5977661",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1160-1190 of 2167 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 1191.\n\n--- FILE CONTENT (truncated) ---\n liked_vectors,\n disliked_vectors,\n torch.from_numpy(numeric).to(device),\n )\n positive_vectors, negative_vectors = candidate_vectors(\n model,\n positive_items[begin:end],\n negatives[begin:end],\n )\n positive_logits = (query_vectors * positive_vectors).sum(dim=1)\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n positive_loss = F.softplus(-positive_logits)\n negative_loss = corrected_negative_loss(\n negative_logits, negative_weights[begin:end], config\n )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n loss_sum = (\n selected_weights * (positive_loss + negative_loss) / 2.0\n ).sum()\n if not bool(torch.isfinite(loss_sum)):\n raise RuntimeError(\"dense next-item training produced non-finite loss\")\n if backward:\n (loss_sum / total_weight).backward()\n if sequence_input.grad is None:\n raise RuntimeError(\"sequence microbatch did not produce a gradient\")\n sequence_gradient[begin:end] = sequence_input.grad\n statistics[\"loss_sum\"] += float(loss_sum.detach().cpu())"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T06:06:50.737Z",
"resultDisplay": "Read lines 1160-1190 of 2167 from solution/train_current_model.py",
"description": "solution/train_current_model.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T06:06:52.225Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1160-1190 of 2167 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 1191.\n\n--- FILE CONTENT (truncated) ---\n liked_vectors,\n disliked_vectors,\n torch.from_numpy(numeric).to(device),\n )\n positive_vectors, negative_vectors = candidate_vectors(\n model,\n positive_items[begin:end],\n negatives[begin:end],\n )\n positive_logits = (query_vectors * positive_vectors).sum(dim=1)\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n positive_loss = F.softplus(-positive_logits)\n negative_loss = corrected_negative_loss(\n negative_logits, negative_weights[begin:end], config\n )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n loss_sum = (\n selected_weights * (positive_loss + negative_loss) / 2.0\n ).sum()\n if not bool(torch.isfinite(loss_sum)):\n raise RuntimeError(\"dense next-item training produced non-finite loss\")\n if backward:\n (loss_sum / total_weight).backward()\n if sequence_input.grad is None:\n raise RuntimeError(\"sequence microbatch did not produce a gradient\")\n sequence_gradient[begin:end] = sequence_input.grad\n statistics[\"loss_sum\"] += float(loss_sum.detach().cpu())",
"isError": false
},
"responseMessageId": "231dab4b-03b3-4789-b1a0-fdaf1e0dd401"
}
]
},
{
"id": 47,
"role": "agent",
"timestamp": "2026-09-29T06:06:57.247Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T06:06:57.247Z",
"callId": "replace__call_149790",
"name": "replace",
"input": {
"instruction": "Support infonce objective alongside bce in score_target_microbatches",
"old_string": " positive_logits = (query_vectors * positive_vectors).sum(dim=1)\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n positive_loss = F.softplus(-positive_logits)\n negative_loss = corrected_negative_loss(\n negative_logits, negative_weights[begin:end], config\n )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n loss_sum = (\n selected_weights * (positive_loss + negative_loss) / 2.0\n ).sum()",
"file_path": "/app/solution/train_current_model.py",
"new_string": " positive_logits = (query_vectors * positive_vectors).sum(dim=1)\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n if config.objective == \"infonce\":\n cat_logits = torch.cat(\n [positive_logits.unsqueeze(1), negative_logits], dim=1\n ) / config.temperature\n labels = torch.zeros(\n len(positive_logits), dtype=torch.long, device=device\n )\n raw_loss = F.cross_entropy(cat_logits, labels, reduction=\"none\")\n loss_sum = (selected_weights * raw_loss).sum()\n else:\n positive_loss = F.softplus(-positive_logits)\n negative_loss = corrected_negative_loss(\n negative_logits, negative_weights[begin:end], config\n )\n loss_sum = (\n selected_weights * (positive_loss + negative_loss) / 2.0\n ).sum()",
"allow_multiple": false
},
"recorded": {
"id": "replace__call_149790",
"name": "replace",
"args": {
"instruction": "Support infonce objective alongside bce in score_target_microbatches",
"old_string": " positive_logits = (query_vectors * positive_vectors).sum(dim=1)\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n positive_loss = F.softplus(-positive_logits)\n negative_loss = corrected_negative_loss(\n negative_logits, negative_weights[begin:end], config\n )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n loss_sum = (\n selected_weights * (positive_loss + negative_loss) / 2.0\n ).sum()",
"file_path": "/app/solution/train_current_model.py",
"new_string": " positive_logits = (query_vectors * positive_vectors).sum(dim=1)\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n if config.objective == \"infonce\":\n cat_logits = torch.cat(\n [positive_logits.unsqueeze(1), negative_logits], dim=1\n ) / config.temperature\n labels = torch.zeros(\n len(positive_logits), dtype=torch.long, device=device\n )\n raw_loss = F.cross_entropy(cat_logits, labels, reduction=\"none\")\n loss_sum = (selected_weights * raw_loss).sum()\n else:\n positive_loss = F.softplus(-positive_logits)\n negative_loss = corrected_negative_loss(\n negative_logits, negative_weights[begin:end], config\n )\n loss_sum = (\n selected_weights * (positive_loss + negative_loss) / 2.0\n ).sum()",
"allow_multiple": false
},
"result": [
{
"functionResponse": {
"id": "replace__call_149790",
"name": "replace",
"response": {
"output": "Successfully modified file: /app/solution/train_current_model.py (1 replacements). Here is the updated code:\n...\n )\n positive_logits = (query_vectors * positive_vectors).sum(dim=1)\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n if config.objective == \"infonce\":\n cat_logits = torch.cat(\n [positive_logits.unsqueeze(1), negative_logits], dim=1\n ) / config.temperature\n labels = torch.zeros(\n len(positive_logits), dtype=torch.long, device=device\n )\n raw_loss = F.cross_entropy(cat_logits, labels, reduction=\"none\")\n loss_sum = (selected_weights * raw_loss).sum()\n else:\n positive_loss = F.softplus(-positive_logits)\n negative_loss = corrected_negative_loss(\n negative_logits, negative_weights[begin:end], config\n )\n loss_sum = (\n selected_weights * (positive_loss + negative_loss) / 2.0\n ).sum()\n if not bool(torch.isfinite(loss_sum)):\n raise RuntimeError(\"dense next-item training produced non-finite loss\")\n if backward:\n (loss_sum / total_weight).backward()\n if sequence_input.grad is None:\n..."
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T06:06:57.264Z",
"resultDisplay": {
"fileDiff": "Index: train_current_model.py\n===================================================================\n--- train_current_model.py\tCurrent\n+++ train_current_model.py\tProposed\n@@ -1170,16 +1170,26 @@\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n- positive_loss = F.softplus(-positive_logits)\n- negative_loss = corrected_negative_loss(\n- negative_logits, negative_weights[begin:end], config\n- )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n- loss_sum = (\n- selected_weights * (positive_loss + negative_loss) / 2.0\n- ).sum()\n+ if config.objective == \"infonce\":\n+ cat_logits = torch.cat(\n+ [positive_logits.unsqueeze(1), negative_logits], dim=1\n+ ) / config.temperature\n+ labels = torch.zeros(\n+ len(positive_logits), dtype=torch.long, device=device\n+ )\n+ raw_loss = F.cross_entropy(cat_logits, labels, reduction=\"none\")\n+ loss_sum = (selected_weights * raw_loss).sum()\n+ else:\n+ positive_loss = F.softplus(-positive_logits)\n+ negative_loss = corrected_negative_loss(\n+ negative_logits, negative_weights[begin:end], config\n+ )\n+ loss_sum = (\n+ selected_weights * (positive_loss + negative_loss) / 2.0\n+ ).sum()\n if not bool(torch.isfinite(loss_sum)):\n raise RuntimeError(\"dense next-item training produced non-finite loss\")\n if backward:\n",
"fileName": "train_current_model.py",
"filePath": "/app/solution/train_current_model.py",
"originalContent": "#!/usr/bin/env python3\n\"\"\"Train the Yambda windowed two-tower baseline.\n\nEvery catalog listen with at least 50 percent completion is a sequence token.\nWithin each selected chronological window, every adjacent target contributes a\nloss. The fast default retains each user's most recent 128 training targets and\n16 held-out validation targets; passing zero for either cap restores the\nuncapped behavior for that side of the split. A 100-step truncated-BPTT block\ncomputes all of its timestep losses and carries the detached GRU hidden state\ninto the next block.\n\nThe model is a pure two-tower retriever: no user-ID embedding, no output\nnormalization, raw dot-product retrieval, sample-weighted binary cross entropy,\nand 24 uniform plus eight propensity-corrected unique in-batch negatives. Every\nselected adjacent target retains a nonzero weight. Only files under the supplied\npublic directory are read.\n\"\"\"\nfrom __future__ import annotations\n\nimport argparse\nimport json\nimport random\nimport sys\nimport time\nfrom dataclasses import asdict, dataclass\nfrom pathlib import Path\n\nimport numpy as np\nimport pandas as pd\nimport torch\nfrom torch import nn\nfrom torch.nn import functional as F\n\nimport two_tower_model as base\n\n\nCACHE_VERSION = 2\n\n\n@dataclass(frozen=True)\nclass Config:\n bptt_steps: int = 100\n user_batch_size: int = 64\n embedding_dim: int = 192\n hidden_dim: int = 192\n user_hidden_1: int = 512\n user_hidden_2: int = 256\n item_hidden_1: int = 512\n item_hidden_2: int = 256\n dropout: float = 0.05\n epochs: int = 1\n learning_rate: float = 1.0e-3\n weight_decay: float = 1.0e-5\n gradient_clip: float = 2.0\n loss_microbatch_size: int = 65_536\n uniform_negatives: int = 24\n inbatch_negatives: int = 8\n inbatch_correction_power: float = 0.5\n inbatch_correction_min: float = 0.25\n inbatch_correction_max: float = 4.0\n negative_filter_scope: str = \"full-training-history\"\n objective: str = \"bce\"\n temperature: float = 0.2\n target_weighting: str = \"organic-strong\"\n ordinary_target_weight: float = 0.25\n organic_strong_target_weight: float = 1.0\n organic_like_target_weight: float = 3.0\n discovery_horizon_days: int = 7\n validation_fraction: float = 0.10\n max_train_targets_per_user: int = 128\n max_validation_targets_per_user: int = 16\n early_stopping_patience: int = 1\n early_stopping_min_delta: float = 1.0e-4\n seed: int = 29\n\n\n@dataclass(frozen=True)\nclass CandidateSet:\n uids: np.ndarray\n item_ids: np.ndarray\n packed: np.ndarray\n\n @classmethod\n def read(cls, path: Path) -> \"CandidateSet\":\n with np.load(path, allow_pickle=False) as artifact:\n result = cls(\n uids=artifact[\"uids\"].astype(np.int64, copy=True),\n item_ids=artifact[\"item_ids\"].astype(np.int64, copy=True),\n packed=artifact[\"packed_eligible\"].astype(np.uint8, copy=True),\n )\n expected = (len(result.uids), (len(result.item_ids) + 7) // 8)\n if result.packed.shape != expected:\n raise ValueError(\n f\"invalid candidate artifact shape {result.packed.shape}; \"\n f\"expected {expected}\"\n )\n return result\n\n\n@dataclass(frozen=True)\nclass SequenceData:\n item_ids: np.ndarray\n item_features: np.ndarray\n user_ids: np.ndarray\n offsets: np.ndarray\n sequence_timestamps: np.ndarray\n sequence_items: np.ndarray\n event_features: np.ndarray\n activity_features: np.ndarray\n state_ids: np.ndarray\n liked_states: np.ndarray\n disliked_states: np.ndarray\n state_counts: np.ndarray\n inference_user_rows: np.ndarray\n inference_liked_items: np.ndarray\n inference_disliked_items: np.ndarray\n inference_user_features: np.ndarray\n\n @property\n def lengths(self) -> np.ndarray:\n return np.diff(self.offsets)\n\n @property\n def target_count(self) -> int:\n return int(np.maximum(self.lengths - 1, 0).sum())\n\n\n@dataclass(frozen=True)\nclass TargetPartition:\n \"\"\"A contiguous recent target window and its chronological BCE split.\"\"\"\n\n window_starts: np.ndarray\n training_counts: np.ndarray\n validation_counts: np.ndarray\n history_training_counts: np.ndarray\n\n @property\n def selected_count(self) -> int:\n return int(self.training_counts.sum() + self.validation_counts.sum())\n\n\ndef log(message: str) -> None:\n print(message, file=sys.stderr, flush=True)\n\n\ndef seed_everything(seed: int) -> None:\n random.seed(seed)\n np.random.seed(seed)\n torch.manual_seed(seed)\n\n\ndef model_config(config: Config) -> base.TowerConfig:\n \"\"\"Map the dense trainer configuration onto the unchanged tower.\"\"\"\n return base.TowerConfig(\n sequence_length=config.bptt_steps,\n embedding_dim=config.embedding_dim,\n hidden_dim=config.hidden_dim,\n user_hidden_1=config.user_hidden_1,\n user_hidden_2=config.user_hidden_2,\n item_hidden_1=config.item_hidden_1,\n item_hidden_2=config.item_hidden_2,\n dropout=config.dropout,\n epochs=config.epochs,\n learning_rate=config.learning_rate,\n weight_decay=config.weight_decay,\n gradient_clip=config.gradient_clip,\n uniform_negatives=config.uniform_negatives,\n inbatch_negatives=config.inbatch_negatives,\n inbatch_correction_power=config.inbatch_correction_power,\n inbatch_correction_min=config.inbatch_correction_min,\n inbatch_correction_max=config.inbatch_correction_max,\n seed=config.seed,\n )\n\n\ndef append_state_snapshot(\n active_likes: dict[int, int],\n active_dislikes: dict[int, int],\n n_items: int,\n liked_states: list[np.ndarray],\n disliked_states: list[np.ndarray],\n state_counts: list[tuple[int, int]],\n) -> int:\n liked_states.append(base.pack_active_items(active_likes, base.LIKE_ITEMS, n_items))\n disliked_states.append(\n base.pack_active_items(active_dislikes, base.DISLIKE_ITEMS, n_items)\n )\n state_counts.append((len(active_likes), len(active_dislikes)))\n return len(liked_states) - 1\n\n\ndef causal_activity_features(\n timestamps: np.ndarray,\n items: np.ndarray,\n organic: np.ndarray,\n completion: np.ndarray,\n) -> np.ndarray:\n \"\"\"Compute the first four reviewed user features at every event cutoff.\"\"\"\n length = len(items)\n result = np.zeros((length, 4), dtype=np.float16)\n if not length:\n return result\n left_rows = np.searchsorted(\n timestamps,\n timestamps - base.PROFILE_LOOKBACK_DAYS * base.DAY,\n side=\"left\",\n )\n rows = np.arange(length, dtype=np.int64)\n counts = rows - left_rows + 1\n organic_sum = np.cumsum(organic.astype(np.float64))\n completion_sum = np.cumsum(np.minimum(completion, 100).astype(np.float64))\n before_organic = np.where(\n left_rows > 0, organic_sum[np.maximum(left_rows - 1, 0)], 0.0\n )\n before_completion = np.where(\n left_rows > 0, completion_sum[np.maximum(left_rows - 1, 0)], 0.0\n )\n\n unique_counts = np.empty(length, dtype=np.int32)\n frequencies: dict[int, int] = {}\n left = 0\n for right, raw_item in enumerate(items):\n item = int(raw_item)\n frequencies[item] = frequencies.get(item, 0) + 1\n required_left = int(left_rows[right])\n while left < required_left:\n old_item = int(items[left])\n remaining = frequencies[old_item] - 1\n if remaining:\n frequencies[old_item] = remaining\n else:\n del frequencies[old_item]\n left += 1\n unique_counts[right] = len(frequencies)\n\n result[:, 0] = (np.log1p(counts) / 8.0).astype(np.float16)\n result[:, 1] = (np.log1p(unique_counts) / 6.0).astype(np.float16)\n result[:, 2] = (\n (organic_sum - before_organic) / np.maximum(counts, 1)\n ).astype(np.float16)\n result[:, 3] = (\n (completion_sum - before_completion)\n / np.maximum(counts, 1)\n / 100.0\n ).astype(np.float16)\n return result\n\n\ndef build_sequence_parts(\n listens: base.ListenEvents,\n explicit: base.ExplicitEvents,\n n_items: int,\n liked_states: list[np.ndarray],\n disliked_states: list[np.ndarray],\n state_counts: list[tuple[int, int]],\n) -> tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray, np.ndarray] | None:\n qualifying = listens.completion_pct >= base.LISTEN_POSITIVE_THRESHOLD\n timestamps = listens.timestamps[qualifying].astype(np.int64, copy=False)\n items = listens.items[qualifying].astype(np.uint16, copy=False)\n organic = listens.organic[qualifying].astype(np.float32, copy=False)\n completion = listens.completion_pct[qualifying].astype(np.float32, copy=False)\n if not len(items):\n return None\n\n event_features = np.zeros((len(items), 3), dtype=np.float16)\n event_features[:, 0] = organic.astype(np.float16)\n event_features[:, 1] = (\n np.minimum(completion, 100.0) / 100.0\n ).astype(np.float16)\n if len(items) > 1:\n gaps = np.maximum(0, timestamps[1:] - timestamps[:-1])\n event_features[1:, 2] = np.exp(\n -gaps.astype(np.float64) / (60.0 * base.DAY)\n ).astype(np.float16)\n\n activity = causal_activity_features(\n timestamps, items, organic, completion\n )\n states = np.zeros(len(items), dtype=np.uint32)\n active_likes: dict[int, int] = {}\n active_dislikes: dict[int, int] = {}\n explicit_position = 0\n current_state = 0\n for row, timestamp in enumerate(timestamps):\n changed = False\n while (\n explicit_position < len(explicit.timestamps)\n and int(explicit.timestamps[explicit_position]) < int(timestamp)\n ):\n item = int(explicit.items[explicit_position])\n event_timestamp = int(explicit.timestamps[explicit_position])\n destination = (\n active_likes\n if int(explicit.channel[explicit_position]) == 0\n else active_dislikes\n )\n if int(explicit.active[explicit_position]) == 1:\n destination[item] = event_timestamp\n else:\n destination.pop(item, None)\n explicit_position += 1\n changed = True\n if changed:\n current_state = append_state_snapshot(\n active_likes,\n active_dislikes,\n n_items,\n liked_states,\n disliked_states,\n state_counts,\n )\n states[row] = np.uint32(current_state)\n return timestamps, items.copy(), event_features, activity, states\n\n\ndef final_user_features(\n timestamps: np.ndarray,\n items: np.ndarray,\n event_features: np.ndarray,\n history_end: int,\n like_count: int,\n dislike_count: int,\n) -> np.ndarray:\n start = int(\n np.searchsorted(\n timestamps,\n history_end - base.PROFILE_LOOKBACK_DAYS * base.DAY,\n side=\"left\",\n )\n )\n result = np.zeros(6, dtype=np.float32)\n recent_items = items[start:]\n if len(recent_items):\n result[0] = np.log1p(len(recent_items)) / 8.0\n result[1] = np.log1p(len(np.unique(recent_items))) / 6.0\n result[2] = float(np.mean(event_features[start:, 0]))\n result[3] = float(np.mean(event_features[start:, 1]))\n result[4] = np.log1p(like_count) / 6.0\n result[5] = np.log1p(dislike_count) / 4.0\n return result\n\n\ndef prepare_data(\n public: Path, stage: str, config: Config\n) -> tuple[SequenceData, CandidateSet]:\n data = public / \"data\"\n catalog = pd.read_parquet(data / \"candidate_catalog.parquet\").sort_values(\n \"popularity_rank\"\n )\n item_ids = catalog.item_id.to_numpy(dtype=np.int64, copy=True)\n item_index = pd.Index(item_ids)\n n_items = len(item_ids)\n if n_items >= np.iinfo(np.uint16).max:\n raise RuntimeError(\"uint16 item storage is too small for this catalog\")\n\n if stage == \"validation\":\n prefixes = (\"train\",)\n history_end = base.TRAIN_END\n candidate_name = \"validation_eligible_candidates.npz\"\n state_name = \"validation_preference_state.parquet\"\n else:\n prefixes = (\"train\", \"val\")\n history_end = base.VALIDATION_END\n candidate_name = \"test_eligible_candidates.npz\"\n state_name = \"test_preference_state.parquet\"\n eligibility = CandidateSet.read(data / candidate_name)\n if not np.array_equal(eligibility.item_ids, item_ids):\n raise RuntimeError(\"catalog and candidate artifact item order differ\")\n\n explicit_by_user = base.load_explicit_events(data, prefixes, item_index)\n inference_likes, inference_dislikes, like_counts, dislike_counts = (\n base.load_inference_state(\n data / state_name,\n item_index,\n eligibility.uids,\n n_items,\n )\n )\n inference_rows = {int(uid): row for row, uid in enumerate(eligibility.uids)}\n inference_user_features = np.zeros((len(eligibility.uids), 6), dtype=np.float32)\n inference_sequence_rows = np.full(len(eligibility.uids), -1, dtype=np.int32)\n\n # Snapshot zero is the shared empty explicit state.\n liked_states = [np.full(base.LIKE_ITEMS, n_items, dtype=np.uint16)]\n disliked_states = [\n np.full(base.DISLIKE_ITEMS, n_items, dtype=np.uint16)\n ]\n state_counts: list[tuple[int, int]] = [(0, 0)]\n user_ids: list[int] = []\n offsets = [0]\n timestamp_parts: list[np.ndarray] = []\n item_parts: list[np.ndarray] = []\n event_parts: list[np.ndarray] = []\n activity_parts: list[np.ndarray] = []\n state_parts: list[np.ndarray] = []\n\n listen_paths = [data / f\"{prefix}_listens.parquet\" for prefix in prefixes]\n started = time.monotonic()\n for seen_users, (uid, listens) in enumerate(\n base.merge_listen_streams(listen_paths, item_index), start=1\n ):\n parts = build_sequence_parts(\n listens,\n explicit_by_user.get(uid, base.empty_explicit_events()),\n n_items,\n liked_states,\n disliked_states,\n state_counts,\n )\n if parts is None:\n continue\n sequence_row = len(user_ids)\n user_ids.append(uid)\n timestamp_parts.append(parts[0].astype(np.uint32, copy=False))\n item_parts.append(parts[1])\n event_parts.append(parts[2])\n activity_parts.append(parts[3])\n state_parts.append(parts[4])\n offsets.append(offsets[-1] + len(parts[1]))\n inference_row = inference_rows.get(uid)\n if inference_row is not None:\n inference_sequence_rows[inference_row] = sequence_row\n inference_user_features[inference_row] = final_user_features(\n parts[0],\n parts[1],\n parts[2],\n history_end,\n int(like_counts[inference_row]),\n int(dislike_counts[inference_row]),\n )\n if seen_users % 1_000 == 0:\n log(\n f\"prepared users={seen_users:,} tokens={offsets[-1]:,} \"\n f\"states={len(liked_states):,} \"\n f\"elapsed={time.monotonic() - started:.1f}s\"\n )\n\n missing = np.flatnonzero(inference_sequence_rows < 0)\n if len(missing):\n inference_user_features[missing, 4] = np.log1p(like_counts[missing]) / 6.0\n inference_user_features[missing, 5] = (\n np.log1p(dislike_counts[missing]) / 4.0\n )\n log(f\"target users without qualifying catalog sequence={len(missing):,}\")\n\n prepared = SequenceData(\n item_ids=item_ids,\n item_features=base.build_item_features(catalog),\n user_ids=np.asarray(user_ids, dtype=np.int64),\n offsets=np.asarray(offsets, dtype=np.int64),\n sequence_timestamps=np.concatenate(timestamp_parts),\n sequence_items=np.concatenate(item_parts),\n event_features=np.concatenate(event_parts),\n activity_features=np.concatenate(activity_parts),\n state_ids=np.concatenate(state_parts),\n liked_states=np.stack(liked_states),\n disliked_states=np.stack(disliked_states),\n state_counts=np.asarray(state_counts, dtype=np.int32),\n inference_user_rows=inference_sequence_rows,\n inference_liked_items=inference_likes,\n inference_disliked_items=inference_dislikes,\n inference_user_features=inference_user_features,\n )\n log(\n f\"prepared stage={stage} sequence_users={len(prepared.user_ids):,} \"\n f\"tokens={len(prepared.sequence_items):,} \"\n f\"next_item_targets={prepared.target_count:,} \"\n f\"states={len(prepared.liked_states):,}\"\n )\n return prepared, eligibility\n\n\ndef cache_metadata(stage: str, config: Config) -> dict[str, object]:\n return {\n \"cache_version\": CACHE_VERSION,\n \"stage\": stage,\n \"positive_listen_threshold\": base.LISTEN_POSITIVE_THRESHOLD,\n \"profile_lookback_days\": base.PROFILE_LOOKBACK_DAYS,\n \"event_time_feature\": \"exp-negative-gap-over-60-days\",\n }\n\n\ndef save_cache(\n path: Path, prepared: SequenceData, stage: str, config: Config\n) -> None:\n path.parent.mkdir(parents=True, exist_ok=True)\n np.savez(\n path,\n metadata=np.asarray(json.dumps(cache_metadata(stage, config), sort_keys=True)),\n **prepared.__dict__,\n )\n log(f\"saved public-derived full-sequence cache: {path}\")\n\n\ndef load_cache(path: Path, stage: str, config: Config) -> SequenceData:\n with np.load(path, allow_pickle=False) as artifact:\n actual = json.loads(str(artifact[\"metadata\"].item()))\n expected = cache_metadata(stage, config)\n if actual != expected:\n raise RuntimeError(f\"sequence cache differs: {actual} != {expected}\")\n values = {\n field: artifact[field].copy()\n for field in SequenceData.__dataclass_fields__\n }\n prepared = SequenceData(**values)\n log(\n f\"loaded full-sequence cache: tokens={len(prepared.sequence_items):,} \"\n f\"targets={prepared.target_count:,}\"\n )\n return prepared\n\n\ndef load_or_prepare(\n public: Path,\n stage: str,\n config: Config,\n cache: Path | None,\n) -> tuple[SequenceData, CandidateSet]:\n candidate_name = (\n \"validation_eligible_candidates.npz\"\n if stage == \"validation\"\n else \"test_eligible_candidates.npz\"\n )\n eligibility = CandidateSet.read(public / \"data\" / candidate_name)\n if cache is not None and cache.is_file():\n prepared = load_cache(cache, stage, config)\n if not np.array_equal(prepared.item_ids, eligibility.item_ids):\n raise RuntimeError(\"cached catalog differs from candidate artifact\")\n return prepared, eligibility\n prepared, built_eligibility = prepare_data(public, stage, config)\n if cache is not None:\n save_cache(cache, prepared, stage, config)\n return prepared, built_eligibility\n\n\ndef build_item_correction(\n prepared: SequenceData,\n config: Config,\n target_starts: np.ndarray,\n target_counts: np.ndarray,\n) -> torch.Tensor:\n \"\"\"Estimate in-batch propensities from the selected training targets.\"\"\"\n counts = np.zeros(len(prepared.item_ids), dtype=np.float64)\n for user_row, count in enumerate(target_counts):\n if not count:\n continue\n offset = int(prepared.offsets[user_row])\n target_start = int(target_starts[user_row])\n counts += np.bincount(\n prepared.sequence_items[\n offset\n + target_start\n + 1 : offset\n + target_start\n + int(count)\n + 1\n ].astype(np.int64),\n minlength=len(prepared.item_ids),\n )\n total = float(counts.sum())\n probabilities = (counts + 1.0) / (total + len(counts))\n raw = np.power(\n (1.0 / len(counts)) / probabilities,\n config.inbatch_correction_power,\n )\n normalization = float(np.sum(counts * raw) / max(total, 1.0))\n correction = np.clip(\n raw / max(normalization, 1.0e-12),\n config.inbatch_correction_min,\n config.inbatch_correction_max,\n ).astype(np.float32)\n mean_on_targets = float(np.sum(counts * correction) / max(total, 1.0))\n log(\n f\"inbatch correction: min={correction.min():.3f} \"\n f\"mean_on_targets={mean_on_targets:.3f} max={correction.max():.3f}\"\n )\n return torch.from_numpy(correction)\n\n\ndef build_known_positive_filter(\n prepared: SequenceData, training_counts: np.ndarray\n) -> np.ndarray:\n \"\"\"Pack each user's complete public-training positive-item set.\"\"\"\n width = (len(prepared.item_ids) + 7) // 8\n packed = np.zeros((len(prepared.user_ids), width), dtype=np.uint8)\n retained = 0\n for user_row, target_count in enumerate(training_counts):\n if not target_count:\n continue\n offset = int(prepared.offsets[user_row])\n # T adjacent targets consume sequence positions 0 through T. The\n # chronological BCE suffix starts at T + 1 and is intentionally absent.\n items = np.unique(\n prepared.sequence_items[\n offset : offset + int(target_count) + 1\n ].astype(np.int64)\n )\n byte_columns = items >> 3\n bit_values = np.left_shift(\n np.uint8(1), (items & 7).astype(np.uint8)\n )\n np.bitwise_or.at(packed[user_row], byte_columns, bit_values)\n retained += len(items)\n log(\n f\"full-history false-negative filter: users={len(prepared.user_ids):,} \"\n f\"unique_user_items={retained:,} bytes={packed.nbytes:,}\"\n )\n return packed\n\n\ndef invalid_full_history_negatives(\n negatives: np.ndarray,\n targets: np.ndarray,\n user_rows: np.ndarray,\n known_positive_packed: np.ndarray,\n liked_items: np.ndarray,\n) -> np.ndarray:\n \"\"\"Return row-level invalid negatives using the complete training prefix.\"\"\"\n byte_columns = negatives >> 3\n bit_columns = negatives & 7\n known = known_positive_packed[user_rows[:, None], byte_columns]\n invalid = ((known >> bit_columns) & 1).astype(bool)\n invalid |= negatives == targets[:, None]\n invalid |= (negatives[:, :, None] == liked_items[:, None, :]).any(axis=2)\n for column in range(1, negatives.shape[1]):\n invalid[:, column] |= (\n negatives[:, column, None] == negatives[:, :column]\n ).any(axis=1)\n return invalid\n\n\ndef draw_full_history_filtered_negatives(\n targets: torch.Tensor,\n user_rows: np.ndarray,\n liked_items: torch.Tensor,\n known_positive_packed: np.ndarray,\n n_items: int,\n config: Config,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n \"\"\"Draw mixed negatives and reject all known public-training positives.\"\"\"\n targets_cpu = targets.detach().cpu().numpy().astype(np.int64, copy=False)\n liked_cpu = liked_items.detach().cpu().numpy().astype(np.int64, copy=False)\n batch_size = len(targets_cpu)\n uniform = torch.randint(\n n_items,\n (batch_size, config.uniform_negatives),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.empty((batch_size, 0), dtype=torch.long)\n corrected_count = 0\n if config.inbatch_negatives:\n unique_targets = torch.unique(targets.detach().cpu())\n corrected_count = min(\n config.inbatch_negatives, max(len(unique_targets) - 1, 0)\n )\n if corrected_count:\n priorities = torch.rand(\n (batch_size, len(unique_targets)), generator=generator\n )\n priorities.masked_fill_(\n unique_targets[None, :].eq(\n torch.from_numpy(targets_cpu)[:, None]\n ),\n -1.0,\n )\n columns = torch.topk(\n priorities, k=corrected_count, dim=1\n ).indices\n inbatch = unique_targets[columns]\n if corrected_count < config.inbatch_negatives:\n filler = torch.randint(\n n_items,\n (batch_size, config.inbatch_negatives - corrected_count),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.cat([inbatch, filler], dim=1)\n\n negatives = torch.cat([uniform, inbatch], dim=1).numpy()\n correction = np.ones_like(negatives, dtype=np.float32)\n if corrected_count:\n correction[\n :, config.uniform_negatives : config.uniform_negatives + corrected_count\n ] = item_correction[inbatch[:, :corrected_count]].numpy()\n for _ in range(32):\n invalid = invalid_full_history_negatives(\n negatives,\n targets_cpu,\n user_rows,\n known_positive_packed,\n liked_cpu,\n )\n if not np.any(invalid):\n return (\n torch.from_numpy(negatives).to(device),\n torch.from_numpy(correction).to(device),\n )\n replacements = torch.randint(\n n_items,\n (int(invalid.sum()),),\n generator=generator,\n device=\"cpu\",\n ).numpy()\n negatives[invalid] = replacements\n # Rejected in-batch items become ordinary uniform draws.\n correction[invalid] = 1.0\n invalid = invalid_full_history_negatives(\n negatives,\n targets_cpu,\n user_rows,\n known_positive_packed,\n liked_cpu,\n )\n if np.any(invalid):\n raise RuntimeError(\"full-history negative resampling failed\")\n return (\n torch.from_numpy(negatives).to(device),\n torch.from_numpy(correction).to(device),\n )\n\n\ndef draw_training_negatives(\n positive_items: torch.Tensor,\n known_history: torch.Tensor,\n liked_items: torch.Tensor,\n user_rows: np.ndarray,\n known_positive_packed: np.ndarray | None,\n prepared: SequenceData,\n config: Config,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n if config.negative_filter_scope == \"full-training-history\":\n if known_positive_packed is None:\n raise RuntimeError(\"full-history negative filter was not built\")\n return draw_full_history_filtered_negatives(\n positive_items,\n user_rows,\n liked_items,\n known_positive_packed,\n len(prepared.item_ids),\n config,\n item_correction,\n generator,\n device,\n )\n return base.draw_filtered_mixed_negatives(\n positive_items,\n known_history,\n liked_items,\n len(prepared.item_ids),\n config.uniform_negatives,\n config.inbatch_negatives,\n item_correction,\n generator,\n device,\n )\n\n\ndef event_sequence(\n model: base.SequentialTwoTower,\n items: torch.Tensor,\n features: torch.Tensor,\n hidden: torch.Tensor,\n) -> tuple[torch.Tensor, torch.Tensor]:\n inputs = torch.cat([model.context_embedding(items), features], dim=-1)\n projected = model.event_projection(inputs)\n return model.sequence_encoder(projected, hidden)\n\n\ndef pool_explicit_states(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n state_ids: np.ndarray,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor, torch.Tensor]:\n unique_states, inverse = np.unique(state_ids, return_inverse=True)\n liked_items = torch.from_numpy(\n prepared.liked_states[unique_states].astype(np.int64)\n ).to(device)\n disliked_items = torch.from_numpy(\n prepared.disliked_states[unique_states].astype(np.int64)\n ).to(device)\n liked_vectors = model._mean_context_channel(liked_items)\n disliked_vectors = model._mean_context_channel(disliked_items)\n inverse_rows = torch.from_numpy(inverse.astype(np.int64)).to(device)\n counts = prepared.state_counts[state_ids]\n return (\n liked_vectors[inverse_rows],\n disliked_vectors[inverse_rows],\n torch.from_numpy(counts.astype(np.float32)).to(device),\n )\n\n\ndef user_queries(\n model: base.SequentialTwoTower,\n sequence_vectors: torch.Tensor,\n liked_vectors: torch.Tensor,\n disliked_vectors: torch.Tensor,\n user_features: torch.Tensor,\n) -> torch.Tensor:\n return model.user_tower(\n torch.cat(\n [sequence_vectors, liked_vectors, disliked_vectors, user_features],\n dim=1,\n )\n )\n\n\ndef candidate_vectors(\n model: base.SequentialTwoTower,\n positives: torch.Tensor,\n negatives: torch.Tensor,\n) -> tuple[torch.Tensor, torch.Tensor]:\n all_items = torch.cat([positives, negatives.reshape(-1)])\n unique_items, inverse = torch.unique(all_items, return_inverse=True)\n encoded = model.encode_items(unique_items)\n positive_vectors = encoded[inverse[: len(positives)]]\n negative_vectors = encoded[inverse[len(positives) :]].reshape(\n len(positives), negatives.shape[1], -1\n )\n return positive_vectors, negative_vectors\n\n\ndef history_windows(\n sequence: np.ndarray,\n start: int,\n count: int,\n width: int,\n padding_item: int,\n) -> np.ndarray:\n base_start = max(0, start - width + 1)\n segment = sequence[base_start : start + count]\n padded = np.concatenate(\n [np.full(width - 1, padding_item, dtype=np.uint16), segment]\n )\n windows = np.lib.stride_tricks.sliding_window_view(padded, width)\n offset = start - base_start\n return windows[offset : offset + count].copy()\n\n\ndef corrected_negative_loss(\n logits: torch.Tensor,\n correction: torch.Tensor,\n config: Config,\n) -> torch.Tensor:\n raw = F.softplus(logits)\n total = config.uniform_negatives + config.inbatch_negatives\n result = torch.zeros(len(logits), device=logits.device)\n if config.uniform_negatives:\n result += config.uniform_negatives * raw[\n :, : config.uniform_negatives\n ].mean(dim=1)\n if config.inbatch_negatives:\n selected = slice(config.uniform_negatives, total)\n weights = correction[:, selected]\n result += config.inbatch_negatives * (\n (raw[:, selected] * weights).sum(dim=1)\n / weights.sum(dim=1).clamp_min(1.0e-6)\n )\n return result / total\n\n\ndef retained_target_window_starts(\n total_targets: np.ndarray,\n validation_counts: np.ndarray,\n training_counts: np.ndarray,\n) -> np.ndarray:\n \"\"\"Return the discarded prefix before the retained chronological suffix.\"\"\"\n training_region_ends = total_targets - validation_counts\n window_starts = training_region_ends - training_counts\n if (\n np.any(window_starts < 0)\n or np.any(training_counts < 0)\n or np.any(validation_counts < 0)\n ):\n raise RuntimeError(\"chronological target partition produced negative counts\")\n if not np.array_equal(\n window_starts + training_counts + validation_counts, total_targets\n ):\n raise RuntimeError(\"chronological target window is not a final contiguous suffix\")\n return window_starts\n\n\ndef chronological_target_partition(\n prepared: SequenceData,\n validation_fraction: float,\n max_train_targets_per_user: int,\n max_validation_targets_per_user: int,\n) -> TargetPartition:\n \"\"\"Select a recent contiguous window with a final out-of-time BCE suffix.\"\"\"\n totals = np.maximum(prepared.lengths - 1, 0).astype(np.int64)\n validation = np.floor(totals * validation_fraction).astype(np.int64)\n validation[(totals >= 2) & (validation == 0)] = 1\n if max_validation_targets_per_user:\n validation = np.minimum(validation, max_validation_targets_per_user)\n\n history_training = totals - validation\n training = history_training.copy()\n if max_train_targets_per_user:\n training = np.minimum(training, max_train_targets_per_user)\n window_starts = retained_target_window_starts(totals, validation, training)\n return TargetPartition(\n window_starts=window_starts,\n training_counts=training,\n validation_counts=validation,\n history_training_counts=history_training,\n )\n\n\ndef load_organic_like_outcomes(\n public: Path, stage: str, item_ids: np.ndarray\n) -> dict[int, dict[int, np.ndarray]]:\n \"\"\"Load mapped organic-like timestamps from public fitting history.\"\"\"\n prefixes = (\"train\",) if stage == \"validation\" else (\"train\", \"val\")\n frames = [\n pd.read_parquet(\n public / \"data\" / f\"{prefix}_likes.parquet\",\n columns=[\"uid\", \"timestamp\", \"item_id\", \"is_organic\"],\n )\n for prefix in prefixes\n ]\n likes = pd.concat(frames, ignore_index=True)\n likes = likes[likes.is_organic == 1].copy()\n mapped = pd.Index(item_ids).get_indexer(\n likes.item_id.to_numpy(dtype=np.int64)\n )\n likes = likes.loc[mapped >= 0, [\"uid\", \"timestamp\"]].copy()\n likes[\"item\"] = mapped[mapped >= 0].astype(np.int32, copy=False)\n likes = likes.sort_values([\"uid\", \"item\", \"timestamp\"], kind=\"stable\")\n nested: dict[int, dict[int, list[int]]] = {}\n for raw_uid, raw_timestamp, raw_item in likes.itertuples(\n index=False, name=None\n ):\n nested.setdefault(int(raw_uid), {}).setdefault(int(raw_item), []).append(\n int(raw_timestamp)\n )\n result = {\n uid: {\n item: np.asarray(timestamps, dtype=np.int64)\n for item, timestamps in items.items()\n }\n for uid, items in nested.items()\n }\n log(\n f\"loaded organic-like outcomes: rows={len(likes):,} \"\n f\"users={len(result):,}\"\n )\n return result\n\n\ndef build_target_weights(\n public: Path,\n stage: str,\n prepared: SequenceData,\n training_counts: np.ndarray,\n config: Config,\n) -> np.ndarray:\n \"\"\"Build nonzero per-token weights without dropping adjacent targets.\"\"\"\n if config.target_weighting == \"uniform\":\n weights = np.ones(len(prepared.sequence_items), dtype=np.float32)\n else:\n weights = np.full(\n len(prepared.sequence_items),\n config.ordinary_target_weight,\n dtype=np.float32,\n )\n organic_strong = (\n (prepared.event_features[:, 0] > 0.5)\n & (prepared.event_features[:, 1] >= 0.8)\n )\n weights[organic_strong] = config.organic_strong_target_weight\n if config.target_weighting == \"discovery-graded\":\n like_outcomes = load_organic_like_outcomes(\n public, stage, prepared.item_ids\n )\n history_end = (\n base.TRAIN_END if stage == \"validation\" else base.VALIDATION_END\n )\n horizon = config.discovery_horizon_days * base.DAY\n grade_three = 0\n for user_row, uid in enumerate(prepared.user_ids):\n start = int(prepared.offsets[user_row])\n end = int(prepared.offsets[user_row + 1])\n if end - start < 2:\n continue\n last_training_target = start + int(training_counts[user_row])\n first_validation_target = last_training_target + 1\n split_timestamp = (\n int(prepared.sequence_timestamps[first_validation_target])\n if first_validation_target < end\n else history_end\n )\n user_likes = like_outcomes.get(int(uid), {})\n seen = {int(prepared.sequence_items[start])}\n for position in range(start + 1, end):\n item = int(prepared.sequence_items[position])\n novel = item not in seen\n seen.add(item)\n if not novel:\n # Repeats remain training examples, but they cannot be\n # eligible discovery outcomes at the corresponding\n # cutoff and therefore keep only the ordinary weight.\n weights[position] = config.ordinary_target_weight\n continue\n timestamps = user_likes.get(item)\n if timestamps is None:\n continue\n timestamp = int(prepared.sequence_timestamps[position])\n upper = min(\n timestamp + horizon,\n split_timestamp\n if position <= last_training_target\n else history_end,\n )\n outcome = int(np.searchsorted(timestamps, timestamp, side=\"left\"))\n if outcome < len(timestamps) and int(timestamps[outcome]) < upper:\n weights[position] = config.organic_like_target_weight\n grade_three += 1\n log(\n f\"grade-3 first-encounter targets={grade_three:,} \"\n f\"horizon_days={config.discovery_horizon_days}\"\n )\n log(\n f\"target weighting={config.target_weighting} \"\n f\"min={weights.min():.3f} mean={weights.mean():.3f} \"\n f\"max={weights.max():.3f}\"\n )\n return weights\n\n\ndef build_dense_block(\n prepared: SequenceData,\n target_weights: np.ndarray,\n users: np.ndarray,\n window_starts: np.ndarray,\n process_counts: np.ndarray,\n score_starts: np.ndarray,\n start: int,\n steps: int,\n config: Config,\n) -> tuple[np.ndarray, ...]:\n \"\"\"Build one causal sequence block and mark the target rows to score.\"\"\"\n batch_size = len(users)\n padding_item = len(prepared.item_ids)\n input_items = np.full((batch_size, steps), padding_item, dtype=np.uint16)\n event_features = np.zeros((batch_size, steps, 3), dtype=np.float16)\n activity_features = np.zeros((batch_size, steps, 4), dtype=np.float16)\n state_ids = np.zeros((batch_size, steps), dtype=np.uint32)\n targets = np.zeros((batch_size, steps), dtype=np.uint16)\n weights = np.ones((batch_size, steps), dtype=np.float32)\n scored = np.zeros((batch_size, steps), dtype=bool)\n histories = np.full(\n (batch_size, steps, config.bptt_steps),\n padding_item,\n dtype=np.uint16,\n )\n processed = np.zeros(batch_size, dtype=np.int32)\n\n lengths = prepared.lengths\n for row, user_row in enumerate(users):\n count = min(steps, max(int(process_counts[row]) - start, 0))\n if not count:\n continue\n offset = int(prepared.offsets[user_row])\n window_start = int(window_starts[row])\n absolute_start = window_start + start\n positions = slice(\n offset + absolute_start,\n offset + absolute_start + count,\n )\n target_positions = slice(\n offset + absolute_start + 1,\n offset + absolute_start + count + 1,\n )\n input_items[row, :count] = prepared.sequence_items[positions]\n event_features[row, :count] = prepared.event_features[positions]\n activity_features[row, :count] = prepared.activity_features[positions]\n state_ids[row, :count] = prepared.state_ids[positions]\n targets[row, :count] = prepared.sequence_items[target_positions]\n weights[row, :count] = target_weights[target_positions]\n score_from = min(count, max(int(score_starts[row]) - start, 0))\n scored[row, score_from:count] = True\n processed[row] = count\n user_sequence = prepared.sequence_items[\n offset : offset + int(lengths[user_row])\n ]\n histories[row, :count] = history_windows(\n user_sequence,\n absolute_start,\n count,\n config.bptt_steps,\n padding_item,\n )\n return (\n input_items,\n event_features,\n activity_features,\n state_ids,\n targets,\n scored,\n histories,\n processed,\n weights,\n )\n\n\ndef score_target_microbatches(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n sequence_vectors: torch.Tensor,\n flat_states: np.ndarray,\n flat_activity: np.ndarray,\n positive_items: torch.Tensor,\n negatives: torch.Tensor,\n negative_weights: torch.Tensor,\n sample_weights: np.ndarray,\n device: torch.device,\n backward: bool,\n) -> tuple[dict[str, float], torch.Tensor | None]:\n \"\"\"Compute one logical block's exact BCE in bounded-memory slices.\"\"\"\n target_count = len(positive_items)\n total_weight = float(sample_weights.sum())\n if total_weight <= 0:\n raise RuntimeError(\"target block has no positive sample weight\")\n sequence_gradient = torch.zeros_like(sequence_vectors) if backward else None\n statistics = {\n \"loss_sum\": 0.0,\n \"positive_sum\": 0.0,\n \"negative_sum\": 0.0,\n \"pair_correct\": 0.0,\n \"weight_sum\": 0.0,\n }\n for begin in range(0, target_count, config.loss_microbatch_size):\n end = min(begin + config.loss_microbatch_size, target_count)\n selected_states = flat_states[begin:end]\n if backward:\n sequence_input = (\n sequence_vectors[begin:end].detach().requires_grad_(True)\n )\n else:\n sequence_input = sequence_vectors[begin:end]\n liked_vectors, disliked_vectors, _ = pool_explicit_states(\n model, prepared, selected_states, device\n )\n counts = prepared.state_counts[selected_states]\n numeric = np.zeros((end - begin, 6), dtype=np.float32)\n numeric[:, :4] = flat_activity[begin:end]\n numeric[:, 4] = np.log1p(counts[:, 0]) / 6.0\n numeric[:, 5] = np.log1p(counts[:, 1]) / 4.0\n query_vectors = user_queries(\n model,\n sequence_input,\n liked_vectors,\n disliked_vectors,\n torch.from_numpy(numeric).to(device),\n )\n positive_vectors, negative_vectors = candidate_vectors(\n model,\n positive_items[begin:end],\n negatives[begin:end],\n )\n positive_logits = (query_vectors * positive_vectors).sum(dim=1)\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n positive_loss = F.softplus(-positive_logits)\n negative_loss = corrected_negative_loss(\n negative_logits, negative_weights[begin:end], config\n )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n loss_sum = (\n selected_weights * (positive_loss + negative_loss) / 2.0\n ).sum()\n if not bool(torch.isfinite(loss_sum)):\n raise RuntimeError(\"dense next-item training produced non-finite loss\")\n if backward:\n (loss_sum / total_weight).backward()\n if sequence_input.grad is None:\n raise RuntimeError(\"sequence microbatch did not produce a gradient\")\n sequence_gradient[begin:end] = sequence_input.grad\n statistics[\"loss_sum\"] += float(loss_sum.detach().cpu())\n statistics[\"positive_sum\"] += float(positive_logits.detach().sum().cpu())\n statistics[\"negative_sum\"] += float(negative_logits.detach().sum().cpu())\n statistics[\"pair_correct\"] += float(\n (positive_logits.detach()[:, None] > negative_logits.detach()).sum().cpu()\n )\n statistics[\"weight_sum\"] += float(selected_weights.sum().detach().cpu())\n return statistics, sequence_gradient\n\n\ndef release_accelerator_cache(device: torch.device) -> None:\n if device.type == \"mps\":\n torch.mps.empty_cache()\n\n\ndef run_training_epoch(\n model: base.SequentialTwoTower,\n optimizer: torch.optim.Optimizer,\n prepared: SequenceData,\n config: Config,\n target_weights: np.ndarray,\n window_starts: np.ndarray,\n target_counts: np.ndarray,\n item_correction: torch.Tensor,\n known_positive_packed: np.ndarray | None,\n generator: torch.Generator,\n device: torch.device,\n epoch: int,\n) -> dict[str, float]:\n model.train()\n train_users = np.flatnonzero(target_counts > 0)\n sorted_users = train_users[\n np.argsort(target_counts[train_users], kind=\"stable\")\n ]\n groups = [\n sorted_users[start : start + config.user_batch_size]\n for start in range(0, len(sorted_users), config.user_batch_size)\n ]\n group_order = torch.randperm(len(groups), generator=generator).numpy()\n negative_count = config.uniform_negatives + config.inbatch_negatives\n total_loss = 0.0\n total_targets = 0\n total_weight = 0.0\n total_positive = 0.0\n total_negative = 0.0\n total_pair_correct = 0.0\n batches = 0\n started = time.monotonic()\n\n for group_number in group_order:\n users = groups[int(group_number)]\n group_counts = target_counts[users]\n hidden = torch.zeros(\n 1, len(users), config.embedding_dim, device=device\n )\n max_targets = int(group_counts.max())\n for start in range(0, max_targets, config.bptt_steps):\n steps = min(config.bptt_steps, max_targets - start)\n block = build_dense_block(\n prepared,\n target_weights,\n users,\n window_starts[users],\n group_counts,\n np.zeros(len(users), dtype=np.int64),\n start,\n steps,\n config,\n )\n sequence_outputs, _ = event_sequence(\n model,\n torch.from_numpy(block[0].astype(np.int64)).to(device),\n torch.from_numpy(block[1].astype(np.float32)).to(device),\n hidden,\n )\n hidden = torch.stack(\n [\n sequence_outputs[row, count - 1]\n if count\n else hidden[0, row]\n for row, count in enumerate(block[7])\n ]\n ).unsqueeze(0).detach()\n flat_scored = block[5].reshape(-1)\n if not np.any(flat_scored):\n continue\n mask = torch.from_numpy(flat_scored).to(device)\n sequence_vectors = sequence_outputs.reshape(\n -1, config.embedding_dim\n )[mask]\n flat_states = block[3].reshape(-1)[flat_scored]\n positive_items = torch.from_numpy(\n block[4].reshape(-1)[flat_scored].astype(np.int64)\n ).to(device)\n known_history = torch.from_numpy(\n block[6].reshape(-1, config.bptt_steps)[flat_scored].astype(\n np.int64\n )\n ).to(device)\n liked_items = torch.from_numpy(\n prepared.liked_states[flat_states].astype(np.int64)\n ).to(device)\n flat_user_rows = np.broadcast_to(\n users[:, None], block[5].shape\n ).reshape(-1)[flat_scored]\n negatives, negative_weights = draw_training_negatives(\n positive_items,\n known_history,\n liked_items,\n flat_user_rows,\n known_positive_packed,\n prepared,\n config,\n item_correction,\n generator,\n device,\n )\n optimizer.zero_grad(set_to_none=True)\n statistics, sequence_gradient = score_target_microbatches(\n model,\n prepared,\n config,\n sequence_vectors,\n flat_states,\n block[2].reshape(-1, 4)[flat_scored].astype(np.float32),\n positive_items,\n negatives,\n negative_weights,\n block[8].reshape(-1)[flat_scored],\n device,\n backward=True,\n )\n if sequence_gradient is None:\n raise RuntimeError(\"training block omitted its sequence gradient\")\n sequence_vectors.backward(sequence_gradient)\n nn.utils.clip_grad_norm_(model.parameters(), config.gradient_clip)\n optimizer.step()\n\n count = len(positive_items)\n total_loss += statistics[\"loss_sum\"]\n total_targets += count\n total_weight += statistics[\"weight_sum\"]\n total_positive += statistics[\"positive_sum\"]\n total_negative += statistics[\"negative_sum\"]\n total_pair_correct += statistics[\"pair_correct\"]\n batches += 1\n if batches % 25 == 0:\n release_accelerator_cache(device)\n if batches % 250 == 0:\n batch_bce = statistics[\"loss_sum\"] / max(\n statistics[\"weight_sum\"], 1.0e-12\n )\n running_bce = total_loss / max(total_weight, 1.0e-12)\n log(\n f\"epoch={epoch}/{config.epochs} batches={batches:,} \"\n f\"targets={total_targets:,} \"\n f\"batch_bce={batch_bce:.6f} \"\n f\"running_bce={running_bce:.6f} \"\n f\"elapsed={time.monotonic() - started:.1f}s\"\n )\n\n return {\n \"epoch\": float(epoch),\n \"loss\": total_loss / max(total_weight, 1.0e-12),\n \"positive_logit\": total_positive / max(total_targets, 1),\n \"negative_logit\": total_negative / max(total_targets * negative_count, 1),\n \"pair_accuracy\": total_pair_correct\n / max(total_targets * negative_count, 1),\n \"targets\": float(total_targets),\n \"target_weight_sum\": total_weight,\n \"batches\": float(batches),\n \"seconds\": time.monotonic() - started,\n }\n\n\n@torch.no_grad()\ndef evaluate_validation_bce(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n target_weights: np.ndarray,\n window_starts: np.ndarray,\n training_counts: np.ndarray,\n validation_counts: np.ndarray,\n item_correction: torch.Tensor,\n known_positive_packed: np.ndarray | None,\n device: torch.device,\n) -> dict[str, float]:\n \"\"\"Measure deterministic BCE on each user's chronological target suffix.\"\"\"\n model.eval()\n totals = training_counts + validation_counts\n users_with_validation = np.flatnonzero(validation_counts > 0)\n ordered = users_with_validation[\n np.argsort(totals[users_with_validation], kind=\"stable\")\n ]\n groups = [\n ordered[start : start + config.user_batch_size]\n for start in range(0, len(ordered), config.user_batch_size)\n ]\n generator = torch.Generator(device=\"cpu\").manual_seed(config.seed + 20_000)\n total_loss = 0.0\n total_targets = 0\n total_weight = 0.0\n started = time.monotonic()\n\n for group_number, users in enumerate(groups, start=1):\n group_totals = totals[users]\n group_starts = training_counts[users]\n hidden = torch.zeros(\n 1, len(users), config.embedding_dim, device=device\n )\n max_targets = int(group_totals.max())\n for start in range(0, max_targets, config.bptt_steps):\n steps = min(config.bptt_steps, max_targets - start)\n block = build_dense_block(\n prepared,\n target_weights,\n users,\n window_starts[users],\n group_totals,\n group_starts,\n start,\n steps,\n config,\n )\n sequence_outputs, _ = event_sequence(\n model,\n torch.from_numpy(block[0].astype(np.int64)).to(device),\n torch.from_numpy(block[1].astype(np.float32)).to(device),\n hidden,\n )\n hidden = torch.stack(\n [\n sequence_outputs[row, count - 1]\n if count\n else hidden[0, row]\n for row, count in enumerate(block[7])\n ]\n ).unsqueeze(0)\n flat_scored = block[5].reshape(-1)\n if not np.any(flat_scored):\n continue\n mask = torch.from_numpy(flat_scored).to(device)\n sequence_vectors = sequence_outputs.reshape(\n -1, config.embedding_dim\n )[mask]\n flat_states = block[3].reshape(-1)[flat_scored]\n positive_items = torch.from_numpy(\n block[4].reshape(-1)[flat_scored].astype(np.int64)\n ).to(device)\n known_history = torch.from_numpy(\n block[6].reshape(-1, config.bptt_steps)[flat_scored].astype(\n np.int64\n )\n ).to(device)\n liked_items = torch.from_numpy(\n prepared.liked_states[flat_states].astype(np.int64)\n ).to(device)\n flat_user_rows = np.broadcast_to(\n users[:, None], block[5].shape\n ).reshape(-1)[flat_scored]\n negatives, negative_weights = draw_training_negatives(\n positive_items,\n known_history,\n liked_items,\n flat_user_rows,\n known_positive_packed,\n prepared,\n config,\n item_correction,\n generator,\n device,\n )\n statistics, _ = score_target_microbatches(\n model,\n prepared,\n config,\n sequence_vectors,\n flat_states,\n block[2].reshape(-1, 4)[flat_scored].astype(np.float32),\n positive_items,\n negatives,\n negative_weights,\n block[8].reshape(-1)[flat_scored],\n device,\n backward=False,\n )\n total_loss += statistics[\"loss_sum\"]\n total_targets += len(positive_items)\n total_weight += statistics[\"weight_sum\"]\n if group_number % 25 == 0:\n release_accelerator_cache(device)\n return {\n \"validation_loss\": total_loss / max(total_weight, 1.0e-12),\n \"validation_targets\": float(total_targets),\n \"validation_target_weight_sum\": total_weight,\n \"validation_seconds\": time.monotonic() - started,\n }\n\n\ndef snapshot_model(model: nn.Module) -> dict[str, torch.Tensor]:\n return {\n name: value.detach().cpu().clone()\n for name, value in model.state_dict().items()\n }\n\n\ndef save_epoch_checkpoint(\n checkpoint_dir: Path,\n epoch: int,\n model_state: dict[str, torch.Tensor],\n config: Config,\n diagnostics: dict[str, float],\n optimizer: torch.optim.Optimizer,\n generator: torch.Generator,\n) -> Path:\n checkpoint_dir.mkdir(parents=True, exist_ok=True)\n path = checkpoint_dir / f\"epoch-{epoch:02d}.pt\"\n torch.save(\n {\n \"epoch\": epoch,\n \"config\": asdict(config),\n \"diagnostics\": diagnostics,\n \"model_state\": model_state,\n \"optimizer_state\": optimizer.state_dict(),\n \"generator_state\": generator.get_state(),\n },\n path,\n )\n log(f\"saved epoch checkpoint: {path}\")\n return path\n\n\ndef train_model(\n public: Path,\n stage: str,\n prepared: SequenceData,\n config: Config,\n device: torch.device,\n checkpoint_dir: Path | None,\n resume_checkpoint: Path | None,\n) -> tuple[base.SequentialTwoTower, list[dict[str, float]]]:\n partition = chronological_target_partition(\n prepared,\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n training_counts = partition.training_counts\n validation_counts = partition.validation_counts\n log(\n f\"chronological BCE window: train={int(training_counts.sum()):,} \"\n f\"validation={int(validation_counts.sum()):,} \"\n f\"selected={partition.selected_count:,} \"\n f\"discarded_old_targets={int(partition.window_starts.sum()):,} \"\n f\"all={prepared.target_count:,}\"\n )\n model = base.SequentialTwoTower(prepared.item_features, model_config(config)).to(\n device\n )\n optimizer = torch.optim.AdamW(\n model.parameters(), lr=config.learning_rate, weight_decay=config.weight_decay\n )\n generator = torch.Generator(device=\"cpu\").manual_seed(config.seed + 10_000)\n item_correction = build_item_correction(\n prepared,\n config,\n partition.window_starts,\n training_counts,\n )\n target_weights = build_target_weights(\n public,\n stage,\n prepared,\n partition.history_training_counts,\n config,\n )\n known_positive_packed = (\n build_known_positive_filter(prepared, partition.history_training_counts)\n if config.negative_filter_scope == \"full-training-history\"\n else None\n )\n history: list[dict[str, float]] = []\n best_loss = float(\"inf\")\n best_epoch = 0\n best_state: dict[str, torch.Tensor] | None = None\n epochs_without_improvement = 0\n start_epoch = 0\n\n if resume_checkpoint is not None:\n payload = torch.load(\n resume_checkpoint, map_location=device, weights_only=False\n )\n stored_config = payload.get(\"config\", {})\n architecture_fields = (\n \"bptt_steps\",\n \"embedding_dim\",\n \"hidden_dim\",\n \"user_hidden_1\",\n \"user_hidden_2\",\n \"item_hidden_1\",\n \"item_hidden_2\",\n )\n mismatches = {\n name: (stored_config.get(name), getattr(config, name))\n for name in architecture_fields\n if stored_config.get(name) != getattr(config, name)\n }\n if mismatches:\n raise RuntimeError(\n f\"resume checkpoint architecture differs: {mismatches}\"\n )\n stored_filter_scope = stored_config.get(\n \"negative_filter_scope\", \"last-100\"\n )\n stored_target_weighting = stored_config.get(\n \"target_weighting\", \"uniform\"\n )\n stored_weight_values = (\n stored_config.get(\"ordinary_target_weight\", 1.0),\n stored_config.get(\"organic_strong_target_weight\", 1.0),\n stored_config.get(\"organic_like_target_weight\", 3.0),\n stored_config.get(\"discovery_horizon_days\", 7),\n )\n current_weight_values = (\n config.ordinary_target_weight,\n config.organic_strong_target_weight,\n config.organic_like_target_weight,\n config.discovery_horizon_days,\n )\n stored_window_values = (\n stored_config.get(\"validation_fraction\", 0.10),\n stored_config.get(\"max_train_targets_per_user\", 0),\n stored_config.get(\"max_validation_targets_per_user\", 0),\n )\n current_window_values = (\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n objective_changed = (\n stored_filter_scope != config.negative_filter_scope\n or stored_target_weighting != config.target_weighting\n or stored_weight_values != current_weight_values\n or stored_window_values != current_window_values\n )\n model.load_state_dict(payload[\"model_state\"])\n if \"optimizer_state\" in payload and not objective_changed:\n optimizer.load_state_dict(payload[\"optimizer_state\"])\n for group in optimizer.param_groups:\n group[\"lr\"] = config.learning_rate\n group[\"weight_decay\"] = config.weight_decay\n log(f\"restored optimizer state from {resume_checkpoint}\")\n else:\n log(\n f\"using a fresh AdamW optimizer after loading {resume_checkpoint}\"\n )\n if \"generator_state\" in payload and not objective_changed:\n generator.set_state(payload[\"generator_state\"])\n start_epoch = int(payload[\"epoch\"])\n if start_epoch >= config.epochs:\n raise RuntimeError(\n f\"resume epoch {start_epoch} must be below --epochs {config.epochs}\"\n )\n prior = {\n str(name): float(value)\n for name, value in payload.get(\"diagnostics\", {}).items()\n }\n if objective_changed:\n prior.pop(\"validation_loss\", None)\n log(\n \"BCE objective changed: filter \"\n f\"{stored_filter_scope}->{config.negative_filter_scope}, \"\n \"target weighting \"\n f\"{stored_target_weighting}->{config.target_weighting}, \"\n f\"target window {stored_window_values}->{current_window_values}; \"\n \"recomputing the resume checkpoint's validation BCE\"\n )\n if \"validation_loss\" not in prior:\n prior.update(\n evaluate_validation_bce(\n model,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n validation_counts,\n item_correction,\n known_positive_packed,\n device,\n )\n )\n history.append(prior)\n best_loss = float(prior[\"validation_loss\"])\n best_epoch = start_epoch\n best_state = snapshot_model(model)\n log(\n f\"resumed validation-BCE training: epoch={start_epoch} \"\n f\"loss={best_loss:.6f}\"\n )\n\n for epoch in range(start_epoch + 1, config.epochs + 1):\n diagnostics = run_training_epoch(\n model,\n optimizer,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n item_correction,\n known_positive_packed,\n generator,\n device,\n epoch,\n )\n validation = evaluate_validation_bce(\n model,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n validation_counts,\n item_correction,\n known_positive_packed,\n device,\n )\n diagnostics.update(validation)\n history.append(diagnostics)\n current_state = snapshot_model(model)\n if checkpoint_dir is not None:\n save_epoch_checkpoint(\n checkpoint_dir,\n epoch,\n current_state,\n config,\n diagnostics,\n optimizer,\n generator,\n )\n improved = (\n diagnostics[\"validation_loss\"]\n < best_loss - config.early_stopping_min_delta\n )\n if improved:\n best_loss = diagnostics[\"validation_loss\"]\n best_epoch = epoch\n best_state = current_state\n epochs_without_improvement = 0\n if checkpoint_dir is not None:\n torch.save(\n {\n \"epoch\": epoch,\n \"config\": asdict(config),\n \"diagnostics\": diagnostics,\n \"model_state\": current_state,\n \"optimizer_state\": optimizer.state_dict(),\n \"generator_state\": generator.get_state(),\n },\n checkpoint_dir / \"best.pt\",\n )\n else:\n epochs_without_improvement += 1\n log(\n f\"epoch={epoch}/{config.epochs} train_loss={diagnostics['loss']:.5f} \"\n f\"validation_loss={diagnostics['validation_loss']:.5f} \"\n f\"pair_accuracy={diagnostics['pair_accuracy']:.3f} \"\n f\"train_targets={int(diagnostics['targets']):,} \"\n f\"validation_targets={int(diagnostics['validation_targets']):,} \"\n f\"train_seconds={diagnostics['seconds']:.1f} \"\n f\"validation_seconds={diagnostics['validation_seconds']:.1f}\"\n )\n release_accelerator_cache(device)\n if epochs_without_improvement >= config.early_stopping_patience:\n log(\n f\"early stopping after epoch={epoch}; \"\n f\"best_epoch={best_epoch} best_validation_loss={best_loss:.6f}\"\n )\n break\n\n if best_state is None:\n raise RuntimeError(\"training did not produce a valid checkpoint\")\n model.load_state_dict(best_state)\n model.to(device).eval()\n log(\n f\"restored best validation-BCE checkpoint: epoch={best_epoch} \"\n f\"loss={best_loss:.6f}\"\n )\n return model, history\n\n\n@torch.inference_mode()\ndef inference_queries(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n device: torch.device,\n) -> torch.Tensor:\n sequence_rows = prepared.inference_user_rows\n lengths = np.where(\n sequence_rows >= 0, prepared.lengths[np.maximum(sequence_rows, 0)], 0\n )\n ordered = np.argsort(lengths, kind=\"stable\")\n result = torch.zeros(\n len(sequence_rows), config.embedding_dim, dtype=torch.float32\n )\n for group_start in range(0, len(ordered), config.user_batch_size):\n output_rows = ordered[group_start : group_start + config.user_batch_size]\n rows = sequence_rows[output_rows]\n group_lengths = lengths[output_rows]\n hidden = torch.zeros(\n 1, len(output_rows), config.embedding_dim, device=device\n )\n for start in range(0, int(group_lengths.max(initial=0)), config.bptt_steps):\n steps = min(\n config.bptt_steps,\n int(group_lengths.max(initial=0)) - start,\n )\n items = np.full(\n (len(output_rows), steps),\n len(prepared.item_ids),\n dtype=np.uint16,\n )\n features = np.zeros((len(output_rows), steps, 3), dtype=np.float16)\n counts = np.zeros(len(output_rows), dtype=np.int32)\n for local_row, sequence_row in enumerate(rows):\n if sequence_row < 0:\n continue\n count = min(steps, max(int(group_lengths[local_row]) - start, 0))\n if not count:\n continue\n offset = int(prepared.offsets[sequence_row])\n selected = slice(offset + start, offset + start + count)\n items[local_row, :count] = prepared.sequence_items[selected]\n features[local_row, :count] = prepared.event_features[selected]\n counts[local_row] = count\n outputs, _ = event_sequence(\n model,\n torch.from_numpy(items.astype(np.int64)).to(device),\n torch.from_numpy(features.astype(np.float32)).to(device),\n hidden,\n )\n next_rows: list[torch.Tensor] = []\n for local_row, count in enumerate(counts):\n next_rows.append(\n outputs[local_row, count - 1]\n if count\n else hidden[0, local_row]\n )\n hidden = torch.stack(next_rows).unsqueeze(0)\n\n liked_items = torch.from_numpy(\n prepared.inference_liked_items[output_rows].astype(np.int64)\n ).to(device)\n disliked_items = torch.from_numpy(\n prepared.inference_disliked_items[output_rows].astype(np.int64)\n ).to(device)\n queries = user_queries(\n model,\n hidden.squeeze(0),\n model._mean_context_channel(liked_items),\n model._mean_context_channel(disliked_items),\n torch.from_numpy(\n prepared.inference_user_features[output_rows].astype(np.float32)\n ).to(device),\n )\n result[torch.from_numpy(output_rows.astype(np.int64))] = queries.cpu()\n return result\n\n\n@torch.inference_mode()\ndef rank_candidates(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n eligibility: CandidateSet,\n config: Config,\n device: torch.device,\n ranking_batch_size: int,\n) -> tuple[pd.DataFrame, dict[str, float]]:\n model.eval()\n queries = inference_queries(model, prepared, config, device)\n item_parts: list[torch.Tensor] = []\n for start in range(0, len(prepared.item_ids), 2_048):\n rows = torch.arange(\n start,\n min(start + 2_048, len(prepared.item_ids)),\n dtype=torch.long,\n device=device,\n )\n item_parts.append(model.encode_items(rows).cpu())\n items = torch.cat(item_parts)\n rows: list[tuple[int, int, int]] = []\n for start in range(0, len(eligibility.uids), ranking_batch_size):\n end = min(start + ranking_batch_size, len(eligibility.uids))\n scores = queries[start:end] @ items.T\n eligible = np.unpackbits(\n eligibility.packed[start:end],\n axis=1,\n count=len(prepared.item_ids),\n bitorder=\"little\",\n ).astype(bool, copy=False)\n scores.masked_fill_(~torch.from_numpy(eligible), -torch.inf)\n selected = torch.topk(scores, k=base.TOP_N, dim=1).indices.numpy()\n for local_row, columns in enumerate(selected):\n uid = int(eligibility.uids[start + local_row])\n rows.extend(\n (uid, int(prepared.item_ids[column]), rank)\n for rank, column in enumerate(columns, start=1)\n )\n if end == len(eligibility.uids) or end % 1_024 == 0:\n log(f\"ranked users={end:,}/{len(eligibility.uids):,}\")\n query_norm = torch.linalg.vector_norm(queries, dim=1).numpy()\n item_norm = torch.linalg.vector_norm(items, dim=1).numpy()\n diagnostics = {\n \"query_norm_mean\": float(query_norm.mean()),\n \"query_norm_std\": float(query_norm.std()),\n \"item_norm_mean\": float(item_norm.mean()),\n \"item_norm_std\": float(item_norm.std()),\n }\n return pd.DataFrame(rows, columns=[\"uid\", \"item_id\", \"rank\"]), diagnostics\n\n\ndef write_ranking(frame: pd.DataFrame, output: Path) -> None:\n output.parent.mkdir(parents=True, exist_ok=True)\n if output.suffix == \".parquet\":\n frame.to_parquet(output, index=False, compression=\"zstd\")\n elif output.suffix == \".csv\":\n frame.to_csv(output, index=False)\n else:\n raise ValueError(\"--output must end in .parquet or .csv\")\n\n\ndef main() -> int:\n parser = argparse.ArgumentParser(description=__doc__)\n parser.add_argument(\n \"--stage\", choices=(\"validation\", \"submission\"), default=\"validation\"\n )\n parser.add_argument(\"--output\", type=Path, required=True)\n parser.add_argument(\"--public-dir\", type=Path, required=True)\n parser.add_argument(\"--prepared-cache\", type=Path)\n parser.add_argument(\"--device\", choices=(\"cpu\", \"mps\"), default=\"cpu\")\n parser.add_argument(\"--threads\", type=int, default=8)\n parser.add_argument(\"--bptt-steps\", type=int, default=Config.bptt_steps)\n parser.add_argument(\n \"--user-batch-size\", type=int, default=Config.user_batch_size\n )\n parser.add_argument(\"--epochs\", type=int, default=Config.epochs)\n parser.add_argument(\"--learning-rate\", type=float, default=Config.learning_rate)\n parser.add_argument(\n \"--loss-microbatch-size\",\n type=int,\n default=Config.loss_microbatch_size,\n )\n parser.add_argument(\n \"--negative-filter-scope\",\n choices=(\"full-training-history\", \"last-100\"),\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n )\n parser.add_argument(\n \"--ordinary-target-weight\",\n type=float,\n default=Config.ordinary_target_weight,\n )\n parser.add_argument(\n \"--organic-strong-target-weight\",\n type=float,\n default=Config.organic_strong_target_weight,\n )\n parser.add_argument(\n \"--organic-like-target-weight\",\n type=float,\n default=Config.organic_like_target_weight,\n )\n parser.add_argument(\n \"--discovery-horizon-days\",\n type=int,\n default=Config.discovery_horizon_days,\n )\n parser.add_argument(\n \"--validation-fraction\",\n type=float,\n default=Config.validation_fraction,\n )\n parser.add_argument(\n \"--max-train-targets-per-user\",\n type=int,\n default=Config.max_train_targets_per_user,\n help=\"retain this many recent pre-validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--max-validation-targets-per-user\",\n type=int,\n default=Config.max_validation_targets_per_user,\n help=\"retain this many final validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--early-stopping-patience\",\n type=int,\n default=Config.early_stopping_patience,\n )\n parser.add_argument(\n \"--early-stopping-min-delta\",\n type=float,\n default=Config.early_stopping_min_delta,\n )\n parser.add_argument(\"--checkpoint-dir\", type=Path)\n parser.add_argument(\"--resume-checkpoint\", type=Path)\n parser.add_argument(\"--seed\", type=int, default=Config.seed)\n parser.add_argument(\"--ranking-batch-size\", type=int, default=64)\n parser.add_argument(\"--prepare-only\", action=\"store_true\")\n args = parser.parse_args()\n\n if args.threads < 1 or args.bptt_steps < 2 or args.user_batch_size < 1:\n parser.error(\"threads, BPTT steps, and user batch size must be positive\")\n if (\n args.epochs < 1\n or args.learning_rate <= 0\n or args.loss_microbatch_size < 1\n or args.early_stopping_patience < 1\n or args.early_stopping_min_delta < 0\n or args.ordinary_target_weight <= 0\n or args.organic_strong_target_weight <= 0\n or args.organic_like_target_weight <= 0\n or args.discovery_horizon_days < 1\n or args.max_train_targets_per_user < 0\n or args.max_validation_targets_per_user < 0\n ):\n parser.error(\"training counts, learning rate, and patience are invalid\")\n if not 0.0 < args.validation_fraction < 1.0:\n parser.error(\"validation fraction must be strictly between zero and one\")\n public = args.public_dir.resolve()\n if not (public / \"data\" / \"candidate_catalog.parquet\").is_file():\n parser.error(f\"public task data not found under {public}\")\n config = Config(\n bptt_steps=args.bptt_steps,\n user_batch_size=args.user_batch_size,\n epochs=args.epochs,\n learning_rate=args.learning_rate,\n loss_microbatch_size=args.loss_microbatch_size,\n negative_filter_scope=args.negative_filter_scope,\n target_weighting=args.target_weighting,\n ordinary_target_weight=args.ordinary_target_weight,\n organic_strong_target_weight=args.organic_strong_target_weight,\n organic_like_target_weight=args.organic_like_target_weight,\n discovery_horizon_days=args.discovery_horizon_days,\n validation_fraction=args.validation_fraction,\n max_train_targets_per_user=args.max_train_targets_per_user,\n max_validation_targets_per_user=args.max_validation_targets_per_user,\n early_stopping_patience=args.early_stopping_patience,\n early_stopping_min_delta=args.early_stopping_min_delta,\n seed=args.seed,\n )\n torch.set_num_threads(args.threads)\n torch.set_num_interop_threads(1)\n seed_everything(config.seed)\n device = torch.device(args.device)\n if device.type == \"mps\" and not torch.backends.mps.is_available():\n parser.error(\"MPS was requested but is unavailable\")\n\n started = time.monotonic()\n prepared, eligibility = load_or_prepare(\n public,\n args.stage,\n config,\n args.prepared_cache.resolve() if args.prepared_cache else None,\n )\n target_partition = chronological_target_partition(\n prepared,\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n if args.prepare_only:\n print(\n json.dumps(\n {\n \"stage\": args.stage,\n \"sequence_users\": len(prepared.user_ids),\n \"sequence_tokens\": len(prepared.sequence_items),\n \"next_item_targets\": prepared.target_count,\n \"selected_training_targets\": int(\n target_partition.training_counts.sum()\n ),\n \"selected_validation_targets\": int(\n target_partition.validation_counts.sum()\n ),\n \"discarded_old_targets\": int(\n target_partition.window_starts.sum()\n ),\n \"target_users\": len(eligibility.uids),\n \"runtime_seconds\": round(time.monotonic() - started, 3),\n \"config\": asdict(config),\n },\n indent=2,\n )\n )\n return 0\n\n model, training_history = train_model(\n public,\n args.stage,\n prepared,\n config,\n device,\n args.checkpoint_dir.resolve() if args.checkpoint_dir else None,\n args.resume_checkpoint.resolve() if args.resume_checkpoint else None,\n )\n ranking, embedding_diagnostics = rank_candidates(\n model,\n prepared,\n eligibility,\n config,\n device,\n args.ranking_batch_size,\n )\n write_ranking(ranking, args.output.resolve())\n print(\n json.dumps(\n {\n \"baseline\": \"yambda-windowed-sequence-two-tower-base\",\n \"stage\": args.stage,\n \"output\": str(args.output.resolve()),\n \"rows\": len(ranking),\n \"users\": int(ranking.uid.nunique()),\n \"runtime_seconds\": round(time.monotonic() - started, 3),\n \"public_inputs_only\": True,\n \"architecture\": {\n \"family\": \"pure-windowed-sequence-two-tower\",\n \"sequence_training\": (\n \"every-adjacent-target-within-recent-chronological-window\"\n ),\n \"sequence_encoder\": \"one-layer-GRU-dense-timestep-loss\",\n \"bptt_hidden_state\": \"carried-and-detached\",\n \"model_selection\": \"chronological-validation-BCE-early-stopping\",\n \"loss_execution\": \"gradient-accumulated-microbatches\",\n \"user_id_embedding\": False,\n \"user_mlp\": \"582-512-256-192\",\n \"item_mlp\": \"196-512-256-192\",\n \"normalization\": \"none\",\n \"score\": \"raw-dot-product\",\n \"objective\": \"sample-weighted-BCE-at-every-timestep\",\n \"negative_sampling\": (\n \"24-uniform-plus-8-propensity-corrected-inbatch\"\n ),\n \"false_negative_filter\": config.negative_filter_scope,\n \"target_weighting\": config.target_weighting,\n },\n \"config\": asdict(config),\n \"sequence_tokens\": len(prepared.sequence_items),\n \"all_next_item_targets\": prepared.target_count,\n \"selected_training_targets\": int(\n target_partition.training_counts.sum()\n ),\n \"selected_validation_targets\": int(\n target_partition.validation_counts.sum()\n ),\n \"discarded_old_targets\": int(target_partition.window_starts.sum()),\n \"training_history\": training_history,\n \"embedding_diagnostics\": embedding_diagnostics,\n \"torch_version\": torch.__version__,\n \"device\": str(device),\n },\n indent=2,\n )\n )\n return 0\n\n\nif __name__ == \"__main__\":\n raise SystemExit(main())\n",
"newContent": "#!/usr/bin/env python3\n\"\"\"Train the Yambda windowed two-tower baseline.\n\nEvery catalog listen with at least 50 percent completion is a sequence token.\nWithin each selected chronological window, every adjacent target contributes a\nloss. The fast default retains each user's most recent 128 training targets and\n16 held-out validation targets; passing zero for either cap restores the\nuncapped behavior for that side of the split. A 100-step truncated-BPTT block\ncomputes all of its timestep losses and carries the detached GRU hidden state\ninto the next block.\n\nThe model is a pure two-tower retriever: no user-ID embedding, no output\nnormalization, raw dot-product retrieval, sample-weighted binary cross entropy,\nand 24 uniform plus eight propensity-corrected unique in-batch negatives. Every\nselected adjacent target retains a nonzero weight. Only files under the supplied\npublic directory are read.\n\"\"\"\nfrom __future__ import annotations\n\nimport argparse\nimport json\nimport random\nimport sys\nimport time\nfrom dataclasses import asdict, dataclass\nfrom pathlib import Path\n\nimport numpy as np\nimport pandas as pd\nimport torch\nfrom torch import nn\nfrom torch.nn import functional as F\n\nimport two_tower_model as base\n\n\nCACHE_VERSION = 2\n\n\n@dataclass(frozen=True)\nclass Config:\n bptt_steps: int = 100\n user_batch_size: int = 64\n embedding_dim: int = 192\n hidden_dim: int = 192\n user_hidden_1: int = 512\n user_hidden_2: int = 256\n item_hidden_1: int = 512\n item_hidden_2: int = 256\n dropout: float = 0.05\n epochs: int = 1\n learning_rate: float = 1.0e-3\n weight_decay: float = 1.0e-5\n gradient_clip: float = 2.0\n loss_microbatch_size: int = 65_536\n uniform_negatives: int = 24\n inbatch_negatives: int = 8\n inbatch_correction_power: float = 0.5\n inbatch_correction_min: float = 0.25\n inbatch_correction_max: float = 4.0\n negative_filter_scope: str = \"full-training-history\"\n objective: str = \"bce\"\n temperature: float = 0.2\n target_weighting: str = \"organic-strong\"\n ordinary_target_weight: float = 0.25\n organic_strong_target_weight: float = 1.0\n organic_like_target_weight: float = 3.0\n discovery_horizon_days: int = 7\n validation_fraction: float = 0.10\n max_train_targets_per_user: int = 128\n max_validation_targets_per_user: int = 16\n early_stopping_patience: int = 1\n early_stopping_min_delta: float = 1.0e-4\n seed: int = 29\n\n\n@dataclass(frozen=True)\nclass CandidateSet:\n uids: np.ndarray\n item_ids: np.ndarray\n packed: np.ndarray\n\n @classmethod\n def read(cls, path: Path) -> \"CandidateSet\":\n with np.load(path, allow_pickle=False) as artifact:\n result = cls(\n uids=artifact[\"uids\"].astype(np.int64, copy=True),\n item_ids=artifact[\"item_ids\"].astype(np.int64, copy=True),\n packed=artifact[\"packed_eligible\"].astype(np.uint8, copy=True),\n )\n expected = (len(result.uids), (len(result.item_ids) + 7) // 8)\n if result.packed.shape != expected:\n raise ValueError(\n f\"invalid candidate artifact shape {result.packed.shape}; \"\n f\"expected {expected}\"\n )\n return result\n\n\n@dataclass(frozen=True)\nclass SequenceData:\n item_ids: np.ndarray\n item_features: np.ndarray\n user_ids: np.ndarray\n offsets: np.ndarray\n sequence_timestamps: np.ndarray\n sequence_items: np.ndarray\n event_features: np.ndarray\n activity_features: np.ndarray\n state_ids: np.ndarray\n liked_states: np.ndarray\n disliked_states: np.ndarray\n state_counts: np.ndarray\n inference_user_rows: np.ndarray\n inference_liked_items: np.ndarray\n inference_disliked_items: np.ndarray\n inference_user_features: np.ndarray\n\n @property\n def lengths(self) -> np.ndarray:\n return np.diff(self.offsets)\n\n @property\n def target_count(self) -> int:\n return int(np.maximum(self.lengths - 1, 0).sum())\n\n\n@dataclass(frozen=True)\nclass TargetPartition:\n \"\"\"A contiguous recent target window and its chronological BCE split.\"\"\"\n\n window_starts: np.ndarray\n training_counts: np.ndarray\n validation_counts: np.ndarray\n history_training_counts: np.ndarray\n\n @property\n def selected_count(self) -> int:\n return int(self.training_counts.sum() + self.validation_counts.sum())\n\n\ndef log(message: str) -> None:\n print(message, file=sys.stderr, flush=True)\n\n\ndef seed_everything(seed: int) -> None:\n random.seed(seed)\n np.random.seed(seed)\n torch.manual_seed(seed)\n\n\ndef model_config(config: Config) -> base.TowerConfig:\n \"\"\"Map the dense trainer configuration onto the unchanged tower.\"\"\"\n return base.TowerConfig(\n sequence_length=config.bptt_steps,\n embedding_dim=config.embedding_dim,\n hidden_dim=config.hidden_dim,\n user_hidden_1=config.user_hidden_1,\n user_hidden_2=config.user_hidden_2,\n item_hidden_1=config.item_hidden_1,\n item_hidden_2=config.item_hidden_2,\n dropout=config.dropout,\n epochs=config.epochs,\n learning_rate=config.learning_rate,\n weight_decay=config.weight_decay,\n gradient_clip=config.gradient_clip,\n uniform_negatives=config.uniform_negatives,\n inbatch_negatives=config.inbatch_negatives,\n inbatch_correction_power=config.inbatch_correction_power,\n inbatch_correction_min=config.inbatch_correction_min,\n inbatch_correction_max=config.inbatch_correction_max,\n seed=config.seed,\n )\n\n\ndef append_state_snapshot(\n active_likes: dict[int, int],\n active_dislikes: dict[int, int],\n n_items: int,\n liked_states: list[np.ndarray],\n disliked_states: list[np.ndarray],\n state_counts: list[tuple[int, int]],\n) -> int:\n liked_states.append(base.pack_active_items(active_likes, base.LIKE_ITEMS, n_items))\n disliked_states.append(\n base.pack_active_items(active_dislikes, base.DISLIKE_ITEMS, n_items)\n )\n state_counts.append((len(active_likes), len(active_dislikes)))\n return len(liked_states) - 1\n\n\ndef causal_activity_features(\n timestamps: np.ndarray,\n items: np.ndarray,\n organic: np.ndarray,\n completion: np.ndarray,\n) -> np.ndarray:\n \"\"\"Compute the first four reviewed user features at every event cutoff.\"\"\"\n length = len(items)\n result = np.zeros((length, 4), dtype=np.float16)\n if not length:\n return result\n left_rows = np.searchsorted(\n timestamps,\n timestamps - base.PROFILE_LOOKBACK_DAYS * base.DAY,\n side=\"left\",\n )\n rows = np.arange(length, dtype=np.int64)\n counts = rows - left_rows + 1\n organic_sum = np.cumsum(organic.astype(np.float64))\n completion_sum = np.cumsum(np.minimum(completion, 100).astype(np.float64))\n before_organic = np.where(\n left_rows > 0, organic_sum[np.maximum(left_rows - 1, 0)], 0.0\n )\n before_completion = np.where(\n left_rows > 0, completion_sum[np.maximum(left_rows - 1, 0)], 0.0\n )\n\n unique_counts = np.empty(length, dtype=np.int32)\n frequencies: dict[int, int] = {}\n left = 0\n for right, raw_item in enumerate(items):\n item = int(raw_item)\n frequencies[item] = frequencies.get(item, 0) + 1\n required_left = int(left_rows[right])\n while left < required_left:\n old_item = int(items[left])\n remaining = frequencies[old_item] - 1\n if remaining:\n frequencies[old_item] = remaining\n else:\n del frequencies[old_item]\n left += 1\n unique_counts[right] = len(frequencies)\n\n result[:, 0] = (np.log1p(counts) / 8.0).astype(np.float16)\n result[:, 1] = (np.log1p(unique_counts) / 6.0).astype(np.float16)\n result[:, 2] = (\n (organic_sum - before_organic) / np.maximum(counts, 1)\n ).astype(np.float16)\n result[:, 3] = (\n (completion_sum - before_completion)\n / np.maximum(counts, 1)\n / 100.0\n ).astype(np.float16)\n return result\n\n\ndef build_sequence_parts(\n listens: base.ListenEvents,\n explicit: base.ExplicitEvents,\n n_items: int,\n liked_states: list[np.ndarray],\n disliked_states: list[np.ndarray],\n state_counts: list[tuple[int, int]],\n) -> tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray, np.ndarray] | None:\n qualifying = listens.completion_pct >= base.LISTEN_POSITIVE_THRESHOLD\n timestamps = listens.timestamps[qualifying].astype(np.int64, copy=False)\n items = listens.items[qualifying].astype(np.uint16, copy=False)\n organic = listens.organic[qualifying].astype(np.float32, copy=False)\n completion = listens.completion_pct[qualifying].astype(np.float32, copy=False)\n if not len(items):\n return None\n\n event_features = np.zeros((len(items), 3), dtype=np.float16)\n event_features[:, 0] = organic.astype(np.float16)\n event_features[:, 1] = (\n np.minimum(completion, 100.0) / 100.0\n ).astype(np.float16)\n if len(items) > 1:\n gaps = np.maximum(0, timestamps[1:] - timestamps[:-1])\n event_features[1:, 2] = np.exp(\n -gaps.astype(np.float64) / (60.0 * base.DAY)\n ).astype(np.float16)\n\n activity = causal_activity_features(\n timestamps, items, organic, completion\n )\n states = np.zeros(len(items), dtype=np.uint32)\n active_likes: dict[int, int] = {}\n active_dislikes: dict[int, int] = {}\n explicit_position = 0\n current_state = 0\n for row, timestamp in enumerate(timestamps):\n changed = False\n while (\n explicit_position < len(explicit.timestamps)\n and int(explicit.timestamps[explicit_position]) < int(timestamp)\n ):\n item = int(explicit.items[explicit_position])\n event_timestamp = int(explicit.timestamps[explicit_position])\n destination = (\n active_likes\n if int(explicit.channel[explicit_position]) == 0\n else active_dislikes\n )\n if int(explicit.active[explicit_position]) == 1:\n destination[item] = event_timestamp\n else:\n destination.pop(item, None)\n explicit_position += 1\n changed = True\n if changed:\n current_state = append_state_snapshot(\n active_likes,\n active_dislikes,\n n_items,\n liked_states,\n disliked_states,\n state_counts,\n )\n states[row] = np.uint32(current_state)\n return timestamps, items.copy(), event_features, activity, states\n\n\ndef final_user_features(\n timestamps: np.ndarray,\n items: np.ndarray,\n event_features: np.ndarray,\n history_end: int,\n like_count: int,\n dislike_count: int,\n) -> np.ndarray:\n start = int(\n np.searchsorted(\n timestamps,\n history_end - base.PROFILE_LOOKBACK_DAYS * base.DAY,\n side=\"left\",\n )\n )\n result = np.zeros(6, dtype=np.float32)\n recent_items = items[start:]\n if len(recent_items):\n result[0] = np.log1p(len(recent_items)) / 8.0\n result[1] = np.log1p(len(np.unique(recent_items))) / 6.0\n result[2] = float(np.mean(event_features[start:, 0]))\n result[3] = float(np.mean(event_features[start:, 1]))\n result[4] = np.log1p(like_count) / 6.0\n result[5] = np.log1p(dislike_count) / 4.0\n return result\n\n\ndef prepare_data(\n public: Path, stage: str, config: Config\n) -> tuple[SequenceData, CandidateSet]:\n data = public / \"data\"\n catalog = pd.read_parquet(data / \"candidate_catalog.parquet\").sort_values(\n \"popularity_rank\"\n )\n item_ids = catalog.item_id.to_numpy(dtype=np.int64, copy=True)\n item_index = pd.Index(item_ids)\n n_items = len(item_ids)\n if n_items >= np.iinfo(np.uint16).max:\n raise RuntimeError(\"uint16 item storage is too small for this catalog\")\n\n if stage == \"validation\":\n prefixes = (\"train\",)\n history_end = base.TRAIN_END\n candidate_name = \"validation_eligible_candidates.npz\"\n state_name = \"validation_preference_state.parquet\"\n else:\n prefixes = (\"train\", \"val\")\n history_end = base.VALIDATION_END\n candidate_name = \"test_eligible_candidates.npz\"\n state_name = \"test_preference_state.parquet\"\n eligibility = CandidateSet.read(data / candidate_name)\n if not np.array_equal(eligibility.item_ids, item_ids):\n raise RuntimeError(\"catalog and candidate artifact item order differ\")\n\n explicit_by_user = base.load_explicit_events(data, prefixes, item_index)\n inference_likes, inference_dislikes, like_counts, dislike_counts = (\n base.load_inference_state(\n data / state_name,\n item_index,\n eligibility.uids,\n n_items,\n )\n )\n inference_rows = {int(uid): row for row, uid in enumerate(eligibility.uids)}\n inference_user_features = np.zeros((len(eligibility.uids), 6), dtype=np.float32)\n inference_sequence_rows = np.full(len(eligibility.uids), -1, dtype=np.int32)\n\n # Snapshot zero is the shared empty explicit state.\n liked_states = [np.full(base.LIKE_ITEMS, n_items, dtype=np.uint16)]\n disliked_states = [\n np.full(base.DISLIKE_ITEMS, n_items, dtype=np.uint16)\n ]\n state_counts: list[tuple[int, int]] = [(0, 0)]\n user_ids: list[int] = []\n offsets = [0]\n timestamp_parts: list[np.ndarray] = []\n item_parts: list[np.ndarray] = []\n event_parts: list[np.ndarray] = []\n activity_parts: list[np.ndarray] = []\n state_parts: list[np.ndarray] = []\n\n listen_paths = [data / f\"{prefix}_listens.parquet\" for prefix in prefixes]\n started = time.monotonic()\n for seen_users, (uid, listens) in enumerate(\n base.merge_listen_streams(listen_paths, item_index), start=1\n ):\n parts = build_sequence_parts(\n listens,\n explicit_by_user.get(uid, base.empty_explicit_events()),\n n_items,\n liked_states,\n disliked_states,\n state_counts,\n )\n if parts is None:\n continue\n sequence_row = len(user_ids)\n user_ids.append(uid)\n timestamp_parts.append(parts[0].astype(np.uint32, copy=False))\n item_parts.append(parts[1])\n event_parts.append(parts[2])\n activity_parts.append(parts[3])\n state_parts.append(parts[4])\n offsets.append(offsets[-1] + len(parts[1]))\n inference_row = inference_rows.get(uid)\n if inference_row is not None:\n inference_sequence_rows[inference_row] = sequence_row\n inference_user_features[inference_row] = final_user_features(\n parts[0],\n parts[1],\n parts[2],\n history_end,\n int(like_counts[inference_row]),\n int(dislike_counts[inference_row]),\n )\n if seen_users % 1_000 == 0:\n log(\n f\"prepared users={seen_users:,} tokens={offsets[-1]:,} \"\n f\"states={len(liked_states):,} \"\n f\"elapsed={time.monotonic() - started:.1f}s\"\n )\n\n missing = np.flatnonzero(inference_sequence_rows < 0)\n if len(missing):\n inference_user_features[missing, 4] = np.log1p(like_counts[missing]) / 6.0\n inference_user_features[missing, 5] = (\n np.log1p(dislike_counts[missing]) / 4.0\n )\n log(f\"target users without qualifying catalog sequence={len(missing):,}\")\n\n prepared = SequenceData(\n item_ids=item_ids,\n item_features=base.build_item_features(catalog),\n user_ids=np.asarray(user_ids, dtype=np.int64),\n offsets=np.asarray(offsets, dtype=np.int64),\n sequence_timestamps=np.concatenate(timestamp_parts),\n sequence_items=np.concatenate(item_parts),\n event_features=np.concatenate(event_parts),\n activity_features=np.concatenate(activity_parts),\n state_ids=np.concatenate(state_parts),\n liked_states=np.stack(liked_states),\n disliked_states=np.stack(disliked_states),\n state_counts=np.asarray(state_counts, dtype=np.int32),\n inference_user_rows=inference_sequence_rows,\n inference_liked_items=inference_likes,\n inference_disliked_items=inference_dislikes,\n inference_user_features=inference_user_features,\n )\n log(\n f\"prepared stage={stage} sequence_users={len(prepared.user_ids):,} \"\n f\"tokens={len(prepared.sequence_items):,} \"\n f\"next_item_targets={prepared.target_count:,} \"\n f\"states={len(prepared.liked_states):,}\"\n )\n return prepared, eligibility\n\n\ndef cache_metadata(stage: str, config: Config) -> dict[str, object]:\n return {\n \"cache_version\": CACHE_VERSION,\n \"stage\": stage,\n \"positive_listen_threshold\": base.LISTEN_POSITIVE_THRESHOLD,\n \"profile_lookback_days\": base.PROFILE_LOOKBACK_DAYS,\n \"event_time_feature\": \"exp-negative-gap-over-60-days\",\n }\n\n\ndef save_cache(\n path: Path, prepared: SequenceData, stage: str, config: Config\n) -> None:\n path.parent.mkdir(parents=True, exist_ok=True)\n np.savez(\n path,\n metadata=np.asarray(json.dumps(cache_metadata(stage, config), sort_keys=True)),\n **prepared.__dict__,\n )\n log(f\"saved public-derived full-sequence cache: {path}\")\n\n\ndef load_cache(path: Path, stage: str, config: Config) -> SequenceData:\n with np.load(path, allow_pickle=False) as artifact:\n actual = json.loads(str(artifact[\"metadata\"].item()))\n expected = cache_metadata(stage, config)\n if actual != expected:\n raise RuntimeError(f\"sequence cache differs: {actual} != {expected}\")\n values = {\n field: artifact[field].copy()\n for field in SequenceData.__dataclass_fields__\n }\n prepared = SequenceData(**values)\n log(\n f\"loaded full-sequence cache: tokens={len(prepared.sequence_items):,} \"\n f\"targets={prepared.target_count:,}\"\n )\n return prepared\n\n\ndef load_or_prepare(\n public: Path,\n stage: str,\n config: Config,\n cache: Path | None,\n) -> tuple[SequenceData, CandidateSet]:\n candidate_name = (\n \"validation_eligible_candidates.npz\"\n if stage == \"validation\"\n else \"test_eligible_candidates.npz\"\n )\n eligibility = CandidateSet.read(public / \"data\" / candidate_name)\n if cache is not None and cache.is_file():\n prepared = load_cache(cache, stage, config)\n if not np.array_equal(prepared.item_ids, eligibility.item_ids):\n raise RuntimeError(\"cached catalog differs from candidate artifact\")\n return prepared, eligibility\n prepared, built_eligibility = prepare_data(public, stage, config)\n if cache is not None:\n save_cache(cache, prepared, stage, config)\n return prepared, built_eligibility\n\n\ndef build_item_correction(\n prepared: SequenceData,\n config: Config,\n target_starts: np.ndarray,\n target_counts: np.ndarray,\n) -> torch.Tensor:\n \"\"\"Estimate in-batch propensities from the selected training targets.\"\"\"\n counts = np.zeros(len(prepared.item_ids), dtype=np.float64)\n for user_row, count in enumerate(target_counts):\n if not count:\n continue\n offset = int(prepared.offsets[user_row])\n target_start = int(target_starts[user_row])\n counts += np.bincount(\n prepared.sequence_items[\n offset\n + target_start\n + 1 : offset\n + target_start\n + int(count)\n + 1\n ].astype(np.int64),\n minlength=len(prepared.item_ids),\n )\n total = float(counts.sum())\n probabilities = (counts + 1.0) / (total + len(counts))\n raw = np.power(\n (1.0 / len(counts)) / probabilities,\n config.inbatch_correction_power,\n )\n normalization = float(np.sum(counts * raw) / max(total, 1.0))\n correction = np.clip(\n raw / max(normalization, 1.0e-12),\n config.inbatch_correction_min,\n config.inbatch_correction_max,\n ).astype(np.float32)\n mean_on_targets = float(np.sum(counts * correction) / max(total, 1.0))\n log(\n f\"inbatch correction: min={correction.min():.3f} \"\n f\"mean_on_targets={mean_on_targets:.3f} max={correction.max():.3f}\"\n )\n return torch.from_numpy(correction)\n\n\ndef build_known_positive_filter(\n prepared: SequenceData, training_counts: np.ndarray\n) -> np.ndarray:\n \"\"\"Pack each user's complete public-training positive-item set.\"\"\"\n width = (len(prepared.item_ids) + 7) // 8\n packed = np.zeros((len(prepared.user_ids), width), dtype=np.uint8)\n retained = 0\n for user_row, target_count in enumerate(training_counts):\n if not target_count:\n continue\n offset = int(prepared.offsets[user_row])\n # T adjacent targets consume sequence positions 0 through T. The\n # chronological BCE suffix starts at T + 1 and is intentionally absent.\n items = np.unique(\n prepared.sequence_items[\n offset : offset + int(target_count) + 1\n ].astype(np.int64)\n )\n byte_columns = items >> 3\n bit_values = np.left_shift(\n np.uint8(1), (items & 7).astype(np.uint8)\n )\n np.bitwise_or.at(packed[user_row], byte_columns, bit_values)\n retained += len(items)\n log(\n f\"full-history false-negative filter: users={len(prepared.user_ids):,} \"\n f\"unique_user_items={retained:,} bytes={packed.nbytes:,}\"\n )\n return packed\n\n\ndef invalid_full_history_negatives(\n negatives: np.ndarray,\n targets: np.ndarray,\n user_rows: np.ndarray,\n known_positive_packed: np.ndarray,\n liked_items: np.ndarray,\n) -> np.ndarray:\n \"\"\"Return row-level invalid negatives using the complete training prefix.\"\"\"\n byte_columns = negatives >> 3\n bit_columns = negatives & 7\n known = known_positive_packed[user_rows[:, None], byte_columns]\n invalid = ((known >> bit_columns) & 1).astype(bool)\n invalid |= negatives == targets[:, None]\n invalid |= (negatives[:, :, None] == liked_items[:, None, :]).any(axis=2)\n for column in range(1, negatives.shape[1]):\n invalid[:, column] |= (\n negatives[:, column, None] == negatives[:, :column]\n ).any(axis=1)\n return invalid\n\n\ndef draw_full_history_filtered_negatives(\n targets: torch.Tensor,\n user_rows: np.ndarray,\n liked_items: torch.Tensor,\n known_positive_packed: np.ndarray,\n n_items: int,\n config: Config,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n \"\"\"Draw mixed negatives and reject all known public-training positives.\"\"\"\n targets_cpu = targets.detach().cpu().numpy().astype(np.int64, copy=False)\n liked_cpu = liked_items.detach().cpu().numpy().astype(np.int64, copy=False)\n batch_size = len(targets_cpu)\n uniform = torch.randint(\n n_items,\n (batch_size, config.uniform_negatives),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.empty((batch_size, 0), dtype=torch.long)\n corrected_count = 0\n if config.inbatch_negatives:\n unique_targets = torch.unique(targets.detach().cpu())\n corrected_count = min(\n config.inbatch_negatives, max(len(unique_targets) - 1, 0)\n )\n if corrected_count:\n priorities = torch.rand(\n (batch_size, len(unique_targets)), generator=generator\n )\n priorities.masked_fill_(\n unique_targets[None, :].eq(\n torch.from_numpy(targets_cpu)[:, None]\n ),\n -1.0,\n )\n columns = torch.topk(\n priorities, k=corrected_count, dim=1\n ).indices\n inbatch = unique_targets[columns]\n if corrected_count < config.inbatch_negatives:\n filler = torch.randint(\n n_items,\n (batch_size, config.inbatch_negatives - corrected_count),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.cat([inbatch, filler], dim=1)\n\n negatives = torch.cat([uniform, inbatch], dim=1).numpy()\n correction = np.ones_like(negatives, dtype=np.float32)\n if corrected_count:\n correction[\n :, config.uniform_negatives : config.uniform_negatives + corrected_count\n ] = item_correction[inbatch[:, :corrected_count]].numpy()\n for _ in range(32):\n invalid = invalid_full_history_negatives(\n negatives,\n targets_cpu,\n user_rows,\n known_positive_packed,\n liked_cpu,\n )\n if not np.any(invalid):\n return (\n torch.from_numpy(negatives).to(device),\n torch.from_numpy(correction).to(device),\n )\n replacements = torch.randint(\n n_items,\n (int(invalid.sum()),),\n generator=generator,\n device=\"cpu\",\n ).numpy()\n negatives[invalid] = replacements\n # Rejected in-batch items become ordinary uniform draws.\n correction[invalid] = 1.0\n invalid = invalid_full_history_negatives(\n negatives,\n targets_cpu,\n user_rows,\n known_positive_packed,\n liked_cpu,\n )\n if np.any(invalid):\n raise RuntimeError(\"full-history negative resampling failed\")\n return (\n torch.from_numpy(negatives).to(device),\n torch.from_numpy(correction).to(device),\n )\n\n\ndef draw_training_negatives(\n positive_items: torch.Tensor,\n known_history: torch.Tensor,\n liked_items: torch.Tensor,\n user_rows: np.ndarray,\n known_positive_packed: np.ndarray | None,\n prepared: SequenceData,\n config: Config,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n if config.negative_filter_scope == \"full-training-history\":\n if known_positive_packed is None:\n raise RuntimeError(\"full-history negative filter was not built\")\n return draw_full_history_filtered_negatives(\n positive_items,\n user_rows,\n liked_items,\n known_positive_packed,\n len(prepared.item_ids),\n config,\n item_correction,\n generator,\n device,\n )\n return base.draw_filtered_mixed_negatives(\n positive_items,\n known_history,\n liked_items,\n len(prepared.item_ids),\n config.uniform_negatives,\n config.inbatch_negatives,\n item_correction,\n generator,\n device,\n )\n\n\ndef event_sequence(\n model: base.SequentialTwoTower,\n items: torch.Tensor,\n features: torch.Tensor,\n hidden: torch.Tensor,\n) -> tuple[torch.Tensor, torch.Tensor]:\n inputs = torch.cat([model.context_embedding(items), features], dim=-1)\n projected = model.event_projection(inputs)\n return model.sequence_encoder(projected, hidden)\n\n\ndef pool_explicit_states(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n state_ids: np.ndarray,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor, torch.Tensor]:\n unique_states, inverse = np.unique(state_ids, return_inverse=True)\n liked_items = torch.from_numpy(\n prepared.liked_states[unique_states].astype(np.int64)\n ).to(device)\n disliked_items = torch.from_numpy(\n prepared.disliked_states[unique_states].astype(np.int64)\n ).to(device)\n liked_vectors = model._mean_context_channel(liked_items)\n disliked_vectors = model._mean_context_channel(disliked_items)\n inverse_rows = torch.from_numpy(inverse.astype(np.int64)).to(device)\n counts = prepared.state_counts[state_ids]\n return (\n liked_vectors[inverse_rows],\n disliked_vectors[inverse_rows],\n torch.from_numpy(counts.astype(np.float32)).to(device),\n )\n\n\ndef user_queries(\n model: base.SequentialTwoTower,\n sequence_vectors: torch.Tensor,\n liked_vectors: torch.Tensor,\n disliked_vectors: torch.Tensor,\n user_features: torch.Tensor,\n) -> torch.Tensor:\n return model.user_tower(\n torch.cat(\n [sequence_vectors, liked_vectors, disliked_vectors, user_features],\n dim=1,\n )\n )\n\n\ndef candidate_vectors(\n model: base.SequentialTwoTower,\n positives: torch.Tensor,\n negatives: torch.Tensor,\n) -> tuple[torch.Tensor, torch.Tensor]:\n all_items = torch.cat([positives, negatives.reshape(-1)])\n unique_items, inverse = torch.unique(all_items, return_inverse=True)\n encoded = model.encode_items(unique_items)\n positive_vectors = encoded[inverse[: len(positives)]]\n negative_vectors = encoded[inverse[len(positives) :]].reshape(\n len(positives), negatives.shape[1], -1\n )\n return positive_vectors, negative_vectors\n\n\ndef history_windows(\n sequence: np.ndarray,\n start: int,\n count: int,\n width: int,\n padding_item: int,\n) -> np.ndarray:\n base_start = max(0, start - width + 1)\n segment = sequence[base_start : start + count]\n padded = np.concatenate(\n [np.full(width - 1, padding_item, dtype=np.uint16), segment]\n )\n windows = np.lib.stride_tricks.sliding_window_view(padded, width)\n offset = start - base_start\n return windows[offset : offset + count].copy()\n\n\ndef corrected_negative_loss(\n logits: torch.Tensor,\n correction: torch.Tensor,\n config: Config,\n) -> torch.Tensor:\n raw = F.softplus(logits)\n total = config.uniform_negatives + config.inbatch_negatives\n result = torch.zeros(len(logits), device=logits.device)\n if config.uniform_negatives:\n result += config.uniform_negatives * raw[\n :, : config.uniform_negatives\n ].mean(dim=1)\n if config.inbatch_negatives:\n selected = slice(config.uniform_negatives, total)\n weights = correction[:, selected]\n result += config.inbatch_negatives * (\n (raw[:, selected] * weights).sum(dim=1)\n / weights.sum(dim=1).clamp_min(1.0e-6)\n )\n return result / total\n\n\ndef retained_target_window_starts(\n total_targets: np.ndarray,\n validation_counts: np.ndarray,\n training_counts: np.ndarray,\n) -> np.ndarray:\n \"\"\"Return the discarded prefix before the retained chronological suffix.\"\"\"\n training_region_ends = total_targets - validation_counts\n window_starts = training_region_ends - training_counts\n if (\n np.any(window_starts < 0)\n or np.any(training_counts < 0)\n or np.any(validation_counts < 0)\n ):\n raise RuntimeError(\"chronological target partition produced negative counts\")\n if not np.array_equal(\n window_starts + training_counts + validation_counts, total_targets\n ):\n raise RuntimeError(\"chronological target window is not a final contiguous suffix\")\n return window_starts\n\n\ndef chronological_target_partition(\n prepared: SequenceData,\n validation_fraction: float,\n max_train_targets_per_user: int,\n max_validation_targets_per_user: int,\n) -> TargetPartition:\n \"\"\"Select a recent contiguous window with a final out-of-time BCE suffix.\"\"\"\n totals = np.maximum(prepared.lengths - 1, 0).astype(np.int64)\n validation = np.floor(totals * validation_fraction).astype(np.int64)\n validation[(totals >= 2) & (validation == 0)] = 1\n if max_validation_targets_per_user:\n validation = np.minimum(validation, max_validation_targets_per_user)\n\n history_training = totals - validation\n training = history_training.copy()\n if max_train_targets_per_user:\n training = np.minimum(training, max_train_targets_per_user)\n window_starts = retained_target_window_starts(totals, validation, training)\n return TargetPartition(\n window_starts=window_starts,\n training_counts=training,\n validation_counts=validation,\n history_training_counts=history_training,\n )\n\n\ndef load_organic_like_outcomes(\n public: Path, stage: str, item_ids: np.ndarray\n) -> dict[int, dict[int, np.ndarray]]:\n \"\"\"Load mapped organic-like timestamps from public fitting history.\"\"\"\n prefixes = (\"train\",) if stage == \"validation\" else (\"train\", \"val\")\n frames = [\n pd.read_parquet(\n public / \"data\" / f\"{prefix}_likes.parquet\",\n columns=[\"uid\", \"timestamp\", \"item_id\", \"is_organic\"],\n )\n for prefix in prefixes\n ]\n likes = pd.concat(frames, ignore_index=True)\n likes = likes[likes.is_organic == 1].copy()\n mapped = pd.Index(item_ids).get_indexer(\n likes.item_id.to_numpy(dtype=np.int64)\n )\n likes = likes.loc[mapped >= 0, [\"uid\", \"timestamp\"]].copy()\n likes[\"item\"] = mapped[mapped >= 0].astype(np.int32, copy=False)\n likes = likes.sort_values([\"uid\", \"item\", \"timestamp\"], kind=\"stable\")\n nested: dict[int, dict[int, list[int]]] = {}\n for raw_uid, raw_timestamp, raw_item in likes.itertuples(\n index=False, name=None\n ):\n nested.setdefault(int(raw_uid), {}).setdefault(int(raw_item), []).append(\n int(raw_timestamp)\n )\n result = {\n uid: {\n item: np.asarray(timestamps, dtype=np.int64)\n for item, timestamps in items.items()\n }\n for uid, items in nested.items()\n }\n log(\n f\"loaded organic-like outcomes: rows={len(likes):,} \"\n f\"users={len(result):,}\"\n )\n return result\n\n\ndef build_target_weights(\n public: Path,\n stage: str,\n prepared: SequenceData,\n training_counts: np.ndarray,\n config: Config,\n) -> np.ndarray:\n \"\"\"Build nonzero per-token weights without dropping adjacent targets.\"\"\"\n if config.target_weighting == \"uniform\":\n weights = np.ones(len(prepared.sequence_items), dtype=np.float32)\n else:\n weights = np.full(\n len(prepared.sequence_items),\n config.ordinary_target_weight,\n dtype=np.float32,\n )\n organic_strong = (\n (prepared.event_features[:, 0] > 0.5)\n & (prepared.event_features[:, 1] >= 0.8)\n )\n weights[organic_strong] = config.organic_strong_target_weight\n if config.target_weighting == \"discovery-graded\":\n like_outcomes = load_organic_like_outcomes(\n public, stage, prepared.item_ids\n )\n history_end = (\n base.TRAIN_END if stage == \"validation\" else base.VALIDATION_END\n )\n horizon = config.discovery_horizon_days * base.DAY\n grade_three = 0\n for user_row, uid in enumerate(prepared.user_ids):\n start = int(prepared.offsets[user_row])\n end = int(prepared.offsets[user_row + 1])\n if end - start < 2:\n continue\n last_training_target = start + int(training_counts[user_row])\n first_validation_target = last_training_target + 1\n split_timestamp = (\n int(prepared.sequence_timestamps[first_validation_target])\n if first_validation_target < end\n else history_end\n )\n user_likes = like_outcomes.get(int(uid), {})\n seen = {int(prepared.sequence_items[start])}\n for position in range(start + 1, end):\n item = int(prepared.sequence_items[position])\n novel = item not in seen\n seen.add(item)\n if not novel:\n # Repeats remain training examples, but they cannot be\n # eligible discovery outcomes at the corresponding\n # cutoff and therefore keep only the ordinary weight.\n weights[position] = config.ordinary_target_weight\n continue\n timestamps = user_likes.get(item)\n if timestamps is None:\n continue\n timestamp = int(prepared.sequence_timestamps[position])\n upper = min(\n timestamp + horizon,\n split_timestamp\n if position <= last_training_target\n else history_end,\n )\n outcome = int(np.searchsorted(timestamps, timestamp, side=\"left\"))\n if outcome < len(timestamps) and int(timestamps[outcome]) < upper:\n weights[position] = config.organic_like_target_weight\n grade_three += 1\n log(\n f\"grade-3 first-encounter targets={grade_three:,} \"\n f\"horizon_days={config.discovery_horizon_days}\"\n )\n log(\n f\"target weighting={config.target_weighting} \"\n f\"min={weights.min():.3f} mean={weights.mean():.3f} \"\n f\"max={weights.max():.3f}\"\n )\n return weights\n\n\ndef build_dense_block(\n prepared: SequenceData,\n target_weights: np.ndarray,\n users: np.ndarray,\n window_starts: np.ndarray,\n process_counts: np.ndarray,\n score_starts: np.ndarray,\n start: int,\n steps: int,\n config: Config,\n) -> tuple[np.ndarray, ...]:\n \"\"\"Build one causal sequence block and mark the target rows to score.\"\"\"\n batch_size = len(users)\n padding_item = len(prepared.item_ids)\n input_items = np.full((batch_size, steps), padding_item, dtype=np.uint16)\n event_features = np.zeros((batch_size, steps, 3), dtype=np.float16)\n activity_features = np.zeros((batch_size, steps, 4), dtype=np.float16)\n state_ids = np.zeros((batch_size, steps), dtype=np.uint32)\n targets = np.zeros((batch_size, steps), dtype=np.uint16)\n weights = np.ones((batch_size, steps), dtype=np.float32)\n scored = np.zeros((batch_size, steps), dtype=bool)\n histories = np.full(\n (batch_size, steps, config.bptt_steps),\n padding_item,\n dtype=np.uint16,\n )\n processed = np.zeros(batch_size, dtype=np.int32)\n\n lengths = prepared.lengths\n for row, user_row in enumerate(users):\n count = min(steps, max(int(process_counts[row]) - start, 0))\n if not count:\n continue\n offset = int(prepared.offsets[user_row])\n window_start = int(window_starts[row])\n absolute_start = window_start + start\n positions = slice(\n offset + absolute_start,\n offset + absolute_start + count,\n )\n target_positions = slice(\n offset + absolute_start + 1,\n offset + absolute_start + count + 1,\n )\n input_items[row, :count] = prepared.sequence_items[positions]\n event_features[row, :count] = prepared.event_features[positions]\n activity_features[row, :count] = prepared.activity_features[positions]\n state_ids[row, :count] = prepared.state_ids[positions]\n targets[row, :count] = prepared.sequence_items[target_positions]\n weights[row, :count] = target_weights[target_positions]\n score_from = min(count, max(int(score_starts[row]) - start, 0))\n scored[row, score_from:count] = True\n processed[row] = count\n user_sequence = prepared.sequence_items[\n offset : offset + int(lengths[user_row])\n ]\n histories[row, :count] = history_windows(\n user_sequence,\n absolute_start,\n count,\n config.bptt_steps,\n padding_item,\n )\n return (\n input_items,\n event_features,\n activity_features,\n state_ids,\n targets,\n scored,\n histories,\n processed,\n weights,\n )\n\n\ndef score_target_microbatches(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n sequence_vectors: torch.Tensor,\n flat_states: np.ndarray,\n flat_activity: np.ndarray,\n positive_items: torch.Tensor,\n negatives: torch.Tensor,\n negative_weights: torch.Tensor,\n sample_weights: np.ndarray,\n device: torch.device,\n backward: bool,\n) -> tuple[dict[str, float], torch.Tensor | None]:\n \"\"\"Compute one logical block's exact BCE in bounded-memory slices.\"\"\"\n target_count = len(positive_items)\n total_weight = float(sample_weights.sum())\n if total_weight <= 0:\n raise RuntimeError(\"target block has no positive sample weight\")\n sequence_gradient = torch.zeros_like(sequence_vectors) if backward else None\n statistics = {\n \"loss_sum\": 0.0,\n \"positive_sum\": 0.0,\n \"negative_sum\": 0.0,\n \"pair_correct\": 0.0,\n \"weight_sum\": 0.0,\n }\n for begin in range(0, target_count, config.loss_microbatch_size):\n end = min(begin + config.loss_microbatch_size, target_count)\n selected_states = flat_states[begin:end]\n if backward:\n sequence_input = (\n sequence_vectors[begin:end].detach().requires_grad_(True)\n )\n else:\n sequence_input = sequence_vectors[begin:end]\n liked_vectors, disliked_vectors, _ = pool_explicit_states(\n model, prepared, selected_states, device\n )\n counts = prepared.state_counts[selected_states]\n numeric = np.zeros((end - begin, 6), dtype=np.float32)\n numeric[:, :4] = flat_activity[begin:end]\n numeric[:, 4] = np.log1p(counts[:, 0]) / 6.0\n numeric[:, 5] = np.log1p(counts[:, 1]) / 4.0\n query_vectors = user_queries(\n model,\n sequence_input,\n liked_vectors,\n disliked_vectors,\n torch.from_numpy(numeric).to(device),\n )\n positive_vectors, negative_vectors = candidate_vectors(\n model,\n positive_items[begin:end],\n negatives[begin:end],\n )\n positive_logits = (query_vectors * positive_vectors).sum(dim=1)\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n if config.objective == \"infonce\":\n cat_logits = torch.cat(\n [positive_logits.unsqueeze(1), negative_logits], dim=1\n ) / config.temperature\n labels = torch.zeros(\n len(positive_logits), dtype=torch.long, device=device\n )\n raw_loss = F.cross_entropy(cat_logits, labels, reduction=\"none\")\n loss_sum = (selected_weights * raw_loss).sum()\n else:\n positive_loss = F.softplus(-positive_logits)\n negative_loss = corrected_negative_loss(\n negative_logits, negative_weights[begin:end], config\n )\n loss_sum = (\n selected_weights * (positive_loss + negative_loss) / 2.0\n ).sum()\n if not bool(torch.isfinite(loss_sum)):\n raise RuntimeError(\"dense next-item training produced non-finite loss\")\n if backward:\n (loss_sum / total_weight).backward()\n if sequence_input.grad is None:\n raise RuntimeError(\"sequence microbatch did not produce a gradient\")\n sequence_gradient[begin:end] = sequence_input.grad\n statistics[\"loss_sum\"] += float(loss_sum.detach().cpu())\n statistics[\"positive_sum\"] += float(positive_logits.detach().sum().cpu())\n statistics[\"negative_sum\"] += float(negative_logits.detach().sum().cpu())\n statistics[\"pair_correct\"] += float(\n (positive_logits.detach()[:, None] > negative_logits.detach()).sum().cpu()\n )\n statistics[\"weight_sum\"] += float(selected_weights.sum().detach().cpu())\n return statistics, sequence_gradient\n\n\ndef release_accelerator_cache(device: torch.device) -> None:\n if device.type == \"mps\":\n torch.mps.empty_cache()\n\n\ndef run_training_epoch(\n model: base.SequentialTwoTower,\n optimizer: torch.optim.Optimizer,\n prepared: SequenceData,\n config: Config,\n target_weights: np.ndarray,\n window_starts: np.ndarray,\n target_counts: np.ndarray,\n item_correction: torch.Tensor,\n known_positive_packed: np.ndarray | None,\n generator: torch.Generator,\n device: torch.device,\n epoch: int,\n) -> dict[str, float]:\n model.train()\n train_users = np.flatnonzero(target_counts > 0)\n sorted_users = train_users[\n np.argsort(target_counts[train_users], kind=\"stable\")\n ]\n groups = [\n sorted_users[start : start + config.user_batch_size]\n for start in range(0, len(sorted_users), config.user_batch_size)\n ]\n group_order = torch.randperm(len(groups), generator=generator).numpy()\n negative_count = config.uniform_negatives + config.inbatch_negatives\n total_loss = 0.0\n total_targets = 0\n total_weight = 0.0\n total_positive = 0.0\n total_negative = 0.0\n total_pair_correct = 0.0\n batches = 0\n started = time.monotonic()\n\n for group_number in group_order:\n users = groups[int(group_number)]\n group_counts = target_counts[users]\n hidden = torch.zeros(\n 1, len(users), config.embedding_dim, device=device\n )\n max_targets = int(group_counts.max())\n for start in range(0, max_targets, config.bptt_steps):\n steps = min(config.bptt_steps, max_targets - start)\n block = build_dense_block(\n prepared,\n target_weights,\n users,\n window_starts[users],\n group_counts,\n np.zeros(len(users), dtype=np.int64),\n start,\n steps,\n config,\n )\n sequence_outputs, _ = event_sequence(\n model,\n torch.from_numpy(block[0].astype(np.int64)).to(device),\n torch.from_numpy(block[1].astype(np.float32)).to(device),\n hidden,\n )\n hidden = torch.stack(\n [\n sequence_outputs[row, count - 1]\n if count\n else hidden[0, row]\n for row, count in enumerate(block[7])\n ]\n ).unsqueeze(0).detach()\n flat_scored = block[5].reshape(-1)\n if not np.any(flat_scored):\n continue\n mask = torch.from_numpy(flat_scored).to(device)\n sequence_vectors = sequence_outputs.reshape(\n -1, config.embedding_dim\n )[mask]\n flat_states = block[3].reshape(-1)[flat_scored]\n positive_items = torch.from_numpy(\n block[4].reshape(-1)[flat_scored].astype(np.int64)\n ).to(device)\n known_history = torch.from_numpy(\n block[6].reshape(-1, config.bptt_steps)[flat_scored].astype(\n np.int64\n )\n ).to(device)\n liked_items = torch.from_numpy(\n prepared.liked_states[flat_states].astype(np.int64)\n ).to(device)\n flat_user_rows = np.broadcast_to(\n users[:, None], block[5].shape\n ).reshape(-1)[flat_scored]\n negatives, negative_weights = draw_training_negatives(\n positive_items,\n known_history,\n liked_items,\n flat_user_rows,\n known_positive_packed,\n prepared,\n config,\n item_correction,\n generator,\n device,\n )\n optimizer.zero_grad(set_to_none=True)\n statistics, sequence_gradient = score_target_microbatches(\n model,\n prepared,\n config,\n sequence_vectors,\n flat_states,\n block[2].reshape(-1, 4)[flat_scored].astype(np.float32),\n positive_items,\n negatives,\n negative_weights,\n block[8].reshape(-1)[flat_scored],\n device,\n backward=True,\n )\n if sequence_gradient is None:\n raise RuntimeError(\"training block omitted its sequence gradient\")\n sequence_vectors.backward(sequence_gradient)\n nn.utils.clip_grad_norm_(model.parameters(), config.gradient_clip)\n optimizer.step()\n\n count = len(positive_items)\n total_loss += statistics[\"loss_sum\"]\n total_targets += count\n total_weight += statistics[\"weight_sum\"]\n total_positive += statistics[\"positive_sum\"]\n total_negative += statistics[\"negative_sum\"]\n total_pair_correct += statistics[\"pair_correct\"]\n batches += 1\n if batches % 25 == 0:\n release_accelerator_cache(device)\n if batches % 250 == 0:\n batch_bce = statistics[\"loss_sum\"] / max(\n statistics[\"weight_sum\"], 1.0e-12\n )\n running_bce = total_loss / max(total_weight, 1.0e-12)\n log(\n f\"epoch={epoch}/{config.epochs} batches={batches:,} \"\n f\"targets={total_targets:,} \"\n f\"batch_bce={batch_bce:.6f} \"\n f\"running_bce={running_bce:.6f} \"\n f\"elapsed={time.monotonic() - started:.1f}s\"\n )\n\n return {\n \"epoch\": float(epoch),\n \"loss\": total_loss / max(total_weight, 1.0e-12),\n \"positive_logit\": total_positive / max(total_targets, 1),\n \"negative_logit\": total_negative / max(total_targets * negative_count, 1),\n \"pair_accuracy\": total_pair_correct\n / max(total_targets * negative_count, 1),\n \"targets\": float(total_targets),\n \"target_weight_sum\": total_weight,\n \"batches\": float(batches),\n \"seconds\": time.monotonic() - started,\n }\n\n\n@torch.no_grad()\ndef evaluate_validation_bce(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n target_weights: np.ndarray,\n window_starts: np.ndarray,\n training_counts: np.ndarray,\n validation_counts: np.ndarray,\n item_correction: torch.Tensor,\n known_positive_packed: np.ndarray | None,\n device: torch.device,\n) -> dict[str, float]:\n \"\"\"Measure deterministic BCE on each user's chronological target suffix.\"\"\"\n model.eval()\n totals = training_counts + validation_counts\n users_with_validation = np.flatnonzero(validation_counts > 0)\n ordered = users_with_validation[\n np.argsort(totals[users_with_validation], kind=\"stable\")\n ]\n groups = [\n ordered[start : start + config.user_batch_size]\n for start in range(0, len(ordered), config.user_batch_size)\n ]\n generator = torch.Generator(device=\"cpu\").manual_seed(config.seed + 20_000)\n total_loss = 0.0\n total_targets = 0\n total_weight = 0.0\n started = time.monotonic()\n\n for group_number, users in enumerate(groups, start=1):\n group_totals = totals[users]\n group_starts = training_counts[users]\n hidden = torch.zeros(\n 1, len(users), config.embedding_dim, device=device\n )\n max_targets = int(group_totals.max())\n for start in range(0, max_targets, config.bptt_steps):\n steps = min(config.bptt_steps, max_targets - start)\n block = build_dense_block(\n prepared,\n target_weights,\n users,\n window_starts[users],\n group_totals,\n group_starts,\n start,\n steps,\n config,\n )\n sequence_outputs, _ = event_sequence(\n model,\n torch.from_numpy(block[0].astype(np.int64)).to(device),\n torch.from_numpy(block[1].astype(np.float32)).to(device),\n hidden,\n )\n hidden = torch.stack(\n [\n sequence_outputs[row, count - 1]\n if count\n else hidden[0, row]\n for row, count in enumerate(block[7])\n ]\n ).unsqueeze(0)\n flat_scored = block[5].reshape(-1)\n if not np.any(flat_scored):\n continue\n mask = torch.from_numpy(flat_scored).to(device)\n sequence_vectors = sequence_outputs.reshape(\n -1, config.embedding_dim\n )[mask]\n flat_states = block[3].reshape(-1)[flat_scored]\n positive_items = torch.from_numpy(\n block[4].reshape(-1)[flat_scored].astype(np.int64)\n ).to(device)\n known_history = torch.from_numpy(\n block[6].reshape(-1, config.bptt_steps)[flat_scored].astype(\n np.int64\n )\n ).to(device)\n liked_items = torch.from_numpy(\n prepared.liked_states[flat_states].astype(np.int64)\n ).to(device)\n flat_user_rows = np.broadcast_to(\n users[:, None], block[5].shape\n ).reshape(-1)[flat_scored]\n negatives, negative_weights = draw_training_negatives(\n positive_items,\n known_history,\n liked_items,\n flat_user_rows,\n known_positive_packed,\n prepared,\n config,\n item_correction,\n generator,\n device,\n )\n statistics, _ = score_target_microbatches(\n model,\n prepared,\n config,\n sequence_vectors,\n flat_states,\n block[2].reshape(-1, 4)[flat_scored].astype(np.float32),\n positive_items,\n negatives,\n negative_weights,\n block[8].reshape(-1)[flat_scored],\n device,\n backward=False,\n )\n total_loss += statistics[\"loss_sum\"]\n total_targets += len(positive_items)\n total_weight += statistics[\"weight_sum\"]\n if group_number % 25 == 0:\n release_accelerator_cache(device)\n return {\n \"validation_loss\": total_loss / max(total_weight, 1.0e-12),\n \"validation_targets\": float(total_targets),\n \"validation_target_weight_sum\": total_weight,\n \"validation_seconds\": time.monotonic() - started,\n }\n\n\ndef snapshot_model(model: nn.Module) -> dict[str, torch.Tensor]:\n return {\n name: value.detach().cpu().clone()\n for name, value in model.state_dict().items()\n }\n\n\ndef save_epoch_checkpoint(\n checkpoint_dir: Path,\n epoch: int,\n model_state: dict[str, torch.Tensor],\n config: Config,\n diagnostics: dict[str, float],\n optimizer: torch.optim.Optimizer,\n generator: torch.Generator,\n) -> Path:\n checkpoint_dir.mkdir(parents=True, exist_ok=True)\n path = checkpoint_dir / f\"epoch-{epoch:02d}.pt\"\n torch.save(\n {\n \"epoch\": epoch,\n \"config\": asdict(config),\n \"diagnostics\": diagnostics,\n \"model_state\": model_state,\n \"optimizer_state\": optimizer.state_dict(),\n \"generator_state\": generator.get_state(),\n },\n path,\n )\n log(f\"saved epoch checkpoint: {path}\")\n return path\n\n\ndef train_model(\n public: Path,\n stage: str,\n prepared: SequenceData,\n config: Config,\n device: torch.device,\n checkpoint_dir: Path | None,\n resume_checkpoint: Path | None,\n) -> tuple[base.SequentialTwoTower, list[dict[str, float]]]:\n partition = chronological_target_partition(\n prepared,\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n training_counts = partition.training_counts\n validation_counts = partition.validation_counts\n log(\n f\"chronological BCE window: train={int(training_counts.sum()):,} \"\n f\"validation={int(validation_counts.sum()):,} \"\n f\"selected={partition.selected_count:,} \"\n f\"discarded_old_targets={int(partition.window_starts.sum()):,} \"\n f\"all={prepared.target_count:,}\"\n )\n model = base.SequentialTwoTower(prepared.item_features, model_config(config)).to(\n device\n )\n optimizer = torch.optim.AdamW(\n model.parameters(), lr=config.learning_rate, weight_decay=config.weight_decay\n )\n generator = torch.Generator(device=\"cpu\").manual_seed(config.seed + 10_000)\n item_correction = build_item_correction(\n prepared,\n config,\n partition.window_starts,\n training_counts,\n )\n target_weights = build_target_weights(\n public,\n stage,\n prepared,\n partition.history_training_counts,\n config,\n )\n known_positive_packed = (\n build_known_positive_filter(prepared, partition.history_training_counts)\n if config.negative_filter_scope == \"full-training-history\"\n else None\n )\n history: list[dict[str, float]] = []\n best_loss = float(\"inf\")\n best_epoch = 0\n best_state: dict[str, torch.Tensor] | None = None\n epochs_without_improvement = 0\n start_epoch = 0\n\n if resume_checkpoint is not None:\n payload = torch.load(\n resume_checkpoint, map_location=device, weights_only=False\n )\n stored_config = payload.get(\"config\", {})\n architecture_fields = (\n \"bptt_steps\",\n \"embedding_dim\",\n \"hidden_dim\",\n \"user_hidden_1\",\n \"user_hidden_2\",\n \"item_hidden_1\",\n \"item_hidden_2\",\n )\n mismatches = {\n name: (stored_config.get(name), getattr(config, name))\n for name in architecture_fields\n if stored_config.get(name) != getattr(config, name)\n }\n if mismatches:\n raise RuntimeError(\n f\"resume checkpoint architecture differs: {mismatches}\"\n )\n stored_filter_scope = stored_config.get(\n \"negative_filter_scope\", \"last-100\"\n )\n stored_target_weighting = stored_config.get(\n \"target_weighting\", \"uniform\"\n )\n stored_weight_values = (\n stored_config.get(\"ordinary_target_weight\", 1.0),\n stored_config.get(\"organic_strong_target_weight\", 1.0),\n stored_config.get(\"organic_like_target_weight\", 3.0),\n stored_config.get(\"discovery_horizon_days\", 7),\n )\n current_weight_values = (\n config.ordinary_target_weight,\n config.organic_strong_target_weight,\n config.organic_like_target_weight,\n config.discovery_horizon_days,\n )\n stored_window_values = (\n stored_config.get(\"validation_fraction\", 0.10),\n stored_config.get(\"max_train_targets_per_user\", 0),\n stored_config.get(\"max_validation_targets_per_user\", 0),\n )\n current_window_values = (\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n objective_changed = (\n stored_filter_scope != config.negative_filter_scope\n or stored_target_weighting != config.target_weighting\n or stored_weight_values != current_weight_values\n or stored_window_values != current_window_values\n )\n model.load_state_dict(payload[\"model_state\"])\n if \"optimizer_state\" in payload and not objective_changed:\n optimizer.load_state_dict(payload[\"optimizer_state\"])\n for group in optimizer.param_groups:\n group[\"lr\"] = config.learning_rate\n group[\"weight_decay\"] = config.weight_decay\n log(f\"restored optimizer state from {resume_checkpoint}\")\n else:\n log(\n f\"using a fresh AdamW optimizer after loading {resume_checkpoint}\"\n )\n if \"generator_state\" in payload and not objective_changed:\n generator.set_state(payload[\"generator_state\"])\n start_epoch = int(payload[\"epoch\"])\n if start_epoch >= config.epochs:\n raise RuntimeError(\n f\"resume epoch {start_epoch} must be below --epochs {config.epochs}\"\n )\n prior = {\n str(name): float(value)\n for name, value in payload.get(\"diagnostics\", {}).items()\n }\n if objective_changed:\n prior.pop(\"validation_loss\", None)\n log(\n \"BCE objective changed: filter \"\n f\"{stored_filter_scope}->{config.negative_filter_scope}, \"\n \"target weighting \"\n f\"{stored_target_weighting}->{config.target_weighting}, \"\n f\"target window {stored_window_values}->{current_window_values}; \"\n \"recomputing the resume checkpoint's validation BCE\"\n )\n if \"validation_loss\" not in prior:\n prior.update(\n evaluate_validation_bce(\n model,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n validation_counts,\n item_correction,\n known_positive_packed,\n device,\n )\n )\n history.append(prior)\n best_loss = float(prior[\"validation_loss\"])\n best_epoch = start_epoch\n best_state = snapshot_model(model)\n log(\n f\"resumed validation-BCE training: epoch={start_epoch} \"\n f\"loss={best_loss:.6f}\"\n )\n\n for epoch in range(start_epoch + 1, config.epochs + 1):\n diagnostics = run_training_epoch(\n model,\n optimizer,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n item_correction,\n known_positive_packed,\n generator,\n device,\n epoch,\n )\n validation = evaluate_validation_bce(\n model,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n validation_counts,\n item_correction,\n known_positive_packed,\n device,\n )\n diagnostics.update(validation)\n history.append(diagnostics)\n current_state = snapshot_model(model)\n if checkpoint_dir is not None:\n save_epoch_checkpoint(\n checkpoint_dir,\n epoch,\n current_state,\n config,\n diagnostics,\n optimizer,\n generator,\n )\n improved = (\n diagnostics[\"validation_loss\"]\n < best_loss - config.early_stopping_min_delta\n )\n if improved:\n best_loss = diagnostics[\"validation_loss\"]\n best_epoch = epoch\n best_state = current_state\n epochs_without_improvement = 0\n if checkpoint_dir is not None:\n torch.save(\n {\n \"epoch\": epoch,\n \"config\": asdict(config),\n \"diagnostics\": diagnostics,\n \"model_state\": current_state,\n \"optimizer_state\": optimizer.state_dict(),\n \"generator_state\": generator.get_state(),\n },\n checkpoint_dir / \"best.pt\",\n )\n else:\n epochs_without_improvement += 1\n log(\n f\"epoch={epoch}/{config.epochs} train_loss={diagnostics['loss']:.5f} \"\n f\"validation_loss={diagnostics['validation_loss']:.5f} \"\n f\"pair_accuracy={diagnostics['pair_accuracy']:.3f} \"\n f\"train_targets={int(diagnostics['targets']):,} \"\n f\"validation_targets={int(diagnostics['validation_targets']):,} \"\n f\"train_seconds={diagnostics['seconds']:.1f} \"\n f\"validation_seconds={diagnostics['validation_seconds']:.1f}\"\n )\n release_accelerator_cache(device)\n if epochs_without_improvement >= config.early_stopping_patience:\n log(\n f\"early stopping after epoch={epoch}; \"\n f\"best_epoch={best_epoch} best_validation_loss={best_loss:.6f}\"\n )\n break\n\n if best_state is None:\n raise RuntimeError(\"training did not produce a valid checkpoint\")\n model.load_state_dict(best_state)\n model.to(device).eval()\n log(\n f\"restored best validation-BCE checkpoint: epoch={best_epoch} \"\n f\"loss={best_loss:.6f}\"\n )\n return model, history\n\n\n@torch.inference_mode()\ndef inference_queries(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n device: torch.device,\n) -> torch.Tensor:\n sequence_rows = prepared.inference_user_rows\n lengths = np.where(\n sequence_rows >= 0, prepared.lengths[np.maximum(sequence_rows, 0)], 0\n )\n ordered = np.argsort(lengths, kind=\"stable\")\n result = torch.zeros(\n len(sequence_rows), config.embedding_dim, dtype=torch.float32\n )\n for group_start in range(0, len(ordered), config.user_batch_size):\n output_rows = ordered[group_start : group_start + config.user_batch_size]\n rows = sequence_rows[output_rows]\n group_lengths = lengths[output_rows]\n hidden = torch.zeros(\n 1, len(output_rows), config.embedding_dim, device=device\n )\n for start in range(0, int(group_lengths.max(initial=0)), config.bptt_steps):\n steps = min(\n config.bptt_steps,\n int(group_lengths.max(initial=0)) - start,\n )\n items = np.full(\n (len(output_rows), steps),\n len(prepared.item_ids),\n dtype=np.uint16,\n )\n features = np.zeros((len(output_rows), steps, 3), dtype=np.float16)\n counts = np.zeros(len(output_rows), dtype=np.int32)\n for local_row, sequence_row in enumerate(rows):\n if sequence_row < 0:\n continue\n count = min(steps, max(int(group_lengths[local_row]) - start, 0))\n if not count:\n continue\n offset = int(prepared.offsets[sequence_row])\n selected = slice(offset + start, offset + start + count)\n items[local_row, :count] = prepared.sequence_items[selected]\n features[local_row, :count] = prepared.event_features[selected]\n counts[local_row] = count\n outputs, _ = event_sequence(\n model,\n torch.from_numpy(items.astype(np.int64)).to(device),\n torch.from_numpy(features.astype(np.float32)).to(device),\n hidden,\n )\n next_rows: list[torch.Tensor] = []\n for local_row, count in enumerate(counts):\n next_rows.append(\n outputs[local_row, count - 1]\n if count\n else hidden[0, local_row]\n )\n hidden = torch.stack(next_rows).unsqueeze(0)\n\n liked_items = torch.from_numpy(\n prepared.inference_liked_items[output_rows].astype(np.int64)\n ).to(device)\n disliked_items = torch.from_numpy(\n prepared.inference_disliked_items[output_rows].astype(np.int64)\n ).to(device)\n queries = user_queries(\n model,\n hidden.squeeze(0),\n model._mean_context_channel(liked_items),\n model._mean_context_channel(disliked_items),\n torch.from_numpy(\n prepared.inference_user_features[output_rows].astype(np.float32)\n ).to(device),\n )\n result[torch.from_numpy(output_rows.astype(np.int64))] = queries.cpu()\n return result\n\n\n@torch.inference_mode()\ndef rank_candidates(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n eligibility: CandidateSet,\n config: Config,\n device: torch.device,\n ranking_batch_size: int,\n) -> tuple[pd.DataFrame, dict[str, float]]:\n model.eval()\n queries = inference_queries(model, prepared, config, device)\n item_parts: list[torch.Tensor] = []\n for start in range(0, len(prepared.item_ids), 2_048):\n rows = torch.arange(\n start,\n min(start + 2_048, len(prepared.item_ids)),\n dtype=torch.long,\n device=device,\n )\n item_parts.append(model.encode_items(rows).cpu())\n items = torch.cat(item_parts)\n rows: list[tuple[int, int, int]] = []\n for start in range(0, len(eligibility.uids), ranking_batch_size):\n end = min(start + ranking_batch_size, len(eligibility.uids))\n scores = queries[start:end] @ items.T\n eligible = np.unpackbits(\n eligibility.packed[start:end],\n axis=1,\n count=len(prepared.item_ids),\n bitorder=\"little\",\n ).astype(bool, copy=False)\n scores.masked_fill_(~torch.from_numpy(eligible), -torch.inf)\n selected = torch.topk(scores, k=base.TOP_N, dim=1).indices.numpy()\n for local_row, columns in enumerate(selected):\n uid = int(eligibility.uids[start + local_row])\n rows.extend(\n (uid, int(prepared.item_ids[column]), rank)\n for rank, column in enumerate(columns, start=1)\n )\n if end == len(eligibility.uids) or end % 1_024 == 0:\n log(f\"ranked users={end:,}/{len(eligibility.uids):,}\")\n query_norm = torch.linalg.vector_norm(queries, dim=1).numpy()\n item_norm = torch.linalg.vector_norm(items, dim=1).numpy()\n diagnostics = {\n \"query_norm_mean\": float(query_norm.mean()),\n \"query_norm_std\": float(query_norm.std()),\n \"item_norm_mean\": float(item_norm.mean()),\n \"item_norm_std\": float(item_norm.std()),\n }\n return pd.DataFrame(rows, columns=[\"uid\", \"item_id\", \"rank\"]), diagnostics\n\n\ndef write_ranking(frame: pd.DataFrame, output: Path) -> None:\n output.parent.mkdir(parents=True, exist_ok=True)\n if output.suffix == \".parquet\":\n frame.to_parquet(output, index=False, compression=\"zstd\")\n elif output.suffix == \".csv\":\n frame.to_csv(output, index=False)\n else:\n raise ValueError(\"--output must end in .parquet or .csv\")\n\n\ndef main() -> int:\n parser = argparse.ArgumentParser(description=__doc__)\n parser.add_argument(\n \"--stage\", choices=(\"validation\", \"submission\"), default=\"validation\"\n )\n parser.add_argument(\"--output\", type=Path, required=True)\n parser.add_argument(\"--public-dir\", type=Path, required=True)\n parser.add_argument(\"--prepared-cache\", type=Path)\n parser.add_argument(\"--device\", choices=(\"cpu\", \"mps\"), default=\"cpu\")\n parser.add_argument(\"--threads\", type=int, default=8)\n parser.add_argument(\"--bptt-steps\", type=int, default=Config.bptt_steps)\n parser.add_argument(\n \"--user-batch-size\", type=int, default=Config.user_batch_size\n )\n parser.add_argument(\"--epochs\", type=int, default=Config.epochs)\n parser.add_argument(\"--learning-rate\", type=float, default=Config.learning_rate)\n parser.add_argument(\n \"--loss-microbatch-size\",\n type=int,\n default=Config.loss_microbatch_size,\n )\n parser.add_argument(\n \"--negative-filter-scope\",\n choices=(\"full-training-history\", \"last-100\"),\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n )\n parser.add_argument(\n \"--ordinary-target-weight\",\n type=float,\n default=Config.ordinary_target_weight,\n )\n parser.add_argument(\n \"--organic-strong-target-weight\",\n type=float,\n default=Config.organic_strong_target_weight,\n )\n parser.add_argument(\n \"--organic-like-target-weight\",\n type=float,\n default=Config.organic_like_target_weight,\n )\n parser.add_argument(\n \"--discovery-horizon-days\",\n type=int,\n default=Config.discovery_horizon_days,\n )\n parser.add_argument(\n \"--validation-fraction\",\n type=float,\n default=Config.validation_fraction,\n )\n parser.add_argument(\n \"--max-train-targets-per-user\",\n type=int,\n default=Config.max_train_targets_per_user,\n help=\"retain this many recent pre-validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--max-validation-targets-per-user\",\n type=int,\n default=Config.max_validation_targets_per_user,\n help=\"retain this many final validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--early-stopping-patience\",\n type=int,\n default=Config.early_stopping_patience,\n )\n parser.add_argument(\n \"--early-stopping-min-delta\",\n type=float,\n default=Config.early_stopping_min_delta,\n )\n parser.add_argument(\"--checkpoint-dir\", type=Path)\n parser.add_argument(\"--resume-checkpoint\", type=Path)\n parser.add_argument(\"--seed\", type=int, default=Config.seed)\n parser.add_argument(\"--ranking-batch-size\", type=int, default=64)\n parser.add_argument(\"--prepare-only\", action=\"store_true\")\n args = parser.parse_args()\n\n if args.threads < 1 or args.bptt_steps < 2 or args.user_batch_size < 1:\n parser.error(\"threads, BPTT steps, and user batch size must be positive\")\n if (\n args.epochs < 1\n or args.learning_rate <= 0\n or args.loss_microbatch_size < 1\n or args.early_stopping_patience < 1\n or args.early_stopping_min_delta < 0\n or args.ordinary_target_weight <= 0\n or args.organic_strong_target_weight <= 0\n or args.organic_like_target_weight <= 0\n or args.discovery_horizon_days < 1\n or args.max_train_targets_per_user < 0\n or args.max_validation_targets_per_user < 0\n ):\n parser.error(\"training counts, learning rate, and patience are invalid\")\n if not 0.0 < args.validation_fraction < 1.0:\n parser.error(\"validation fraction must be strictly between zero and one\")\n public = args.public_dir.resolve()\n if not (public / \"data\" / \"candidate_catalog.parquet\").is_file():\n parser.error(f\"public task data not found under {public}\")\n config = Config(\n bptt_steps=args.bptt_steps,\n user_batch_size=args.user_batch_size,\n epochs=args.epochs,\n learning_rate=args.learning_rate,\n loss_microbatch_size=args.loss_microbatch_size,\n negative_filter_scope=args.negative_filter_scope,\n target_weighting=args.target_weighting,\n ordinary_target_weight=args.ordinary_target_weight,\n organic_strong_target_weight=args.organic_strong_target_weight,\n organic_like_target_weight=args.organic_like_target_weight,\n discovery_horizon_days=args.discovery_horizon_days,\n validation_fraction=args.validation_fraction,\n max_train_targets_per_user=args.max_train_targets_per_user,\n max_validation_targets_per_user=args.max_validation_targets_per_user,\n early_stopping_patience=args.early_stopping_patience,\n early_stopping_min_delta=args.early_stopping_min_delta,\n seed=args.seed,\n )\n torch.set_num_threads(args.threads)\n torch.set_num_interop_threads(1)\n seed_everything(config.seed)\n device = torch.device(args.device)\n if device.type == \"mps\" and not torch.backends.mps.is_available():\n parser.error(\"MPS was requested but is unavailable\")\n\n started = time.monotonic()\n prepared, eligibility = load_or_prepare(\n public,\n args.stage,\n config,\n args.prepared_cache.resolve() if args.prepared_cache else None,\n )\n target_partition = chronological_target_partition(\n prepared,\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n if args.prepare_only:\n print(\n json.dumps(\n {\n \"stage\": args.stage,\n \"sequence_users\": len(prepared.user_ids),\n \"sequence_tokens\": len(prepared.sequence_items),\n \"next_item_targets\": prepared.target_count,\n \"selected_training_targets\": int(\n target_partition.training_counts.sum()\n ),\n \"selected_validation_targets\": int(\n target_partition.validation_counts.sum()\n ),\n \"discarded_old_targets\": int(\n target_partition.window_starts.sum()\n ),\n \"target_users\": len(eligibility.uids),\n \"runtime_seconds\": round(time.monotonic() - started, 3),\n \"config\": asdict(config),\n },\n indent=2,\n )\n )\n return 0\n\n model, training_history = train_model(\n public,\n args.stage,\n prepared,\n config,\n device,\n args.checkpoint_dir.resolve() if args.checkpoint_dir else None,\n args.resume_checkpoint.resolve() if args.resume_checkpoint else None,\n )\n ranking, embedding_diagnostics = rank_candidates(\n model,\n prepared,\n eligibility,\n config,\n device,\n args.ranking_batch_size,\n )\n write_ranking(ranking, args.output.resolve())\n print(\n json.dumps(\n {\n \"baseline\": \"yambda-windowed-sequence-two-tower-base\",\n \"stage\": args.stage,\n \"output\": str(args.output.resolve()),\n \"rows\": len(ranking),\n \"users\": int(ranking.uid.nunique()),\n \"runtime_seconds\": round(time.monotonic() - started, 3),\n \"public_inputs_only\": True,\n \"architecture\": {\n \"family\": \"pure-windowed-sequence-two-tower\",\n \"sequence_training\": (\n \"every-adjacent-target-within-recent-chronological-window\"\n ),\n \"sequence_encoder\": \"one-layer-GRU-dense-timestep-loss\",\n \"bptt_hidden_state\": \"carried-and-detached\",\n \"model_selection\": \"chronological-validation-BCE-early-stopping\",\n \"loss_execution\": \"gradient-accumulated-microbatches\",\n \"user_id_embedding\": False,\n \"user_mlp\": \"582-512-256-192\",\n \"item_mlp\": \"196-512-256-192\",\n \"normalization\": \"none\",\n \"score\": \"raw-dot-product\",\n \"objective\": \"sample-weighted-BCE-at-every-timestep\",\n \"negative_sampling\": (\n \"24-uniform-plus-8-propensity-corrected-inbatch\"\n ),\n \"false_negative_filter\": config.negative_filter_scope,\n \"target_weighting\": config.target_weighting,\n },\n \"config\": asdict(config),\n \"sequence_tokens\": len(prepared.sequence_items),\n \"all_next_item_targets\": prepared.target_count,\n \"selected_training_targets\": int(\n target_partition.training_counts.sum()\n ),\n \"selected_validation_targets\": int(\n target_partition.validation_counts.sum()\n ),\n \"discarded_old_targets\": int(target_partition.window_starts.sum()),\n \"training_history\": training_history,\n \"embedding_diagnostics\": embedding_diagnostics,\n \"torch_version\": torch.__version__,\n \"device\": str(device),\n },\n indent=2,\n )\n )\n return 0\n\n\nif __name__ == \"__main__\":\n raise SystemExit(main())\n",
"diffStat": {
"model_added_lines": 17,
"model_removed_lines": 7,
"model_added_chars": 746,
"model_removed_chars": 276,
"user_added_lines": 0,
"user_removed_lines": 0,
"user_added_chars": 0,
"user_removed_chars": 0
},
"isNewFile": false,
"isBuildFile": false
},
"description": "solution/train_current_model.py: positive_logits = (que... => positive_logits = (que...",
"displayName": "Edit",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T06:06:57.267Z",
"content": "Successfully modified file: /app/solution/train_current_model.py (1 replacements). Here is the updated code:\n...\n )\n positive_logits = (query_vectors * positive_vectors).sum(dim=1)\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n if config.objective == \"infonce\":\n cat_logits = torch.cat(\n [positive_logits.unsqueeze(1), negative_logits], dim=1\n ) / config.temperature\n labels = torch.zeros(\n len(positive_logits), dtype=torch.long, device=device\n )\n raw_loss = F.cross_entropy(cat_logits, labels, reduction=\"none\")\n loss_sum = (selected_weights * raw_loss).sum()\n else:\n positive_loss = F.softplus(-positive_logits)\n negative_loss = corrected_negative_loss(\n negative_logits, negative_weights[begin:end], config\n )\n loss_sum = (\n selected_weights * (positive_loss + negative_loss) / 2.0\n ).sum()\n if not bool(torch.isfinite(loss_sum)):\n raise RuntimeError(\"dense next-item training produced non-finite loss\")\n if backward:\n (loss_sum / total_weight).backward()\n if sequence_input.grad is None:\n...",
"isError": false
},
"responseMessageId": "ccb9f932-1a73-41de-a3cb-409a1af8e7fb"
}
]
},
{
"id": 48,
"role": "agent",
"timestamp": "2026-09-29T06:07:01.210Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T06:07:01.210Z",
"callId": "read_file__call_165772",
"name": "read_file",
"input": {
"file_path": "/app/solution/train_current_model.py",
"end_line": 2050,
"start_line": 1940
},
"recorded": {
"id": "read_file__call_165772",
"name": "read_file",
"args": {
"file_path": "/app/solution/train_current_model.py",
"end_line": 2050,
"start_line": 1940
},
"result": [
{
"functionResponse": {
"id": "read_file__call_165772",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1940-2050 of 2177 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 2051.\n\n--- FILE CONTENT (truncated) ---\n parser.add_argument(\"--public-dir\", type=Path, required=True)\n parser.add_argument(\"--prepared-cache\", type=Path)\n parser.add_argument(\"--device\", choices=(\"cpu\", \"mps\"), default=\"cpu\")\n parser.add_argument(\"--threads\", type=int, default=8)\n parser.add_argument(\"--bptt-steps\", type=int, default=Config.bptt_steps)\n parser.add_argument(\n \"--user-batch-size\", type=int, default=Config.user_batch_size\n )\n parser.add_argument(\"--epochs\", type=int, default=Config.epochs)\n parser.add_argument(\"--learning-rate\", type=float, default=Config.learning_rate)\n parser.add_argument(\n \"--loss-microbatch-size\",\n type=int,\n default=Config.loss_microbatch_size,\n )\n parser.add_argument(\n \"--negative-filter-scope\",\n choices=(\"full-training-history\", \"last-100\"),\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n )\n parser.add_argument(\n \"--ordinary-target-weight\",\n type=float,\n default=Config.ordinary_target_weight,\n )\n parser.add_argument(\n \"--organic-strong-target-weight\",\n type=float,\n default=Config.organic_strong_target_weight,\n )\n parser.add_argument(\n \"--organic-like-target-weight\",\n type=float,\n default=Config.organic_like_target_weight,\n )\n parser.add_argument(\n \"--discovery-horizon-days\",\n type=int,\n default=Config.discovery_horizon_days,\n )\n parser.add_argument(\n \"--validation-fraction\",\n type=float,\n default=Config.validation_fraction,\n )\n parser.add_argument(\n \"--max-train-targets-per-user\",\n type=int,\n default=Config.max_train_targets_per_user,\n help=\"retain this many recent pre-validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--max-validation-targets-per-user\",\n type=int,\n default=Config.max_validation_targets_per_user,\n help=\"retain this many final validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--early-stopping-patience\",\n type=int,\n default=Config.early_stopping_patience,\n )\n parser.add_argument(\n \"--early-stopping-min-delta\",\n type=float,\n default=Config.early_stopping_min_delta,\n )\n parser.add_argument(\"--checkpoint-dir\", type=Path)\n parser.add_argument(\"--resume-checkpoint\", type=Path)\n parser.add_argument(\"--seed\", type=int, default=Config.seed)\n parser.add_argument(\"--ranking-batch-size\", type=int, default=64)\n parser.add_argument(\"--prepare-only\", action=\"store_true\")\n args = parser.parse_args()\n\n if args.threads < 1 or args.bptt_steps < 2 or args.user_batch_size < 1:\n parser.error(\"threads, BPTT steps, and user batch size must be positive\")\n if (\n args.epochs < 1\n or args.learning_rate <= 0\n or args.loss_microbatch_size < 1\n or args.early_stopping_patience < 1\n or args.early_stopping_min_delta < 0\n or args.ordinary_target_weight <= 0\n or args.organic_strong_target_weight <= 0\n or args.organic_like_target_weight <= 0\n or args.discovery_horizon_days < 1\n or args.max_train_targets_per_user < 0\n or args.max_validation_targets_per_user < 0\n ):\n parser.error(\"training counts, learning rate, and patience are invalid\")\n if not 0.0 < args.validation_fraction < 1.0:\n parser.error(\"validation fraction must be strictly between zero and one\")\n public = args.public_dir.resolve()\n if not (public / \"data\" / \"candidate_catalog.parquet\").is_file():\n parser.error(f\"public task data not found under {public}\")\n config = Config(\n bptt_steps=args.bptt_steps,\n user_batch_size=args.user_batch_size,\n epochs=args.epochs,\n learning_rate=args.learning_rate,\n loss_microbatch_size=args.loss_microbatch_size,\n negative_filter_scope=args.negative_filter_scope,\n target_weighting=args.target_weighting,\n ordinary_target_weight=args.ordinary_target_weight,\n organic_strong_target_weight=args.organic_strong_target_weight,\n organic_like_target_weight=args.organic_like_target_weight,"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T06:07:01.215Z",
"resultDisplay": "Read lines 1940-2050 of 2177 from solution/train_current_model.py",
"description": "solution/train_current_model.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T06:07:01.217Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 1940-2050 of 2177 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 2051.\n\n--- FILE CONTENT (truncated) ---\n parser.add_argument(\"--public-dir\", type=Path, required=True)\n parser.add_argument(\"--prepared-cache\", type=Path)\n parser.add_argument(\"--device\", choices=(\"cpu\", \"mps\"), default=\"cpu\")\n parser.add_argument(\"--threads\", type=int, default=8)\n parser.add_argument(\"--bptt-steps\", type=int, default=Config.bptt_steps)\n parser.add_argument(\n \"--user-batch-size\", type=int, default=Config.user_batch_size\n )\n parser.add_argument(\"--epochs\", type=int, default=Config.epochs)\n parser.add_argument(\"--learning-rate\", type=float, default=Config.learning_rate)\n parser.add_argument(\n \"--loss-microbatch-size\",\n type=int,\n default=Config.loss_microbatch_size,\n )\n parser.add_argument(\n \"--negative-filter-scope\",\n choices=(\"full-training-history\", \"last-100\"),\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n )\n parser.add_argument(\n \"--ordinary-target-weight\",\n type=float,\n default=Config.ordinary_target_weight,\n )\n parser.add_argument(\n \"--organic-strong-target-weight\",\n type=float,\n default=Config.organic_strong_target_weight,\n )\n parser.add_argument(\n \"--organic-like-target-weight\",\n type=float,\n default=Config.organic_like_target_weight,\n )\n parser.add_argument(\n \"--discovery-horizon-days\",\n type=int,\n default=Config.discovery_horizon_days,\n )\n parser.add_argument(\n \"--validation-fraction\",\n type=float,\n default=Config.validation_fraction,\n )\n parser.add_argument(\n \"--max-train-targets-per-user\",\n type=int,\n default=Config.max_train_targets_per_user,\n help=\"retain this many recent pre-validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--max-validation-targets-per-user\",\n type=int,\n default=Config.max_validation_targets_per_user,\n help=\"retain this many final validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--early-stopping-patience\",\n type=int,\n default=Config.early_stopping_patience,\n )\n parser.add_argument(\n \"--early-stopping-min-delta\",\n type=float,\n default=Config.early_stopping_min_delta,\n )\n parser.add_argument(\"--checkpoint-dir\", type=Path)\n parser.add_argument(\"--resume-checkpoint\", type=Path)\n parser.add_argument(\"--seed\", type=int, default=Config.seed)\n parser.add_argument(\"--ranking-batch-size\", type=int, default=64)\n parser.add_argument(\"--prepare-only\", action=\"store_true\")\n args = parser.parse_args()\n\n if args.threads < 1 or args.bptt_steps < 2 or args.user_batch_size < 1:\n parser.error(\"threads, BPTT steps, and user batch size must be positive\")\n if (\n args.epochs < 1\n or args.learning_rate <= 0\n or args.loss_microbatch_size < 1\n or args.early_stopping_patience < 1\n or args.early_stopping_min_delta < 0\n or args.ordinary_target_weight <= 0\n or args.organic_strong_target_weight <= 0\n or args.organic_like_target_weight <= 0\n or args.discovery_horizon_days < 1\n or args.max_train_targets_per_user < 0\n or args.max_validation_targets_per_user < 0\n ):\n parser.error(\"training counts, learning rate, and patience are invalid\")\n if not 0.0 < args.validation_fraction < 1.0:\n parser.error(\"validation fraction must be strictly between zero and one\")\n public = args.public_dir.resolve()\n if not (public / \"data\" / \"candidate_catalog.parquet\").is_file():\n parser.error(f\"public task data not found under {public}\")\n config = Config(\n bptt_steps=args.bptt_steps,\n user_batch_size=args.user_batch_size,\n epochs=args.epochs,\n learning_rate=args.learning_rate,\n loss_microbatch_size=args.loss_microbatch_size,\n negative_filter_scope=args.negative_filter_scope,\n target_weighting=args.target_weighting,\n ordinary_target_weight=args.ordinary_target_weight,\n organic_strong_target_weight=args.organic_strong_target_weight,\n organic_like_target_weight=args.organic_like_target_weight,",
"isError": false
},
"responseMessageId": "687e783e-4804-4115-9300-3d44dea0998b"
}
]
},
{
"id": 49,
"role": "agent",
"timestamp": "2026-09-29T06:07:04.043Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T06:07:04.043Z",
"callId": "read_file__call_171486",
"name": "read_file",
"input": {
"end_line": 2070,
"file_path": "/app/solution/train_current_model.py",
"start_line": 2045
},
"recorded": {
"id": "read_file__call_171486",
"name": "read_file",
"args": {
"end_line": 2070,
"file_path": "/app/solution/train_current_model.py",
"start_line": 2045
},
"result": [
{
"functionResponse": {
"id": "read_file__call_171486",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 2045-2070 of 2177 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 2071.\n\n--- FILE CONTENT (truncated) ---\n loss_microbatch_size=args.loss_microbatch_size,\n negative_filter_scope=args.negative_filter_scope,\n target_weighting=args.target_weighting,\n ordinary_target_weight=args.ordinary_target_weight,\n organic_strong_target_weight=args.organic_strong_target_weight,\n organic_like_target_weight=args.organic_like_target_weight,\n discovery_horizon_days=args.discovery_horizon_days,\n validation_fraction=args.validation_fraction,\n max_train_targets_per_user=args.max_train_targets_per_user,\n max_validation_targets_per_user=args.max_validation_targets_per_user,\n early_stopping_patience=args.early_stopping_patience,\n early_stopping_min_delta=args.early_stopping_min_delta,\n seed=args.seed,\n )\n torch.set_num_threads(args.threads)\n torch.set_num_interop_threads(1)\n seed_everything(config.seed)\n device = torch.device(args.device)\n if device.type == \"mps\" and not torch.backends.mps.is_available():\n parser.error(\"MPS was requested but is unavailable\")\n\n started = time.monotonic()\n prepared, eligibility = load_or_prepare(\n public,\n args.stage,\n config,"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T06:07:04.048Z",
"resultDisplay": "Read lines 2045-2070 of 2177 from solution/train_current_model.py",
"description": "solution/train_current_model.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T06:07:04.049Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 2045-2070 of 2177 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 2071.\n\n--- FILE CONTENT (truncated) ---\n loss_microbatch_size=args.loss_microbatch_size,\n negative_filter_scope=args.negative_filter_scope,\n target_weighting=args.target_weighting,\n ordinary_target_weight=args.ordinary_target_weight,\n organic_strong_target_weight=args.organic_strong_target_weight,\n organic_like_target_weight=args.organic_like_target_weight,\n discovery_horizon_days=args.discovery_horizon_days,\n validation_fraction=args.validation_fraction,\n max_train_targets_per_user=args.max_train_targets_per_user,\n max_validation_targets_per_user=args.max_validation_targets_per_user,\n early_stopping_patience=args.early_stopping_patience,\n early_stopping_min_delta=args.early_stopping_min_delta,\n seed=args.seed,\n )\n torch.set_num_threads(args.threads)\n torch.set_num_interop_threads(1)\n seed_everything(config.seed)\n device = torch.device(args.device)\n if device.type == \"mps\" and not torch.backends.mps.is_available():\n parser.error(\"MPS was requested but is unavailable\")\n\n started = time.monotonic()\n prepared, eligibility = load_or_prepare(\n public,\n args.stage,\n config,",
"isError": false
},
"responseMessageId": "c222fdeb-12fa-43bc-9e28-39fe370268bd"
}
]
},
{
"id": 50,
"role": "agent",
"timestamp": "2026-09-29T06:07:10.132Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T06:07:10.132Z",
"callId": "replace__call_49464",
"name": "replace",
"input": {
"instruction": "Add CLI arguments for objective and temperature",
"file_path": "/app/solution/train_current_model.py",
"new_string": " parser.add_argument(\n \"--negative-filter-scope\",\n choices=(\"full-training-history\", \"last-100\"),\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n \"--objective\",\n choices=(\"bce\", \"infonce\"),\n default=Config.objective,\n )\n parser.add_argument(\n \"--temperature\",\n type=float,\n default=Config.temperature,\n )\n parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n )",
"old_string": " parser.add_argument(\n \"--negative-filter-scope\",\n choices=(\"full-training-history\", \"last-100\"),\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n )",
"allow_multiple": false
},
"recorded": {
"id": "replace__call_49464",
"name": "replace",
"args": {
"instruction": "Add CLI arguments for objective and temperature",
"file_path": "/app/solution/train_current_model.py",
"new_string": " parser.add_argument(\n \"--negative-filter-scope\",\n choices=(\"full-training-history\", \"last-100\"),\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n \"--objective\",\n choices=(\"bce\", \"infonce\"),\n default=Config.objective,\n )\n parser.add_argument(\n \"--temperature\",\n type=float,\n default=Config.temperature,\n )\n parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n )",
"old_string": " parser.add_argument(\n \"--negative-filter-scope\",\n choices=(\"full-training-history\", \"last-100\"),\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n )",
"allow_multiple": false
},
"result": [
{
"functionResponse": {
"id": "replace__call_49464",
"name": "replace",
"response": {
"output": "Successfully modified file: /app/solution/train_current_model.py (1 replacements). Here is the updated code:\n...\n \"--negative-filter-scope\",\n choices=(\"full-training-history\", \"last-100\"),\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n \"--objective\",\n choices=(\"bce\", \"infonce\"),\n default=Config.objective,\n )\n parser.add_argument(\n \"--temperature\",\n type=float,\n default=Config.temperature,\n )\n parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n )\n parser.add_argument(\n..."
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T06:07:10.146Z",
"resultDisplay": {
"fileDiff": "Index: train_current_model.py\n===================================================================\n--- train_current_model.py\tCurrent\n+++ train_current_model.py\tProposed\n@@ -1958,6 +1958,16 @@\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n+ \"--objective\",\n+ choices=(\"bce\", \"infonce\"),\n+ default=Config.objective,\n+ )\n+ parser.add_argument(\n+ \"--temperature\",\n+ type=float,\n+ default=Config.temperature,\n+ )\n+ parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n",
"fileName": "train_current_model.py",
"filePath": "/app/solution/train_current_model.py",
"originalContent": "#!/usr/bin/env python3\n\"\"\"Train the Yambda windowed two-tower baseline.\n\nEvery catalog listen with at least 50 percent completion is a sequence token.\nWithin each selected chronological window, every adjacent target contributes a\nloss. The fast default retains each user's most recent 128 training targets and\n16 held-out validation targets; passing zero for either cap restores the\nuncapped behavior for that side of the split. A 100-step truncated-BPTT block\ncomputes all of its timestep losses and carries the detached GRU hidden state\ninto the next block.\n\nThe model is a pure two-tower retriever: no user-ID embedding, no output\nnormalization, raw dot-product retrieval, sample-weighted binary cross entropy,\nand 24 uniform plus eight propensity-corrected unique in-batch negatives. Every\nselected adjacent target retains a nonzero weight. Only files under the supplied\npublic directory are read.\n\"\"\"\nfrom __future__ import annotations\n\nimport argparse\nimport json\nimport random\nimport sys\nimport time\nfrom dataclasses import asdict, dataclass\nfrom pathlib import Path\n\nimport numpy as np\nimport pandas as pd\nimport torch\nfrom torch import nn\nfrom torch.nn import functional as F\n\nimport two_tower_model as base\n\n\nCACHE_VERSION = 2\n\n\n@dataclass(frozen=True)\nclass Config:\n bptt_steps: int = 100\n user_batch_size: int = 64\n embedding_dim: int = 192\n hidden_dim: int = 192\n user_hidden_1: int = 512\n user_hidden_2: int = 256\n item_hidden_1: int = 512\n item_hidden_2: int = 256\n dropout: float = 0.05\n epochs: int = 1\n learning_rate: float = 1.0e-3\n weight_decay: float = 1.0e-5\n gradient_clip: float = 2.0\n loss_microbatch_size: int = 65_536\n uniform_negatives: int = 24\n inbatch_negatives: int = 8\n inbatch_correction_power: float = 0.5\n inbatch_correction_min: float = 0.25\n inbatch_correction_max: float = 4.0\n negative_filter_scope: str = \"full-training-history\"\n objective: str = \"bce\"\n temperature: float = 0.2\n target_weighting: str = \"organic-strong\"\n ordinary_target_weight: float = 0.25\n organic_strong_target_weight: float = 1.0\n organic_like_target_weight: float = 3.0\n discovery_horizon_days: int = 7\n validation_fraction: float = 0.10\n max_train_targets_per_user: int = 128\n max_validation_targets_per_user: int = 16\n early_stopping_patience: int = 1\n early_stopping_min_delta: float = 1.0e-4\n seed: int = 29\n\n\n@dataclass(frozen=True)\nclass CandidateSet:\n uids: np.ndarray\n item_ids: np.ndarray\n packed: np.ndarray\n\n @classmethod\n def read(cls, path: Path) -> \"CandidateSet\":\n with np.load(path, allow_pickle=False) as artifact:\n result = cls(\n uids=artifact[\"uids\"].astype(np.int64, copy=True),\n item_ids=artifact[\"item_ids\"].astype(np.int64, copy=True),\n packed=artifact[\"packed_eligible\"].astype(np.uint8, copy=True),\n )\n expected = (len(result.uids), (len(result.item_ids) + 7) // 8)\n if result.packed.shape != expected:\n raise ValueError(\n f\"invalid candidate artifact shape {result.packed.shape}; \"\n f\"expected {expected}\"\n )\n return result\n\n\n@dataclass(frozen=True)\nclass SequenceData:\n item_ids: np.ndarray\n item_features: np.ndarray\n user_ids: np.ndarray\n offsets: np.ndarray\n sequence_timestamps: np.ndarray\n sequence_items: np.ndarray\n event_features: np.ndarray\n activity_features: np.ndarray\n state_ids: np.ndarray\n liked_states: np.ndarray\n disliked_states: np.ndarray\n state_counts: np.ndarray\n inference_user_rows: np.ndarray\n inference_liked_items: np.ndarray\n inference_disliked_items: np.ndarray\n inference_user_features: np.ndarray\n\n @property\n def lengths(self) -> np.ndarray:\n return np.diff(self.offsets)\n\n @property\n def target_count(self) -> int:\n return int(np.maximum(self.lengths - 1, 0).sum())\n\n\n@dataclass(frozen=True)\nclass TargetPartition:\n \"\"\"A contiguous recent target window and its chronological BCE split.\"\"\"\n\n window_starts: np.ndarray\n training_counts: np.ndarray\n validation_counts: np.ndarray\n history_training_counts: np.ndarray\n\n @property\n def selected_count(self) -> int:\n return int(self.training_counts.sum() + self.validation_counts.sum())\n\n\ndef log(message: str) -> None:\n print(message, file=sys.stderr, flush=True)\n\n\ndef seed_everything(seed: int) -> None:\n random.seed(seed)\n np.random.seed(seed)\n torch.manual_seed(seed)\n\n\ndef model_config(config: Config) -> base.TowerConfig:\n \"\"\"Map the dense trainer configuration onto the unchanged tower.\"\"\"\n return base.TowerConfig(\n sequence_length=config.bptt_steps,\n embedding_dim=config.embedding_dim,\n hidden_dim=config.hidden_dim,\n user_hidden_1=config.user_hidden_1,\n user_hidden_2=config.user_hidden_2,\n item_hidden_1=config.item_hidden_1,\n item_hidden_2=config.item_hidden_2,\n dropout=config.dropout,\n epochs=config.epochs,\n learning_rate=config.learning_rate,\n weight_decay=config.weight_decay,\n gradient_clip=config.gradient_clip,\n uniform_negatives=config.uniform_negatives,\n inbatch_negatives=config.inbatch_negatives,\n inbatch_correction_power=config.inbatch_correction_power,\n inbatch_correction_min=config.inbatch_correction_min,\n inbatch_correction_max=config.inbatch_correction_max,\n seed=config.seed,\n )\n\n\ndef append_state_snapshot(\n active_likes: dict[int, int],\n active_dislikes: dict[int, int],\n n_items: int,\n liked_states: list[np.ndarray],\n disliked_states: list[np.ndarray],\n state_counts: list[tuple[int, int]],\n) -> int:\n liked_states.append(base.pack_active_items(active_likes, base.LIKE_ITEMS, n_items))\n disliked_states.append(\n base.pack_active_items(active_dislikes, base.DISLIKE_ITEMS, n_items)\n )\n state_counts.append((len(active_likes), len(active_dislikes)))\n return len(liked_states) - 1\n\n\ndef causal_activity_features(\n timestamps: np.ndarray,\n items: np.ndarray,\n organic: np.ndarray,\n completion: np.ndarray,\n) -> np.ndarray:\n \"\"\"Compute the first four reviewed user features at every event cutoff.\"\"\"\n length = len(items)\n result = np.zeros((length, 4), dtype=np.float16)\n if not length:\n return result\n left_rows = np.searchsorted(\n timestamps,\n timestamps - base.PROFILE_LOOKBACK_DAYS * base.DAY,\n side=\"left\",\n )\n rows = np.arange(length, dtype=np.int64)\n counts = rows - left_rows + 1\n organic_sum = np.cumsum(organic.astype(np.float64))\n completion_sum = np.cumsum(np.minimum(completion, 100).astype(np.float64))\n before_organic = np.where(\n left_rows > 0, organic_sum[np.maximum(left_rows - 1, 0)], 0.0\n )\n before_completion = np.where(\n left_rows > 0, completion_sum[np.maximum(left_rows - 1, 0)], 0.0\n )\n\n unique_counts = np.empty(length, dtype=np.int32)\n frequencies: dict[int, int] = {}\n left = 0\n for right, raw_item in enumerate(items):\n item = int(raw_item)\n frequencies[item] = frequencies.get(item, 0) + 1\n required_left = int(left_rows[right])\n while left < required_left:\n old_item = int(items[left])\n remaining = frequencies[old_item] - 1\n if remaining:\n frequencies[old_item] = remaining\n else:\n del frequencies[old_item]\n left += 1\n unique_counts[right] = len(frequencies)\n\n result[:, 0] = (np.log1p(counts) / 8.0).astype(np.float16)\n result[:, 1] = (np.log1p(unique_counts) / 6.0).astype(np.float16)\n result[:, 2] = (\n (organic_sum - before_organic) / np.maximum(counts, 1)\n ).astype(np.float16)\n result[:, 3] = (\n (completion_sum - before_completion)\n / np.maximum(counts, 1)\n / 100.0\n ).astype(np.float16)\n return result\n\n\ndef build_sequence_parts(\n listens: base.ListenEvents,\n explicit: base.ExplicitEvents,\n n_items: int,\n liked_states: list[np.ndarray],\n disliked_states: list[np.ndarray],\n state_counts: list[tuple[int, int]],\n) -> tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray, np.ndarray] | None:\n qualifying = listens.completion_pct >= base.LISTEN_POSITIVE_THRESHOLD\n timestamps = listens.timestamps[qualifying].astype(np.int64, copy=False)\n items = listens.items[qualifying].astype(np.uint16, copy=False)\n organic = listens.organic[qualifying].astype(np.float32, copy=False)\n completion = listens.completion_pct[qualifying].astype(np.float32, copy=False)\n if not len(items):\n return None\n\n event_features = np.zeros((len(items), 3), dtype=np.float16)\n event_features[:, 0] = organic.astype(np.float16)\n event_features[:, 1] = (\n np.minimum(completion, 100.0) / 100.0\n ).astype(np.float16)\n if len(items) > 1:\n gaps = np.maximum(0, timestamps[1:] - timestamps[:-1])\n event_features[1:, 2] = np.exp(\n -gaps.astype(np.float64) / (60.0 * base.DAY)\n ).astype(np.float16)\n\n activity = causal_activity_features(\n timestamps, items, organic, completion\n )\n states = np.zeros(len(items), dtype=np.uint32)\n active_likes: dict[int, int] = {}\n active_dislikes: dict[int, int] = {}\n explicit_position = 0\n current_state = 0\n for row, timestamp in enumerate(timestamps):\n changed = False\n while (\n explicit_position < len(explicit.timestamps)\n and int(explicit.timestamps[explicit_position]) < int(timestamp)\n ):\n item = int(explicit.items[explicit_position])\n event_timestamp = int(explicit.timestamps[explicit_position])\n destination = (\n active_likes\n if int(explicit.channel[explicit_position]) == 0\n else active_dislikes\n )\n if int(explicit.active[explicit_position]) == 1:\n destination[item] = event_timestamp\n else:\n destination.pop(item, None)\n explicit_position += 1\n changed = True\n if changed:\n current_state = append_state_snapshot(\n active_likes,\n active_dislikes,\n n_items,\n liked_states,\n disliked_states,\n state_counts,\n )\n states[row] = np.uint32(current_state)\n return timestamps, items.copy(), event_features, activity, states\n\n\ndef final_user_features(\n timestamps: np.ndarray,\n items: np.ndarray,\n event_features: np.ndarray,\n history_end: int,\n like_count: int,\n dislike_count: int,\n) -> np.ndarray:\n start = int(\n np.searchsorted(\n timestamps,\n history_end - base.PROFILE_LOOKBACK_DAYS * base.DAY,\n side=\"left\",\n )\n )\n result = np.zeros(6, dtype=np.float32)\n recent_items = items[start:]\n if len(recent_items):\n result[0] = np.log1p(len(recent_items)) / 8.0\n result[1] = np.log1p(len(np.unique(recent_items))) / 6.0\n result[2] = float(np.mean(event_features[start:, 0]))\n result[3] = float(np.mean(event_features[start:, 1]))\n result[4] = np.log1p(like_count) / 6.0\n result[5] = np.log1p(dislike_count) / 4.0\n return result\n\n\ndef prepare_data(\n public: Path, stage: str, config: Config\n) -> tuple[SequenceData, CandidateSet]:\n data = public / \"data\"\n catalog = pd.read_parquet(data / \"candidate_catalog.parquet\").sort_values(\n \"popularity_rank\"\n )\n item_ids = catalog.item_id.to_numpy(dtype=np.int64, copy=True)\n item_index = pd.Index(item_ids)\n n_items = len(item_ids)\n if n_items >= np.iinfo(np.uint16).max:\n raise RuntimeError(\"uint16 item storage is too small for this catalog\")\n\n if stage == \"validation\":\n prefixes = (\"train\",)\n history_end = base.TRAIN_END\n candidate_name = \"validation_eligible_candidates.npz\"\n state_name = \"validation_preference_state.parquet\"\n else:\n prefixes = (\"train\", \"val\")\n history_end = base.VALIDATION_END\n candidate_name = \"test_eligible_candidates.npz\"\n state_name = \"test_preference_state.parquet\"\n eligibility = CandidateSet.read(data / candidate_name)\n if not np.array_equal(eligibility.item_ids, item_ids):\n raise RuntimeError(\"catalog and candidate artifact item order differ\")\n\n explicit_by_user = base.load_explicit_events(data, prefixes, item_index)\n inference_likes, inference_dislikes, like_counts, dislike_counts = (\n base.load_inference_state(\n data / state_name,\n item_index,\n eligibility.uids,\n n_items,\n )\n )\n inference_rows = {int(uid): row for row, uid in enumerate(eligibility.uids)}\n inference_user_features = np.zeros((len(eligibility.uids), 6), dtype=np.float32)\n inference_sequence_rows = np.full(len(eligibility.uids), -1, dtype=np.int32)\n\n # Snapshot zero is the shared empty explicit state.\n liked_states = [np.full(base.LIKE_ITEMS, n_items, dtype=np.uint16)]\n disliked_states = [\n np.full(base.DISLIKE_ITEMS, n_items, dtype=np.uint16)\n ]\n state_counts: list[tuple[int, int]] = [(0, 0)]\n user_ids: list[int] = []\n offsets = [0]\n timestamp_parts: list[np.ndarray] = []\n item_parts: list[np.ndarray] = []\n event_parts: list[np.ndarray] = []\n activity_parts: list[np.ndarray] = []\n state_parts: list[np.ndarray] = []\n\n listen_paths = [data / f\"{prefix}_listens.parquet\" for prefix in prefixes]\n started = time.monotonic()\n for seen_users, (uid, listens) in enumerate(\n base.merge_listen_streams(listen_paths, item_index), start=1\n ):\n parts = build_sequence_parts(\n listens,\n explicit_by_user.get(uid, base.empty_explicit_events()),\n n_items,\n liked_states,\n disliked_states,\n state_counts,\n )\n if parts is None:\n continue\n sequence_row = len(user_ids)\n user_ids.append(uid)\n timestamp_parts.append(parts[0].astype(np.uint32, copy=False))\n item_parts.append(parts[1])\n event_parts.append(parts[2])\n activity_parts.append(parts[3])\n state_parts.append(parts[4])\n offsets.append(offsets[-1] + len(parts[1]))\n inference_row = inference_rows.get(uid)\n if inference_row is not None:\n inference_sequence_rows[inference_row] = sequence_row\n inference_user_features[inference_row] = final_user_features(\n parts[0],\n parts[1],\n parts[2],\n history_end,\n int(like_counts[inference_row]),\n int(dislike_counts[inference_row]),\n )\n if seen_users % 1_000 == 0:\n log(\n f\"prepared users={seen_users:,} tokens={offsets[-1]:,} \"\n f\"states={len(liked_states):,} \"\n f\"elapsed={time.monotonic() - started:.1f}s\"\n )\n\n missing = np.flatnonzero(inference_sequence_rows < 0)\n if len(missing):\n inference_user_features[missing, 4] = np.log1p(like_counts[missing]) / 6.0\n inference_user_features[missing, 5] = (\n np.log1p(dislike_counts[missing]) / 4.0\n )\n log(f\"target users without qualifying catalog sequence={len(missing):,}\")\n\n prepared = SequenceData(\n item_ids=item_ids,\n item_features=base.build_item_features(catalog),\n user_ids=np.asarray(user_ids, dtype=np.int64),\n offsets=np.asarray(offsets, dtype=np.int64),\n sequence_timestamps=np.concatenate(timestamp_parts),\n sequence_items=np.concatenate(item_parts),\n event_features=np.concatenate(event_parts),\n activity_features=np.concatenate(activity_parts),\n state_ids=np.concatenate(state_parts),\n liked_states=np.stack(liked_states),\n disliked_states=np.stack(disliked_states),\n state_counts=np.asarray(state_counts, dtype=np.int32),\n inference_user_rows=inference_sequence_rows,\n inference_liked_items=inference_likes,\n inference_disliked_items=inference_dislikes,\n inference_user_features=inference_user_features,\n )\n log(\n f\"prepared stage={stage} sequence_users={len(prepared.user_ids):,} \"\n f\"tokens={len(prepared.sequence_items):,} \"\n f\"next_item_targets={prepared.target_count:,} \"\n f\"states={len(prepared.liked_states):,}\"\n )\n return prepared, eligibility\n\n\ndef cache_metadata(stage: str, config: Config) -> dict[str, object]:\n return {\n \"cache_version\": CACHE_VERSION,\n \"stage\": stage,\n \"positive_listen_threshold\": base.LISTEN_POSITIVE_THRESHOLD,\n \"profile_lookback_days\": base.PROFILE_LOOKBACK_DAYS,\n \"event_time_feature\": \"exp-negative-gap-over-60-days\",\n }\n\n\ndef save_cache(\n path: Path, prepared: SequenceData, stage: str, config: Config\n) -> None:\n path.parent.mkdir(parents=True, exist_ok=True)\n np.savez(\n path,\n metadata=np.asarray(json.dumps(cache_metadata(stage, config), sort_keys=True)),\n **prepared.__dict__,\n )\n log(f\"saved public-derived full-sequence cache: {path}\")\n\n\ndef load_cache(path: Path, stage: str, config: Config) -> SequenceData:\n with np.load(path, allow_pickle=False) as artifact:\n actual = json.loads(str(artifact[\"metadata\"].item()))\n expected = cache_metadata(stage, config)\n if actual != expected:\n raise RuntimeError(f\"sequence cache differs: {actual} != {expected}\")\n values = {\n field: artifact[field].copy()\n for field in SequenceData.__dataclass_fields__\n }\n prepared = SequenceData(**values)\n log(\n f\"loaded full-sequence cache: tokens={len(prepared.sequence_items):,} \"\n f\"targets={prepared.target_count:,}\"\n )\n return prepared\n\n\ndef load_or_prepare(\n public: Path,\n stage: str,\n config: Config,\n cache: Path | None,\n) -> tuple[SequenceData, CandidateSet]:\n candidate_name = (\n \"validation_eligible_candidates.npz\"\n if stage == \"validation\"\n else \"test_eligible_candidates.npz\"\n )\n eligibility = CandidateSet.read(public / \"data\" / candidate_name)\n if cache is not None and cache.is_file():\n prepared = load_cache(cache, stage, config)\n if not np.array_equal(prepared.item_ids, eligibility.item_ids):\n raise RuntimeError(\"cached catalog differs from candidate artifact\")\n return prepared, eligibility\n prepared, built_eligibility = prepare_data(public, stage, config)\n if cache is not None:\n save_cache(cache, prepared, stage, config)\n return prepared, built_eligibility\n\n\ndef build_item_correction(\n prepared: SequenceData,\n config: Config,\n target_starts: np.ndarray,\n target_counts: np.ndarray,\n) -> torch.Tensor:\n \"\"\"Estimate in-batch propensities from the selected training targets.\"\"\"\n counts = np.zeros(len(prepared.item_ids), dtype=np.float64)\n for user_row, count in enumerate(target_counts):\n if not count:\n continue\n offset = int(prepared.offsets[user_row])\n target_start = int(target_starts[user_row])\n counts += np.bincount(\n prepared.sequence_items[\n offset\n + target_start\n + 1 : offset\n + target_start\n + int(count)\n + 1\n ].astype(np.int64),\n minlength=len(prepared.item_ids),\n )\n total = float(counts.sum())\n probabilities = (counts + 1.0) / (total + len(counts))\n raw = np.power(\n (1.0 / len(counts)) / probabilities,\n config.inbatch_correction_power,\n )\n normalization = float(np.sum(counts * raw) / max(total, 1.0))\n correction = np.clip(\n raw / max(normalization, 1.0e-12),\n config.inbatch_correction_min,\n config.inbatch_correction_max,\n ).astype(np.float32)\n mean_on_targets = float(np.sum(counts * correction) / max(total, 1.0))\n log(\n f\"inbatch correction: min={correction.min():.3f} \"\n f\"mean_on_targets={mean_on_targets:.3f} max={correction.max():.3f}\"\n )\n return torch.from_numpy(correction)\n\n\ndef build_known_positive_filter(\n prepared: SequenceData, training_counts: np.ndarray\n) -> np.ndarray:\n \"\"\"Pack each user's complete public-training positive-item set.\"\"\"\n width = (len(prepared.item_ids) + 7) // 8\n packed = np.zeros((len(prepared.user_ids), width), dtype=np.uint8)\n retained = 0\n for user_row, target_count in enumerate(training_counts):\n if not target_count:\n continue\n offset = int(prepared.offsets[user_row])\n # T adjacent targets consume sequence positions 0 through T. The\n # chronological BCE suffix starts at T + 1 and is intentionally absent.\n items = np.unique(\n prepared.sequence_items[\n offset : offset + int(target_count) + 1\n ].astype(np.int64)\n )\n byte_columns = items >> 3\n bit_values = np.left_shift(\n np.uint8(1), (items & 7).astype(np.uint8)\n )\n np.bitwise_or.at(packed[user_row], byte_columns, bit_values)\n retained += len(items)\n log(\n f\"full-history false-negative filter: users={len(prepared.user_ids):,} \"\n f\"unique_user_items={retained:,} bytes={packed.nbytes:,}\"\n )\n return packed\n\n\ndef invalid_full_history_negatives(\n negatives: np.ndarray,\n targets: np.ndarray,\n user_rows: np.ndarray,\n known_positive_packed: np.ndarray,\n liked_items: np.ndarray,\n) -> np.ndarray:\n \"\"\"Return row-level invalid negatives using the complete training prefix.\"\"\"\n byte_columns = negatives >> 3\n bit_columns = negatives & 7\n known = known_positive_packed[user_rows[:, None], byte_columns]\n invalid = ((known >> bit_columns) & 1).astype(bool)\n invalid |= negatives == targets[:, None]\n invalid |= (negatives[:, :, None] == liked_items[:, None, :]).any(axis=2)\n for column in range(1, negatives.shape[1]):\n invalid[:, column] |= (\n negatives[:, column, None] == negatives[:, :column]\n ).any(axis=1)\n return invalid\n\n\ndef draw_full_history_filtered_negatives(\n targets: torch.Tensor,\n user_rows: np.ndarray,\n liked_items: torch.Tensor,\n known_positive_packed: np.ndarray,\n n_items: int,\n config: Config,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n \"\"\"Draw mixed negatives and reject all known public-training positives.\"\"\"\n targets_cpu = targets.detach().cpu().numpy().astype(np.int64, copy=False)\n liked_cpu = liked_items.detach().cpu().numpy().astype(np.int64, copy=False)\n batch_size = len(targets_cpu)\n uniform = torch.randint(\n n_items,\n (batch_size, config.uniform_negatives),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.empty((batch_size, 0), dtype=torch.long)\n corrected_count = 0\n if config.inbatch_negatives:\n unique_targets = torch.unique(targets.detach().cpu())\n corrected_count = min(\n config.inbatch_negatives, max(len(unique_targets) - 1, 0)\n )\n if corrected_count:\n priorities = torch.rand(\n (batch_size, len(unique_targets)), generator=generator\n )\n priorities.masked_fill_(\n unique_targets[None, :].eq(\n torch.from_numpy(targets_cpu)[:, None]\n ),\n -1.0,\n )\n columns = torch.topk(\n priorities, k=corrected_count, dim=1\n ).indices\n inbatch = unique_targets[columns]\n if corrected_count < config.inbatch_negatives:\n filler = torch.randint(\n n_items,\n (batch_size, config.inbatch_negatives - corrected_count),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.cat([inbatch, filler], dim=1)\n\n negatives = torch.cat([uniform, inbatch], dim=1).numpy()\n correction = np.ones_like(negatives, dtype=np.float32)\n if corrected_count:\n correction[\n :, config.uniform_negatives : config.uniform_negatives + corrected_count\n ] = item_correction[inbatch[:, :corrected_count]].numpy()\n for _ in range(32):\n invalid = invalid_full_history_negatives(\n negatives,\n targets_cpu,\n user_rows,\n known_positive_packed,\n liked_cpu,\n )\n if not np.any(invalid):\n return (\n torch.from_numpy(negatives).to(device),\n torch.from_numpy(correction).to(device),\n )\n replacements = torch.randint(\n n_items,\n (int(invalid.sum()),),\n generator=generator,\n device=\"cpu\",\n ).numpy()\n negatives[invalid] = replacements\n # Rejected in-batch items become ordinary uniform draws.\n correction[invalid] = 1.0\n invalid = invalid_full_history_negatives(\n negatives,\n targets_cpu,\n user_rows,\n known_positive_packed,\n liked_cpu,\n )\n if np.any(invalid):\n raise RuntimeError(\"full-history negative resampling failed\")\n return (\n torch.from_numpy(negatives).to(device),\n torch.from_numpy(correction).to(device),\n )\n\n\ndef draw_training_negatives(\n positive_items: torch.Tensor,\n known_history: torch.Tensor,\n liked_items: torch.Tensor,\n user_rows: np.ndarray,\n known_positive_packed: np.ndarray | None,\n prepared: SequenceData,\n config: Config,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n if config.negative_filter_scope == \"full-training-history\":\n if known_positive_packed is None:\n raise RuntimeError(\"full-history negative filter was not built\")\n return draw_full_history_filtered_negatives(\n positive_items,\n user_rows,\n liked_items,\n known_positive_packed,\n len(prepared.item_ids),\n config,\n item_correction,\n generator,\n device,\n )\n return base.draw_filtered_mixed_negatives(\n positive_items,\n known_history,\n liked_items,\n len(prepared.item_ids),\n config.uniform_negatives,\n config.inbatch_negatives,\n item_correction,\n generator,\n device,\n )\n\n\ndef event_sequence(\n model: base.SequentialTwoTower,\n items: torch.Tensor,\n features: torch.Tensor,\n hidden: torch.Tensor,\n) -> tuple[torch.Tensor, torch.Tensor]:\n inputs = torch.cat([model.context_embedding(items), features], dim=-1)\n projected = model.event_projection(inputs)\n return model.sequence_encoder(projected, hidden)\n\n\ndef pool_explicit_states(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n state_ids: np.ndarray,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor, torch.Tensor]:\n unique_states, inverse = np.unique(state_ids, return_inverse=True)\n liked_items = torch.from_numpy(\n prepared.liked_states[unique_states].astype(np.int64)\n ).to(device)\n disliked_items = torch.from_numpy(\n prepared.disliked_states[unique_states].astype(np.int64)\n ).to(device)\n liked_vectors = model._mean_context_channel(liked_items)\n disliked_vectors = model._mean_context_channel(disliked_items)\n inverse_rows = torch.from_numpy(inverse.astype(np.int64)).to(device)\n counts = prepared.state_counts[state_ids]\n return (\n liked_vectors[inverse_rows],\n disliked_vectors[inverse_rows],\n torch.from_numpy(counts.astype(np.float32)).to(device),\n )\n\n\ndef user_queries(\n model: base.SequentialTwoTower,\n sequence_vectors: torch.Tensor,\n liked_vectors: torch.Tensor,\n disliked_vectors: torch.Tensor,\n user_features: torch.Tensor,\n) -> torch.Tensor:\n return model.user_tower(\n torch.cat(\n [sequence_vectors, liked_vectors, disliked_vectors, user_features],\n dim=1,\n )\n )\n\n\ndef candidate_vectors(\n model: base.SequentialTwoTower,\n positives: torch.Tensor,\n negatives: torch.Tensor,\n) -> tuple[torch.Tensor, torch.Tensor]:\n all_items = torch.cat([positives, negatives.reshape(-1)])\n unique_items, inverse = torch.unique(all_items, return_inverse=True)\n encoded = model.encode_items(unique_items)\n positive_vectors = encoded[inverse[: len(positives)]]\n negative_vectors = encoded[inverse[len(positives) :]].reshape(\n len(positives), negatives.shape[1], -1\n )\n return positive_vectors, negative_vectors\n\n\ndef history_windows(\n sequence: np.ndarray,\n start: int,\n count: int,\n width: int,\n padding_item: int,\n) -> np.ndarray:\n base_start = max(0, start - width + 1)\n segment = sequence[base_start : start + count]\n padded = np.concatenate(\n [np.full(width - 1, padding_item, dtype=np.uint16), segment]\n )\n windows = np.lib.stride_tricks.sliding_window_view(padded, width)\n offset = start - base_start\n return windows[offset : offset + count].copy()\n\n\ndef corrected_negative_loss(\n logits: torch.Tensor,\n correction: torch.Tensor,\n config: Config,\n) -> torch.Tensor:\n raw = F.softplus(logits)\n total = config.uniform_negatives + config.inbatch_negatives\n result = torch.zeros(len(logits), device=logits.device)\n if config.uniform_negatives:\n result += config.uniform_negatives * raw[\n :, : config.uniform_negatives\n ].mean(dim=1)\n if config.inbatch_negatives:\n selected = slice(config.uniform_negatives, total)\n weights = correction[:, selected]\n result += config.inbatch_negatives * (\n (raw[:, selected] * weights).sum(dim=1)\n / weights.sum(dim=1).clamp_min(1.0e-6)\n )\n return result / total\n\n\ndef retained_target_window_starts(\n total_targets: np.ndarray,\n validation_counts: np.ndarray,\n training_counts: np.ndarray,\n) -> np.ndarray:\n \"\"\"Return the discarded prefix before the retained chronological suffix.\"\"\"\n training_region_ends = total_targets - validation_counts\n window_starts = training_region_ends - training_counts\n if (\n np.any(window_starts < 0)\n or np.any(training_counts < 0)\n or np.any(validation_counts < 0)\n ):\n raise RuntimeError(\"chronological target partition produced negative counts\")\n if not np.array_equal(\n window_starts + training_counts + validation_counts, total_targets\n ):\n raise RuntimeError(\"chronological target window is not a final contiguous suffix\")\n return window_starts\n\n\ndef chronological_target_partition(\n prepared: SequenceData,\n validation_fraction: float,\n max_train_targets_per_user: int,\n max_validation_targets_per_user: int,\n) -> TargetPartition:\n \"\"\"Select a recent contiguous window with a final out-of-time BCE suffix.\"\"\"\n totals = np.maximum(prepared.lengths - 1, 0).astype(np.int64)\n validation = np.floor(totals * validation_fraction).astype(np.int64)\n validation[(totals >= 2) & (validation == 0)] = 1\n if max_validation_targets_per_user:\n validation = np.minimum(validation, max_validation_targets_per_user)\n\n history_training = totals - validation\n training = history_training.copy()\n if max_train_targets_per_user:\n training = np.minimum(training, max_train_targets_per_user)\n window_starts = retained_target_window_starts(totals, validation, training)\n return TargetPartition(\n window_starts=window_starts,\n training_counts=training,\n validation_counts=validation,\n history_training_counts=history_training,\n )\n\n\ndef load_organic_like_outcomes(\n public: Path, stage: str, item_ids: np.ndarray\n) -> dict[int, dict[int, np.ndarray]]:\n \"\"\"Load mapped organic-like timestamps from public fitting history.\"\"\"\n prefixes = (\"train\",) if stage == \"validation\" else (\"train\", \"val\")\n frames = [\n pd.read_parquet(\n public / \"data\" / f\"{prefix}_likes.parquet\",\n columns=[\"uid\", \"timestamp\", \"item_id\", \"is_organic\"],\n )\n for prefix in prefixes\n ]\n likes = pd.concat(frames, ignore_index=True)\n likes = likes[likes.is_organic == 1].copy()\n mapped = pd.Index(item_ids).get_indexer(\n likes.item_id.to_numpy(dtype=np.int64)\n )\n likes = likes.loc[mapped >= 0, [\"uid\", \"timestamp\"]].copy()\n likes[\"item\"] = mapped[mapped >= 0].astype(np.int32, copy=False)\n likes = likes.sort_values([\"uid\", \"item\", \"timestamp\"], kind=\"stable\")\n nested: dict[int, dict[int, list[int]]] = {}\n for raw_uid, raw_timestamp, raw_item in likes.itertuples(\n index=False, name=None\n ):\n nested.setdefault(int(raw_uid), {}).setdefault(int(raw_item), []).append(\n int(raw_timestamp)\n )\n result = {\n uid: {\n item: np.asarray(timestamps, dtype=np.int64)\n for item, timestamps in items.items()\n }\n for uid, items in nested.items()\n }\n log(\n f\"loaded organic-like outcomes: rows={len(likes):,} \"\n f\"users={len(result):,}\"\n )\n return result\n\n\ndef build_target_weights(\n public: Path,\n stage: str,\n prepared: SequenceData,\n training_counts: np.ndarray,\n config: Config,\n) -> np.ndarray:\n \"\"\"Build nonzero per-token weights without dropping adjacent targets.\"\"\"\n if config.target_weighting == \"uniform\":\n weights = np.ones(len(prepared.sequence_items), dtype=np.float32)\n else:\n weights = np.full(\n len(prepared.sequence_items),\n config.ordinary_target_weight,\n dtype=np.float32,\n )\n organic_strong = (\n (prepared.event_features[:, 0] > 0.5)\n & (prepared.event_features[:, 1] >= 0.8)\n )\n weights[organic_strong] = config.organic_strong_target_weight\n if config.target_weighting == \"discovery-graded\":\n like_outcomes = load_organic_like_outcomes(\n public, stage, prepared.item_ids\n )\n history_end = (\n base.TRAIN_END if stage == \"validation\" else base.VALIDATION_END\n )\n horizon = config.discovery_horizon_days * base.DAY\n grade_three = 0\n for user_row, uid in enumerate(prepared.user_ids):\n start = int(prepared.offsets[user_row])\n end = int(prepared.offsets[user_row + 1])\n if end - start < 2:\n continue\n last_training_target = start + int(training_counts[user_row])\n first_validation_target = last_training_target + 1\n split_timestamp = (\n int(prepared.sequence_timestamps[first_validation_target])\n if first_validation_target < end\n else history_end\n )\n user_likes = like_outcomes.get(int(uid), {})\n seen = {int(prepared.sequence_items[start])}\n for position in range(start + 1, end):\n item = int(prepared.sequence_items[position])\n novel = item not in seen\n seen.add(item)\n if not novel:\n # Repeats remain training examples, but they cannot be\n # eligible discovery outcomes at the corresponding\n # cutoff and therefore keep only the ordinary weight.\n weights[position] = config.ordinary_target_weight\n continue\n timestamps = user_likes.get(item)\n if timestamps is None:\n continue\n timestamp = int(prepared.sequence_timestamps[position])\n upper = min(\n timestamp + horizon,\n split_timestamp\n if position <= last_training_target\n else history_end,\n )\n outcome = int(np.searchsorted(timestamps, timestamp, side=\"left\"))\n if outcome < len(timestamps) and int(timestamps[outcome]) < upper:\n weights[position] = config.organic_like_target_weight\n grade_three += 1\n log(\n f\"grade-3 first-encounter targets={grade_three:,} \"\n f\"horizon_days={config.discovery_horizon_days}\"\n )\n log(\n f\"target weighting={config.target_weighting} \"\n f\"min={weights.min():.3f} mean={weights.mean():.3f} \"\n f\"max={weights.max():.3f}\"\n )\n return weights\n\n\ndef build_dense_block(\n prepared: SequenceData,\n target_weights: np.ndarray,\n users: np.ndarray,\n window_starts: np.ndarray,\n process_counts: np.ndarray,\n score_starts: np.ndarray,\n start: int,\n steps: int,\n config: Config,\n) -> tuple[np.ndarray, ...]:\n \"\"\"Build one causal sequence block and mark the target rows to score.\"\"\"\n batch_size = len(users)\n padding_item = len(prepared.item_ids)\n input_items = np.full((batch_size, steps), padding_item, dtype=np.uint16)\n event_features = np.zeros((batch_size, steps, 3), dtype=np.float16)\n activity_features = np.zeros((batch_size, steps, 4), dtype=np.float16)\n state_ids = np.zeros((batch_size, steps), dtype=np.uint32)\n targets = np.zeros((batch_size, steps), dtype=np.uint16)\n weights = np.ones((batch_size, steps), dtype=np.float32)\n scored = np.zeros((batch_size, steps), dtype=bool)\n histories = np.full(\n (batch_size, steps, config.bptt_steps),\n padding_item,\n dtype=np.uint16,\n )\n processed = np.zeros(batch_size, dtype=np.int32)\n\n lengths = prepared.lengths\n for row, user_row in enumerate(users):\n count = min(steps, max(int(process_counts[row]) - start, 0))\n if not count:\n continue\n offset = int(prepared.offsets[user_row])\n window_start = int(window_starts[row])\n absolute_start = window_start + start\n positions = slice(\n offset + absolute_start,\n offset + absolute_start + count,\n )\n target_positions = slice(\n offset + absolute_start + 1,\n offset + absolute_start + count + 1,\n )\n input_items[row, :count] = prepared.sequence_items[positions]\n event_features[row, :count] = prepared.event_features[positions]\n activity_features[row, :count] = prepared.activity_features[positions]\n state_ids[row, :count] = prepared.state_ids[positions]\n targets[row, :count] = prepared.sequence_items[target_positions]\n weights[row, :count] = target_weights[target_positions]\n score_from = min(count, max(int(score_starts[row]) - start, 0))\n scored[row, score_from:count] = True\n processed[row] = count\n user_sequence = prepared.sequence_items[\n offset : offset + int(lengths[user_row])\n ]\n histories[row, :count] = history_windows(\n user_sequence,\n absolute_start,\n count,\n config.bptt_steps,\n padding_item,\n )\n return (\n input_items,\n event_features,\n activity_features,\n state_ids,\n targets,\n scored,\n histories,\n processed,\n weights,\n )\n\n\ndef score_target_microbatches(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n sequence_vectors: torch.Tensor,\n flat_states: np.ndarray,\n flat_activity: np.ndarray,\n positive_items: torch.Tensor,\n negatives: torch.Tensor,\n negative_weights: torch.Tensor,\n sample_weights: np.ndarray,\n device: torch.device,\n backward: bool,\n) -> tuple[dict[str, float], torch.Tensor | None]:\n \"\"\"Compute one logical block's exact BCE in bounded-memory slices.\"\"\"\n target_count = len(positive_items)\n total_weight = float(sample_weights.sum())\n if total_weight <= 0:\n raise RuntimeError(\"target block has no positive sample weight\")\n sequence_gradient = torch.zeros_like(sequence_vectors) if backward else None\n statistics = {\n \"loss_sum\": 0.0,\n \"positive_sum\": 0.0,\n \"negative_sum\": 0.0,\n \"pair_correct\": 0.0,\n \"weight_sum\": 0.0,\n }\n for begin in range(0, target_count, config.loss_microbatch_size):\n end = min(begin + config.loss_microbatch_size, target_count)\n selected_states = flat_states[begin:end]\n if backward:\n sequence_input = (\n sequence_vectors[begin:end].detach().requires_grad_(True)\n )\n else:\n sequence_input = sequence_vectors[begin:end]\n liked_vectors, disliked_vectors, _ = pool_explicit_states(\n model, prepared, selected_states, device\n )\n counts = prepared.state_counts[selected_states]\n numeric = np.zeros((end - begin, 6), dtype=np.float32)\n numeric[:, :4] = flat_activity[begin:end]\n numeric[:, 4] = np.log1p(counts[:, 0]) / 6.0\n numeric[:, 5] = np.log1p(counts[:, 1]) / 4.0\n query_vectors = user_queries(\n model,\n sequence_input,\n liked_vectors,\n disliked_vectors,\n torch.from_numpy(numeric).to(device),\n )\n positive_vectors, negative_vectors = candidate_vectors(\n model,\n positive_items[begin:end],\n negatives[begin:end],\n )\n positive_logits = (query_vectors * positive_vectors).sum(dim=1)\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n if config.objective == \"infonce\":\n cat_logits = torch.cat(\n [positive_logits.unsqueeze(1), negative_logits], dim=1\n ) / config.temperature\n labels = torch.zeros(\n len(positive_logits), dtype=torch.long, device=device\n )\n raw_loss = F.cross_entropy(cat_logits, labels, reduction=\"none\")\n loss_sum = (selected_weights * raw_loss).sum()\n else:\n positive_loss = F.softplus(-positive_logits)\n negative_loss = corrected_negative_loss(\n negative_logits, negative_weights[begin:end], config\n )\n loss_sum = (\n selected_weights * (positive_loss + negative_loss) / 2.0\n ).sum()\n if not bool(torch.isfinite(loss_sum)):\n raise RuntimeError(\"dense next-item training produced non-finite loss\")\n if backward:\n (loss_sum / total_weight).backward()\n if sequence_input.grad is None:\n raise RuntimeError(\"sequence microbatch did not produce a gradient\")\n sequence_gradient[begin:end] = sequence_input.grad\n statistics[\"loss_sum\"] += float(loss_sum.detach().cpu())\n statistics[\"positive_sum\"] += float(positive_logits.detach().sum().cpu())\n statistics[\"negative_sum\"] += float(negative_logits.detach().sum().cpu())\n statistics[\"pair_correct\"] += float(\n (positive_logits.detach()[:, None] > negative_logits.detach()).sum().cpu()\n )\n statistics[\"weight_sum\"] += float(selected_weights.sum().detach().cpu())\n return statistics, sequence_gradient\n\n\ndef release_accelerator_cache(device: torch.device) -> None:\n if device.type == \"mps\":\n torch.mps.empty_cache()\n\n\ndef run_training_epoch(\n model: base.SequentialTwoTower,\n optimizer: torch.optim.Optimizer,\n prepared: SequenceData,\n config: Config,\n target_weights: np.ndarray,\n window_starts: np.ndarray,\n target_counts: np.ndarray,\n item_correction: torch.Tensor,\n known_positive_packed: np.ndarray | None,\n generator: torch.Generator,\n device: torch.device,\n epoch: int,\n) -> dict[str, float]:\n model.train()\n train_users = np.flatnonzero(target_counts > 0)\n sorted_users = train_users[\n np.argsort(target_counts[train_users], kind=\"stable\")\n ]\n groups = [\n sorted_users[start : start + config.user_batch_size]\n for start in range(0, len(sorted_users), config.user_batch_size)\n ]\n group_order = torch.randperm(len(groups), generator=generator).numpy()\n negative_count = config.uniform_negatives + config.inbatch_negatives\n total_loss = 0.0\n total_targets = 0\n total_weight = 0.0\n total_positive = 0.0\n total_negative = 0.0\n total_pair_correct = 0.0\n batches = 0\n started = time.monotonic()\n\n for group_number in group_order:\n users = groups[int(group_number)]\n group_counts = target_counts[users]\n hidden = torch.zeros(\n 1, len(users), config.embedding_dim, device=device\n )\n max_targets = int(group_counts.max())\n for start in range(0, max_targets, config.bptt_steps):\n steps = min(config.bptt_steps, max_targets - start)\n block = build_dense_block(\n prepared,\n target_weights,\n users,\n window_starts[users],\n group_counts,\n np.zeros(len(users), dtype=np.int64),\n start,\n steps,\n config,\n )\n sequence_outputs, _ = event_sequence(\n model,\n torch.from_numpy(block[0].astype(np.int64)).to(device),\n torch.from_numpy(block[1].astype(np.float32)).to(device),\n hidden,\n )\n hidden = torch.stack(\n [\n sequence_outputs[row, count - 1]\n if count\n else hidden[0, row]\n for row, count in enumerate(block[7])\n ]\n ).unsqueeze(0).detach()\n flat_scored = block[5].reshape(-1)\n if not np.any(flat_scored):\n continue\n mask = torch.from_numpy(flat_scored).to(device)\n sequence_vectors = sequence_outputs.reshape(\n -1, config.embedding_dim\n )[mask]\n flat_states = block[3].reshape(-1)[flat_scored]\n positive_items = torch.from_numpy(\n block[4].reshape(-1)[flat_scored].astype(np.int64)\n ).to(device)\n known_history = torch.from_numpy(\n block[6].reshape(-1, config.bptt_steps)[flat_scored].astype(\n np.int64\n )\n ).to(device)\n liked_items = torch.from_numpy(\n prepared.liked_states[flat_states].astype(np.int64)\n ).to(device)\n flat_user_rows = np.broadcast_to(\n users[:, None], block[5].shape\n ).reshape(-1)[flat_scored]\n negatives, negative_weights = draw_training_negatives(\n positive_items,\n known_history,\n liked_items,\n flat_user_rows,\n known_positive_packed,\n prepared,\n config,\n item_correction,\n generator,\n device,\n )\n optimizer.zero_grad(set_to_none=True)\n statistics, sequence_gradient = score_target_microbatches(\n model,\n prepared,\n config,\n sequence_vectors,\n flat_states,\n block[2].reshape(-1, 4)[flat_scored].astype(np.float32),\n positive_items,\n negatives,\n negative_weights,\n block[8].reshape(-1)[flat_scored],\n device,\n backward=True,\n )\n if sequence_gradient is None:\n raise RuntimeError(\"training block omitted its sequence gradient\")\n sequence_vectors.backward(sequence_gradient)\n nn.utils.clip_grad_norm_(model.parameters(), config.gradient_clip)\n optimizer.step()\n\n count = len(positive_items)\n total_loss += statistics[\"loss_sum\"]\n total_targets += count\n total_weight += statistics[\"weight_sum\"]\n total_positive += statistics[\"positive_sum\"]\n total_negative += statistics[\"negative_sum\"]\n total_pair_correct += statistics[\"pair_correct\"]\n batches += 1\n if batches % 25 == 0:\n release_accelerator_cache(device)\n if batches % 250 == 0:\n batch_bce = statistics[\"loss_sum\"] / max(\n statistics[\"weight_sum\"], 1.0e-12\n )\n running_bce = total_loss / max(total_weight, 1.0e-12)\n log(\n f\"epoch={epoch}/{config.epochs} batches={batches:,} \"\n f\"targets={total_targets:,} \"\n f\"batch_bce={batch_bce:.6f} \"\n f\"running_bce={running_bce:.6f} \"\n f\"elapsed={time.monotonic() - started:.1f}s\"\n )\n\n return {\n \"epoch\": float(epoch),\n \"loss\": total_loss / max(total_weight, 1.0e-12),\n \"positive_logit\": total_positive / max(total_targets, 1),\n \"negative_logit\": total_negative / max(total_targets * negative_count, 1),\n \"pair_accuracy\": total_pair_correct\n / max(total_targets * negative_count, 1),\n \"targets\": float(total_targets),\n \"target_weight_sum\": total_weight,\n \"batches\": float(batches),\n \"seconds\": time.monotonic() - started,\n }\n\n\n@torch.no_grad()\ndef evaluate_validation_bce(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n target_weights: np.ndarray,\n window_starts: np.ndarray,\n training_counts: np.ndarray,\n validation_counts: np.ndarray,\n item_correction: torch.Tensor,\n known_positive_packed: np.ndarray | None,\n device: torch.device,\n) -> dict[str, float]:\n \"\"\"Measure deterministic BCE on each user's chronological target suffix.\"\"\"\n model.eval()\n totals = training_counts + validation_counts\n users_with_validation = np.flatnonzero(validation_counts > 0)\n ordered = users_with_validation[\n np.argsort(totals[users_with_validation], kind=\"stable\")\n ]\n groups = [\n ordered[start : start + config.user_batch_size]\n for start in range(0, len(ordered), config.user_batch_size)\n ]\n generator = torch.Generator(device=\"cpu\").manual_seed(config.seed + 20_000)\n total_loss = 0.0\n total_targets = 0\n total_weight = 0.0\n started = time.monotonic()\n\n for group_number, users in enumerate(groups, start=1):\n group_totals = totals[users]\n group_starts = training_counts[users]\n hidden = torch.zeros(\n 1, len(users), config.embedding_dim, device=device\n )\n max_targets = int(group_totals.max())\n for start in range(0, max_targets, config.bptt_steps):\n steps = min(config.bptt_steps, max_targets - start)\n block = build_dense_block(\n prepared,\n target_weights,\n users,\n window_starts[users],\n group_totals,\n group_starts,\n start,\n steps,\n config,\n )\n sequence_outputs, _ = event_sequence(\n model,\n torch.from_numpy(block[0].astype(np.int64)).to(device),\n torch.from_numpy(block[1].astype(np.float32)).to(device),\n hidden,\n )\n hidden = torch.stack(\n [\n sequence_outputs[row, count - 1]\n if count\n else hidden[0, row]\n for row, count in enumerate(block[7])\n ]\n ).unsqueeze(0)\n flat_scored = block[5].reshape(-1)\n if not np.any(flat_scored):\n continue\n mask = torch.from_numpy(flat_scored).to(device)\n sequence_vectors = sequence_outputs.reshape(\n -1, config.embedding_dim\n )[mask]\n flat_states = block[3].reshape(-1)[flat_scored]\n positive_items = torch.from_numpy(\n block[4].reshape(-1)[flat_scored].astype(np.int64)\n ).to(device)\n known_history = torch.from_numpy(\n block[6].reshape(-1, config.bptt_steps)[flat_scored].astype(\n np.int64\n )\n ).to(device)\n liked_items = torch.from_numpy(\n prepared.liked_states[flat_states].astype(np.int64)\n ).to(device)\n flat_user_rows = np.broadcast_to(\n users[:, None], block[5].shape\n ).reshape(-1)[flat_scored]\n negatives, negative_weights = draw_training_negatives(\n positive_items,\n known_history,\n liked_items,\n flat_user_rows,\n known_positive_packed,\n prepared,\n config,\n item_correction,\n generator,\n device,\n )\n statistics, _ = score_target_microbatches(\n model,\n prepared,\n config,\n sequence_vectors,\n flat_states,\n block[2].reshape(-1, 4)[flat_scored].astype(np.float32),\n positive_items,\n negatives,\n negative_weights,\n block[8].reshape(-1)[flat_scored],\n device,\n backward=False,\n )\n total_loss += statistics[\"loss_sum\"]\n total_targets += len(positive_items)\n total_weight += statistics[\"weight_sum\"]\n if group_number % 25 == 0:\n release_accelerator_cache(device)\n return {\n \"validation_loss\": total_loss / max(total_weight, 1.0e-12),\n \"validation_targets\": float(total_targets),\n \"validation_target_weight_sum\": total_weight,\n \"validation_seconds\": time.monotonic() - started,\n }\n\n\ndef snapshot_model(model: nn.Module) -> dict[str, torch.Tensor]:\n return {\n name: value.detach().cpu().clone()\n for name, value in model.state_dict().items()\n }\n\n\ndef save_epoch_checkpoint(\n checkpoint_dir: Path,\n epoch: int,\n model_state: dict[str, torch.Tensor],\n config: Config,\n diagnostics: dict[str, float],\n optimizer: torch.optim.Optimizer,\n generator: torch.Generator,\n) -> Path:\n checkpoint_dir.mkdir(parents=True, exist_ok=True)\n path = checkpoint_dir / f\"epoch-{epoch:02d}.pt\"\n torch.save(\n {\n \"epoch\": epoch,\n \"config\": asdict(config),\n \"diagnostics\": diagnostics,\n \"model_state\": model_state,\n \"optimizer_state\": optimizer.state_dict(),\n \"generator_state\": generator.get_state(),\n },\n path,\n )\n log(f\"saved epoch checkpoint: {path}\")\n return path\n\n\ndef train_model(\n public: Path,\n stage: str,\n prepared: SequenceData,\n config: Config,\n device: torch.device,\n checkpoint_dir: Path | None,\n resume_checkpoint: Path | None,\n) -> tuple[base.SequentialTwoTower, list[dict[str, float]]]:\n partition = chronological_target_partition(\n prepared,\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n training_counts = partition.training_counts\n validation_counts = partition.validation_counts\n log(\n f\"chronological BCE window: train={int(training_counts.sum()):,} \"\n f\"validation={int(validation_counts.sum()):,} \"\n f\"selected={partition.selected_count:,} \"\n f\"discarded_old_targets={int(partition.window_starts.sum()):,} \"\n f\"all={prepared.target_count:,}\"\n )\n model = base.SequentialTwoTower(prepared.item_features, model_config(config)).to(\n device\n )\n optimizer = torch.optim.AdamW(\n model.parameters(), lr=config.learning_rate, weight_decay=config.weight_decay\n )\n generator = torch.Generator(device=\"cpu\").manual_seed(config.seed + 10_000)\n item_correction = build_item_correction(\n prepared,\n config,\n partition.window_starts,\n training_counts,\n )\n target_weights = build_target_weights(\n public,\n stage,\n prepared,\n partition.history_training_counts,\n config,\n )\n known_positive_packed = (\n build_known_positive_filter(prepared, partition.history_training_counts)\n if config.negative_filter_scope == \"full-training-history\"\n else None\n )\n history: list[dict[str, float]] = []\n best_loss = float(\"inf\")\n best_epoch = 0\n best_state: dict[str, torch.Tensor] | None = None\n epochs_without_improvement = 0\n start_epoch = 0\n\n if resume_checkpoint is not None:\n payload = torch.load(\n resume_checkpoint, map_location=device, weights_only=False\n )\n stored_config = payload.get(\"config\", {})\n architecture_fields = (\n \"bptt_steps\",\n \"embedding_dim\",\n \"hidden_dim\",\n \"user_hidden_1\",\n \"user_hidden_2\",\n \"item_hidden_1\",\n \"item_hidden_2\",\n )\n mismatches = {\n name: (stored_config.get(name), getattr(config, name))\n for name in architecture_fields\n if stored_config.get(name) != getattr(config, name)\n }\n if mismatches:\n raise RuntimeError(\n f\"resume checkpoint architecture differs: {mismatches}\"\n )\n stored_filter_scope = stored_config.get(\n \"negative_filter_scope\", \"last-100\"\n )\n stored_target_weighting = stored_config.get(\n \"target_weighting\", \"uniform\"\n )\n stored_weight_values = (\n stored_config.get(\"ordinary_target_weight\", 1.0),\n stored_config.get(\"organic_strong_target_weight\", 1.0),\n stored_config.get(\"organic_like_target_weight\", 3.0),\n stored_config.get(\"discovery_horizon_days\", 7),\n )\n current_weight_values = (\n config.ordinary_target_weight,\n config.organic_strong_target_weight,\n config.organic_like_target_weight,\n config.discovery_horizon_days,\n )\n stored_window_values = (\n stored_config.get(\"validation_fraction\", 0.10),\n stored_config.get(\"max_train_targets_per_user\", 0),\n stored_config.get(\"max_validation_targets_per_user\", 0),\n )\n current_window_values = (\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n objective_changed = (\n stored_filter_scope != config.negative_filter_scope\n or stored_target_weighting != config.target_weighting\n or stored_weight_values != current_weight_values\n or stored_window_values != current_window_values\n )\n model.load_state_dict(payload[\"model_state\"])\n if \"optimizer_state\" in payload and not objective_changed:\n optimizer.load_state_dict(payload[\"optimizer_state\"])\n for group in optimizer.param_groups:\n group[\"lr\"] = config.learning_rate\n group[\"weight_decay\"] = config.weight_decay\n log(f\"restored optimizer state from {resume_checkpoint}\")\n else:\n log(\n f\"using a fresh AdamW optimizer after loading {resume_checkpoint}\"\n )\n if \"generator_state\" in payload and not objective_changed:\n generator.set_state(payload[\"generator_state\"])\n start_epoch = int(payload[\"epoch\"])\n if start_epoch >= config.epochs:\n raise RuntimeError(\n f\"resume epoch {start_epoch} must be below --epochs {config.epochs}\"\n )\n prior = {\n str(name): float(value)\n for name, value in payload.get(\"diagnostics\", {}).items()\n }\n if objective_changed:\n prior.pop(\"validation_loss\", None)\n log(\n \"BCE objective changed: filter \"\n f\"{stored_filter_scope}->{config.negative_filter_scope}, \"\n \"target weighting \"\n f\"{stored_target_weighting}->{config.target_weighting}, \"\n f\"target window {stored_window_values}->{current_window_values}; \"\n \"recomputing the resume checkpoint's validation BCE\"\n )\n if \"validation_loss\" not in prior:\n prior.update(\n evaluate_validation_bce(\n model,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n validation_counts,\n item_correction,\n known_positive_packed,\n device,\n )\n )\n history.append(prior)\n best_loss = float(prior[\"validation_loss\"])\n best_epoch = start_epoch\n best_state = snapshot_model(model)\n log(\n f\"resumed validation-BCE training: epoch={start_epoch} \"\n f\"loss={best_loss:.6f}\"\n )\n\n for epoch in range(start_epoch + 1, config.epochs + 1):\n diagnostics = run_training_epoch(\n model,\n optimizer,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n item_correction,\n known_positive_packed,\n generator,\n device,\n epoch,\n )\n validation = evaluate_validation_bce(\n model,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n validation_counts,\n item_correction,\n known_positive_packed,\n device,\n )\n diagnostics.update(validation)\n history.append(diagnostics)\n current_state = snapshot_model(model)\n if checkpoint_dir is not None:\n save_epoch_checkpoint(\n checkpoint_dir,\n epoch,\n current_state,\n config,\n diagnostics,\n optimizer,\n generator,\n )\n improved = (\n diagnostics[\"validation_loss\"]\n < best_loss - config.early_stopping_min_delta\n )\n if improved:\n best_loss = diagnostics[\"validation_loss\"]\n best_epoch = epoch\n best_state = current_state\n epochs_without_improvement = 0\n if checkpoint_dir is not None:\n torch.save(\n {\n \"epoch\": epoch,\n \"config\": asdict(config),\n \"diagnostics\": diagnostics,\n \"model_state\": current_state,\n \"optimizer_state\": optimizer.state_dict(),\n \"generator_state\": generator.get_state(),\n },\n checkpoint_dir / \"best.pt\",\n )\n else:\n epochs_without_improvement += 1\n log(\n f\"epoch={epoch}/{config.epochs} train_loss={diagnostics['loss']:.5f} \"\n f\"validation_loss={diagnostics['validation_loss']:.5f} \"\n f\"pair_accuracy={diagnostics['pair_accuracy']:.3f} \"\n f\"train_targets={int(diagnostics['targets']):,} \"\n f\"validation_targets={int(diagnostics['validation_targets']):,} \"\n f\"train_seconds={diagnostics['seconds']:.1f} \"\n f\"validation_seconds={diagnostics['validation_seconds']:.1f}\"\n )\n release_accelerator_cache(device)\n if epochs_without_improvement >= config.early_stopping_patience:\n log(\n f\"early stopping after epoch={epoch}; \"\n f\"best_epoch={best_epoch} best_validation_loss={best_loss:.6f}\"\n )\n break\n\n if best_state is None:\n raise RuntimeError(\"training did not produce a valid checkpoint\")\n model.load_state_dict(best_state)\n model.to(device).eval()\n log(\n f\"restored best validation-BCE checkpoint: epoch={best_epoch} \"\n f\"loss={best_loss:.6f}\"\n )\n return model, history\n\n\n@torch.inference_mode()\ndef inference_queries(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n device: torch.device,\n) -> torch.Tensor:\n sequence_rows = prepared.inference_user_rows\n lengths = np.where(\n sequence_rows >= 0, prepared.lengths[np.maximum(sequence_rows, 0)], 0\n )\n ordered = np.argsort(lengths, kind=\"stable\")\n result = torch.zeros(\n len(sequence_rows), config.embedding_dim, dtype=torch.float32\n )\n for group_start in range(0, len(ordered), config.user_batch_size):\n output_rows = ordered[group_start : group_start + config.user_batch_size]\n rows = sequence_rows[output_rows]\n group_lengths = lengths[output_rows]\n hidden = torch.zeros(\n 1, len(output_rows), config.embedding_dim, device=device\n )\n for start in range(0, int(group_lengths.max(initial=0)), config.bptt_steps):\n steps = min(\n config.bptt_steps,\n int(group_lengths.max(initial=0)) - start,\n )\n items = np.full(\n (len(output_rows), steps),\n len(prepared.item_ids),\n dtype=np.uint16,\n )\n features = np.zeros((len(output_rows), steps, 3), dtype=np.float16)\n counts = np.zeros(len(output_rows), dtype=np.int32)\n for local_row, sequence_row in enumerate(rows):\n if sequence_row < 0:\n continue\n count = min(steps, max(int(group_lengths[local_row]) - start, 0))\n if not count:\n continue\n offset = int(prepared.offsets[sequence_row])\n selected = slice(offset + start, offset + start + count)\n items[local_row, :count] = prepared.sequence_items[selected]\n features[local_row, :count] = prepared.event_features[selected]\n counts[local_row] = count\n outputs, _ = event_sequence(\n model,\n torch.from_numpy(items.astype(np.int64)).to(device),\n torch.from_numpy(features.astype(np.float32)).to(device),\n hidden,\n )\n next_rows: list[torch.Tensor] = []\n for local_row, count in enumerate(counts):\n next_rows.append(\n outputs[local_row, count - 1]\n if count\n else hidden[0, local_row]\n )\n hidden = torch.stack(next_rows).unsqueeze(0)\n\n liked_items = torch.from_numpy(\n prepared.inference_liked_items[output_rows].astype(np.int64)\n ).to(device)\n disliked_items = torch.from_numpy(\n prepared.inference_disliked_items[output_rows].astype(np.int64)\n ).to(device)\n queries = user_queries(\n model,\n hidden.squeeze(0),\n model._mean_context_channel(liked_items),\n model._mean_context_channel(disliked_items),\n torch.from_numpy(\n prepared.inference_user_features[output_rows].astype(np.float32)\n ).to(device),\n )\n result[torch.from_numpy(output_rows.astype(np.int64))] = queries.cpu()\n return result\n\n\n@torch.inference_mode()\ndef rank_candidates(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n eligibility: CandidateSet,\n config: Config,\n device: torch.device,\n ranking_batch_size: int,\n) -> tuple[pd.DataFrame, dict[str, float]]:\n model.eval()\n queries = inference_queries(model, prepared, config, device)\n item_parts: list[torch.Tensor] = []\n for start in range(0, len(prepared.item_ids), 2_048):\n rows = torch.arange(\n start,\n min(start + 2_048, len(prepared.item_ids)),\n dtype=torch.long,\n device=device,\n )\n item_parts.append(model.encode_items(rows).cpu())\n items = torch.cat(item_parts)\n rows: list[tuple[int, int, int]] = []\n for start in range(0, len(eligibility.uids), ranking_batch_size):\n end = min(start + ranking_batch_size, len(eligibility.uids))\n scores = queries[start:end] @ items.T\n eligible = np.unpackbits(\n eligibility.packed[start:end],\n axis=1,\n count=len(prepared.item_ids),\n bitorder=\"little\",\n ).astype(bool, copy=False)\n scores.masked_fill_(~torch.from_numpy(eligible), -torch.inf)\n selected = torch.topk(scores, k=base.TOP_N, dim=1).indices.numpy()\n for local_row, columns in enumerate(selected):\n uid = int(eligibility.uids[start + local_row])\n rows.extend(\n (uid, int(prepared.item_ids[column]), rank)\n for rank, column in enumerate(columns, start=1)\n )\n if end == len(eligibility.uids) or end % 1_024 == 0:\n log(f\"ranked users={end:,}/{len(eligibility.uids):,}\")\n query_norm = torch.linalg.vector_norm(queries, dim=1).numpy()\n item_norm = torch.linalg.vector_norm(items, dim=1).numpy()\n diagnostics = {\n \"query_norm_mean\": float(query_norm.mean()),\n \"query_norm_std\": float(query_norm.std()),\n \"item_norm_mean\": float(item_norm.mean()),\n \"item_norm_std\": float(item_norm.std()),\n }\n return pd.DataFrame(rows, columns=[\"uid\", \"item_id\", \"rank\"]), diagnostics\n\n\ndef write_ranking(frame: pd.DataFrame, output: Path) -> None:\n output.parent.mkdir(parents=True, exist_ok=True)\n if output.suffix == \".parquet\":\n frame.to_parquet(output, index=False, compression=\"zstd\")\n elif output.suffix == \".csv\":\n frame.to_csv(output, index=False)\n else:\n raise ValueError(\"--output must end in .parquet or .csv\")\n\n\ndef main() -> int:\n parser = argparse.ArgumentParser(description=__doc__)\n parser.add_argument(\n \"--stage\", choices=(\"validation\", \"submission\"), default=\"validation\"\n )\n parser.add_argument(\"--output\", type=Path, required=True)\n parser.add_argument(\"--public-dir\", type=Path, required=True)\n parser.add_argument(\"--prepared-cache\", type=Path)\n parser.add_argument(\"--device\", choices=(\"cpu\", \"mps\"), default=\"cpu\")\n parser.add_argument(\"--threads\", type=int, default=8)\n parser.add_argument(\"--bptt-steps\", type=int, default=Config.bptt_steps)\n parser.add_argument(\n \"--user-batch-size\", type=int, default=Config.user_batch_size\n )\n parser.add_argument(\"--epochs\", type=int, default=Config.epochs)\n parser.add_argument(\"--learning-rate\", type=float, default=Config.learning_rate)\n parser.add_argument(\n \"--loss-microbatch-size\",\n type=int,\n default=Config.loss_microbatch_size,\n )\n parser.add_argument(\n \"--negative-filter-scope\",\n choices=(\"full-training-history\", \"last-100\"),\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n )\n parser.add_argument(\n \"--ordinary-target-weight\",\n type=float,\n default=Config.ordinary_target_weight,\n )\n parser.add_argument(\n \"--organic-strong-target-weight\",\n type=float,\n default=Config.organic_strong_target_weight,\n )\n parser.add_argument(\n \"--organic-like-target-weight\",\n type=float,\n default=Config.organic_like_target_weight,\n )\n parser.add_argument(\n \"--discovery-horizon-days\",\n type=int,\n default=Config.discovery_horizon_days,\n )\n parser.add_argument(\n \"--validation-fraction\",\n type=float,\n default=Config.validation_fraction,\n )\n parser.add_argument(\n \"--max-train-targets-per-user\",\n type=int,\n default=Config.max_train_targets_per_user,\n help=\"retain this many recent pre-validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--max-validation-targets-per-user\",\n type=int,\n default=Config.max_validation_targets_per_user,\n help=\"retain this many final validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--early-stopping-patience\",\n type=int,\n default=Config.early_stopping_patience,\n )\n parser.add_argument(\n \"--early-stopping-min-delta\",\n type=float,\n default=Config.early_stopping_min_delta,\n )\n parser.add_argument(\"--checkpoint-dir\", type=Path)\n parser.add_argument(\"--resume-checkpoint\", type=Path)\n parser.add_argument(\"--seed\", type=int, default=Config.seed)\n parser.add_argument(\"--ranking-batch-size\", type=int, default=64)\n parser.add_argument(\"--prepare-only\", action=\"store_true\")\n args = parser.parse_args()\n\n if args.threads < 1 or args.bptt_steps < 2 or args.user_batch_size < 1:\n parser.error(\"threads, BPTT steps, and user batch size must be positive\")\n if (\n args.epochs < 1\n or args.learning_rate <= 0\n or args.loss_microbatch_size < 1\n or args.early_stopping_patience < 1\n or args.early_stopping_min_delta < 0\n or args.ordinary_target_weight <= 0\n or args.organic_strong_target_weight <= 0\n or args.organic_like_target_weight <= 0\n or args.discovery_horizon_days < 1\n or args.max_train_targets_per_user < 0\n or args.max_validation_targets_per_user < 0\n ):\n parser.error(\"training counts, learning rate, and patience are invalid\")\n if not 0.0 < args.validation_fraction < 1.0:\n parser.error(\"validation fraction must be strictly between zero and one\")\n public = args.public_dir.resolve()\n if not (public / \"data\" / \"candidate_catalog.parquet\").is_file():\n parser.error(f\"public task data not found under {public}\")\n config = Config(\n bptt_steps=args.bptt_steps,\n user_batch_size=args.user_batch_size,\n epochs=args.epochs,\n learning_rate=args.learning_rate,\n loss_microbatch_size=args.loss_microbatch_size,\n negative_filter_scope=args.negative_filter_scope,\n target_weighting=args.target_weighting,\n ordinary_target_weight=args.ordinary_target_weight,\n organic_strong_target_weight=args.organic_strong_target_weight,\n organic_like_target_weight=args.organic_like_target_weight,\n discovery_horizon_days=args.discovery_horizon_days,\n validation_fraction=args.validation_fraction,\n max_train_targets_per_user=args.max_train_targets_per_user,\n max_validation_targets_per_user=args.max_validation_targets_per_user,\n early_stopping_patience=args.early_stopping_patience,\n early_stopping_min_delta=args.early_stopping_min_delta,\n seed=args.seed,\n )\n torch.set_num_threads(args.threads)\n torch.set_num_interop_threads(1)\n seed_everything(config.seed)\n device = torch.device(args.device)\n if device.type == \"mps\" and not torch.backends.mps.is_available():\n parser.error(\"MPS was requested but is unavailable\")\n\n started = time.monotonic()\n prepared, eligibility = load_or_prepare(\n public,\n args.stage,\n config,\n args.prepared_cache.resolve() if args.prepared_cache else None,\n )\n target_partition = chronological_target_partition(\n prepared,\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n if args.prepare_only:\n print(\n json.dumps(\n {\n \"stage\": args.stage,\n \"sequence_users\": len(prepared.user_ids),\n \"sequence_tokens\": len(prepared.sequence_items),\n \"next_item_targets\": prepared.target_count,\n \"selected_training_targets\": int(\n target_partition.training_counts.sum()\n ),\n \"selected_validation_targets\": int(\n target_partition.validation_counts.sum()\n ),\n \"discarded_old_targets\": int(\n target_partition.window_starts.sum()\n ),\n \"target_users\": len(eligibility.uids),\n \"runtime_seconds\": round(time.monotonic() - started, 3),\n \"config\": asdict(config),\n },\n indent=2,\n )\n )\n return 0\n\n model, training_history = train_model(\n public,\n args.stage,\n prepared,\n config,\n device,\n args.checkpoint_dir.resolve() if args.checkpoint_dir else None,\n args.resume_checkpoint.resolve() if args.resume_checkpoint else None,\n )\n ranking, embedding_diagnostics = rank_candidates(\n model,\n prepared,\n eligibility,\n config,\n device,\n args.ranking_batch_size,\n )\n write_ranking(ranking, args.output.resolve())\n print(\n json.dumps(\n {\n \"baseline\": \"yambda-windowed-sequence-two-tower-base\",\n \"stage\": args.stage,\n \"output\": str(args.output.resolve()),\n \"rows\": len(ranking),\n \"users\": int(ranking.uid.nunique()),\n \"runtime_seconds\": round(time.monotonic() - started, 3),\n \"public_inputs_only\": True,\n \"architecture\": {\n \"family\": \"pure-windowed-sequence-two-tower\",\n \"sequence_training\": (\n \"every-adjacent-target-within-recent-chronological-window\"\n ),\n \"sequence_encoder\": \"one-layer-GRU-dense-timestep-loss\",\n \"bptt_hidden_state\": \"carried-and-detached\",\n \"model_selection\": \"chronological-validation-BCE-early-stopping\",\n \"loss_execution\": \"gradient-accumulated-microbatches\",\n \"user_id_embedding\": False,\n \"user_mlp\": \"582-512-256-192\",\n \"item_mlp\": \"196-512-256-192\",\n \"normalization\": \"none\",\n \"score\": \"raw-dot-product\",\n \"objective\": \"sample-weighted-BCE-at-every-timestep\",\n \"negative_sampling\": (\n \"24-uniform-plus-8-propensity-corrected-inbatch\"\n ),\n \"false_negative_filter\": config.negative_filter_scope,\n \"target_weighting\": config.target_weighting,\n },\n \"config\": asdict(config),\n \"sequence_tokens\": len(prepared.sequence_items),\n \"all_next_item_targets\": prepared.target_count,\n \"selected_training_targets\": int(\n target_partition.training_counts.sum()\n ),\n \"selected_validation_targets\": int(\n target_partition.validation_counts.sum()\n ),\n \"discarded_old_targets\": int(target_partition.window_starts.sum()),\n \"training_history\": training_history,\n \"embedding_diagnostics\": embedding_diagnostics,\n \"torch_version\": torch.__version__,\n \"device\": str(device),\n },\n indent=2,\n )\n )\n return 0\n\n\nif __name__ == \"__main__\":\n raise SystemExit(main())\n",
"newContent": "#!/usr/bin/env python3\n\"\"\"Train the Yambda windowed two-tower baseline.\n\nEvery catalog listen with at least 50 percent completion is a sequence token.\nWithin each selected chronological window, every adjacent target contributes a\nloss. The fast default retains each user's most recent 128 training targets and\n16 held-out validation targets; passing zero for either cap restores the\nuncapped behavior for that side of the split. A 100-step truncated-BPTT block\ncomputes all of its timestep losses and carries the detached GRU hidden state\ninto the next block.\n\nThe model is a pure two-tower retriever: no user-ID embedding, no output\nnormalization, raw dot-product retrieval, sample-weighted binary cross entropy,\nand 24 uniform plus eight propensity-corrected unique in-batch negatives. Every\nselected adjacent target retains a nonzero weight. Only files under the supplied\npublic directory are read.\n\"\"\"\nfrom __future__ import annotations\n\nimport argparse\nimport json\nimport random\nimport sys\nimport time\nfrom dataclasses import asdict, dataclass\nfrom pathlib import Path\n\nimport numpy as np\nimport pandas as pd\nimport torch\nfrom torch import nn\nfrom torch.nn import functional as F\n\nimport two_tower_model as base\n\n\nCACHE_VERSION = 2\n\n\n@dataclass(frozen=True)\nclass Config:\n bptt_steps: int = 100\n user_batch_size: int = 64\n embedding_dim: int = 192\n hidden_dim: int = 192\n user_hidden_1: int = 512\n user_hidden_2: int = 256\n item_hidden_1: int = 512\n item_hidden_2: int = 256\n dropout: float = 0.05\n epochs: int = 1\n learning_rate: float = 1.0e-3\n weight_decay: float = 1.0e-5\n gradient_clip: float = 2.0\n loss_microbatch_size: int = 65_536\n uniform_negatives: int = 24\n inbatch_negatives: int = 8\n inbatch_correction_power: float = 0.5\n inbatch_correction_min: float = 0.25\n inbatch_correction_max: float = 4.0\n negative_filter_scope: str = \"full-training-history\"\n objective: str = \"bce\"\n temperature: float = 0.2\n target_weighting: str = \"organic-strong\"\n ordinary_target_weight: float = 0.25\n organic_strong_target_weight: float = 1.0\n organic_like_target_weight: float = 3.0\n discovery_horizon_days: int = 7\n validation_fraction: float = 0.10\n max_train_targets_per_user: int = 128\n max_validation_targets_per_user: int = 16\n early_stopping_patience: int = 1\n early_stopping_min_delta: float = 1.0e-4\n seed: int = 29\n\n\n@dataclass(frozen=True)\nclass CandidateSet:\n uids: np.ndarray\n item_ids: np.ndarray\n packed: np.ndarray\n\n @classmethod\n def read(cls, path: Path) -> \"CandidateSet\":\n with np.load(path, allow_pickle=False) as artifact:\n result = cls(\n uids=artifact[\"uids\"].astype(np.int64, copy=True),\n item_ids=artifact[\"item_ids\"].astype(np.int64, copy=True),\n packed=artifact[\"packed_eligible\"].astype(np.uint8, copy=True),\n )\n expected = (len(result.uids), (len(result.item_ids) + 7) // 8)\n if result.packed.shape != expected:\n raise ValueError(\n f\"invalid candidate artifact shape {result.packed.shape}; \"\n f\"expected {expected}\"\n )\n return result\n\n\n@dataclass(frozen=True)\nclass SequenceData:\n item_ids: np.ndarray\n item_features: np.ndarray\n user_ids: np.ndarray\n offsets: np.ndarray\n sequence_timestamps: np.ndarray\n sequence_items: np.ndarray\n event_features: np.ndarray\n activity_features: np.ndarray\n state_ids: np.ndarray\n liked_states: np.ndarray\n disliked_states: np.ndarray\n state_counts: np.ndarray\n inference_user_rows: np.ndarray\n inference_liked_items: np.ndarray\n inference_disliked_items: np.ndarray\n inference_user_features: np.ndarray\n\n @property\n def lengths(self) -> np.ndarray:\n return np.diff(self.offsets)\n\n @property\n def target_count(self) -> int:\n return int(np.maximum(self.lengths - 1, 0).sum())\n\n\n@dataclass(frozen=True)\nclass TargetPartition:\n \"\"\"A contiguous recent target window and its chronological BCE split.\"\"\"\n\n window_starts: np.ndarray\n training_counts: np.ndarray\n validation_counts: np.ndarray\n history_training_counts: np.ndarray\n\n @property\n def selected_count(self) -> int:\n return int(self.training_counts.sum() + self.validation_counts.sum())\n\n\ndef log(message: str) -> None:\n print(message, file=sys.stderr, flush=True)\n\n\ndef seed_everything(seed: int) -> None:\n random.seed(seed)\n np.random.seed(seed)\n torch.manual_seed(seed)\n\n\ndef model_config(config: Config) -> base.TowerConfig:\n \"\"\"Map the dense trainer configuration onto the unchanged tower.\"\"\"\n return base.TowerConfig(\n sequence_length=config.bptt_steps,\n embedding_dim=config.embedding_dim,\n hidden_dim=config.hidden_dim,\n user_hidden_1=config.user_hidden_1,\n user_hidden_2=config.user_hidden_2,\n item_hidden_1=config.item_hidden_1,\n item_hidden_2=config.item_hidden_2,\n dropout=config.dropout,\n epochs=config.epochs,\n learning_rate=config.learning_rate,\n weight_decay=config.weight_decay,\n gradient_clip=config.gradient_clip,\n uniform_negatives=config.uniform_negatives,\n inbatch_negatives=config.inbatch_negatives,\n inbatch_correction_power=config.inbatch_correction_power,\n inbatch_correction_min=config.inbatch_correction_min,\n inbatch_correction_max=config.inbatch_correction_max,\n seed=config.seed,\n )\n\n\ndef append_state_snapshot(\n active_likes: dict[int, int],\n active_dislikes: dict[int, int],\n n_items: int,\n liked_states: list[np.ndarray],\n disliked_states: list[np.ndarray],\n state_counts: list[tuple[int, int]],\n) -> int:\n liked_states.append(base.pack_active_items(active_likes, base.LIKE_ITEMS, n_items))\n disliked_states.append(\n base.pack_active_items(active_dislikes, base.DISLIKE_ITEMS, n_items)\n )\n state_counts.append((len(active_likes), len(active_dislikes)))\n return len(liked_states) - 1\n\n\ndef causal_activity_features(\n timestamps: np.ndarray,\n items: np.ndarray,\n organic: np.ndarray,\n completion: np.ndarray,\n) -> np.ndarray:\n \"\"\"Compute the first four reviewed user features at every event cutoff.\"\"\"\n length = len(items)\n result = np.zeros((length, 4), dtype=np.float16)\n if not length:\n return result\n left_rows = np.searchsorted(\n timestamps,\n timestamps - base.PROFILE_LOOKBACK_DAYS * base.DAY,\n side=\"left\",\n )\n rows = np.arange(length, dtype=np.int64)\n counts = rows - left_rows + 1\n organic_sum = np.cumsum(organic.astype(np.float64))\n completion_sum = np.cumsum(np.minimum(completion, 100).astype(np.float64))\n before_organic = np.where(\n left_rows > 0, organic_sum[np.maximum(left_rows - 1, 0)], 0.0\n )\n before_completion = np.where(\n left_rows > 0, completion_sum[np.maximum(left_rows - 1, 0)], 0.0\n )\n\n unique_counts = np.empty(length, dtype=np.int32)\n frequencies: dict[int, int] = {}\n left = 0\n for right, raw_item in enumerate(items):\n item = int(raw_item)\n frequencies[item] = frequencies.get(item, 0) + 1\n required_left = int(left_rows[right])\n while left < required_left:\n old_item = int(items[left])\n remaining = frequencies[old_item] - 1\n if remaining:\n frequencies[old_item] = remaining\n else:\n del frequencies[old_item]\n left += 1\n unique_counts[right] = len(frequencies)\n\n result[:, 0] = (np.log1p(counts) / 8.0).astype(np.float16)\n result[:, 1] = (np.log1p(unique_counts) / 6.0).astype(np.float16)\n result[:, 2] = (\n (organic_sum - before_organic) / np.maximum(counts, 1)\n ).astype(np.float16)\n result[:, 3] = (\n (completion_sum - before_completion)\n / np.maximum(counts, 1)\n / 100.0\n ).astype(np.float16)\n return result\n\n\ndef build_sequence_parts(\n listens: base.ListenEvents,\n explicit: base.ExplicitEvents,\n n_items: int,\n liked_states: list[np.ndarray],\n disliked_states: list[np.ndarray],\n state_counts: list[tuple[int, int]],\n) -> tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray, np.ndarray] | None:\n qualifying = listens.completion_pct >= base.LISTEN_POSITIVE_THRESHOLD\n timestamps = listens.timestamps[qualifying].astype(np.int64, copy=False)\n items = listens.items[qualifying].astype(np.uint16, copy=False)\n organic = listens.organic[qualifying].astype(np.float32, copy=False)\n completion = listens.completion_pct[qualifying].astype(np.float32, copy=False)\n if not len(items):\n return None\n\n event_features = np.zeros((len(items), 3), dtype=np.float16)\n event_features[:, 0] = organic.astype(np.float16)\n event_features[:, 1] = (\n np.minimum(completion, 100.0) / 100.0\n ).astype(np.float16)\n if len(items) > 1:\n gaps = np.maximum(0, timestamps[1:] - timestamps[:-1])\n event_features[1:, 2] = np.exp(\n -gaps.astype(np.float64) / (60.0 * base.DAY)\n ).astype(np.float16)\n\n activity = causal_activity_features(\n timestamps, items, organic, completion\n )\n states = np.zeros(len(items), dtype=np.uint32)\n active_likes: dict[int, int] = {}\n active_dislikes: dict[int, int] = {}\n explicit_position = 0\n current_state = 0\n for row, timestamp in enumerate(timestamps):\n changed = False\n while (\n explicit_position < len(explicit.timestamps)\n and int(explicit.timestamps[explicit_position]) < int(timestamp)\n ):\n item = int(explicit.items[explicit_position])\n event_timestamp = int(explicit.timestamps[explicit_position])\n destination = (\n active_likes\n if int(explicit.channel[explicit_position]) == 0\n else active_dislikes\n )\n if int(explicit.active[explicit_position]) == 1:\n destination[item] = event_timestamp\n else:\n destination.pop(item, None)\n explicit_position += 1\n changed = True\n if changed:\n current_state = append_state_snapshot(\n active_likes,\n active_dislikes,\n n_items,\n liked_states,\n disliked_states,\n state_counts,\n )\n states[row] = np.uint32(current_state)\n return timestamps, items.copy(), event_features, activity, states\n\n\ndef final_user_features(\n timestamps: np.ndarray,\n items: np.ndarray,\n event_features: np.ndarray,\n history_end: int,\n like_count: int,\n dislike_count: int,\n) -> np.ndarray:\n start = int(\n np.searchsorted(\n timestamps,\n history_end - base.PROFILE_LOOKBACK_DAYS * base.DAY,\n side=\"left\",\n )\n )\n result = np.zeros(6, dtype=np.float32)\n recent_items = items[start:]\n if len(recent_items):\n result[0] = np.log1p(len(recent_items)) / 8.0\n result[1] = np.log1p(len(np.unique(recent_items))) / 6.0\n result[2] = float(np.mean(event_features[start:, 0]))\n result[3] = float(np.mean(event_features[start:, 1]))\n result[4] = np.log1p(like_count) / 6.0\n result[5] = np.log1p(dislike_count) / 4.0\n return result\n\n\ndef prepare_data(\n public: Path, stage: str, config: Config\n) -> tuple[SequenceData, CandidateSet]:\n data = public / \"data\"\n catalog = pd.read_parquet(data / \"candidate_catalog.parquet\").sort_values(\n \"popularity_rank\"\n )\n item_ids = catalog.item_id.to_numpy(dtype=np.int64, copy=True)\n item_index = pd.Index(item_ids)\n n_items = len(item_ids)\n if n_items >= np.iinfo(np.uint16).max:\n raise RuntimeError(\"uint16 item storage is too small for this catalog\")\n\n if stage == \"validation\":\n prefixes = (\"train\",)\n history_end = base.TRAIN_END\n candidate_name = \"validation_eligible_candidates.npz\"\n state_name = \"validation_preference_state.parquet\"\n else:\n prefixes = (\"train\", \"val\")\n history_end = base.VALIDATION_END\n candidate_name = \"test_eligible_candidates.npz\"\n state_name = \"test_preference_state.parquet\"\n eligibility = CandidateSet.read(data / candidate_name)\n if not np.array_equal(eligibility.item_ids, item_ids):\n raise RuntimeError(\"catalog and candidate artifact item order differ\")\n\n explicit_by_user = base.load_explicit_events(data, prefixes, item_index)\n inference_likes, inference_dislikes, like_counts, dislike_counts = (\n base.load_inference_state(\n data / state_name,\n item_index,\n eligibility.uids,\n n_items,\n )\n )\n inference_rows = {int(uid): row for row, uid in enumerate(eligibility.uids)}\n inference_user_features = np.zeros((len(eligibility.uids), 6), dtype=np.float32)\n inference_sequence_rows = np.full(len(eligibility.uids), -1, dtype=np.int32)\n\n # Snapshot zero is the shared empty explicit state.\n liked_states = [np.full(base.LIKE_ITEMS, n_items, dtype=np.uint16)]\n disliked_states = [\n np.full(base.DISLIKE_ITEMS, n_items, dtype=np.uint16)\n ]\n state_counts: list[tuple[int, int]] = [(0, 0)]\n user_ids: list[int] = []\n offsets = [0]\n timestamp_parts: list[np.ndarray] = []\n item_parts: list[np.ndarray] = []\n event_parts: list[np.ndarray] = []\n activity_parts: list[np.ndarray] = []\n state_parts: list[np.ndarray] = []\n\n listen_paths = [data / f\"{prefix}_listens.parquet\" for prefix in prefixes]\n started = time.monotonic()\n for seen_users, (uid, listens) in enumerate(\n base.merge_listen_streams(listen_paths, item_index), start=1\n ):\n parts = build_sequence_parts(\n listens,\n explicit_by_user.get(uid, base.empty_explicit_events()),\n n_items,\n liked_states,\n disliked_states,\n state_counts,\n )\n if parts is None:\n continue\n sequence_row = len(user_ids)\n user_ids.append(uid)\n timestamp_parts.append(parts[0].astype(np.uint32, copy=False))\n item_parts.append(parts[1])\n event_parts.append(parts[2])\n activity_parts.append(parts[3])\n state_parts.append(parts[4])\n offsets.append(offsets[-1] + len(parts[1]))\n inference_row = inference_rows.get(uid)\n if inference_row is not None:\n inference_sequence_rows[inference_row] = sequence_row\n inference_user_features[inference_row] = final_user_features(\n parts[0],\n parts[1],\n parts[2],\n history_end,\n int(like_counts[inference_row]),\n int(dislike_counts[inference_row]),\n )\n if seen_users % 1_000 == 0:\n log(\n f\"prepared users={seen_users:,} tokens={offsets[-1]:,} \"\n f\"states={len(liked_states):,} \"\n f\"elapsed={time.monotonic() - started:.1f}s\"\n )\n\n missing = np.flatnonzero(inference_sequence_rows < 0)\n if len(missing):\n inference_user_features[missing, 4] = np.log1p(like_counts[missing]) / 6.0\n inference_user_features[missing, 5] = (\n np.log1p(dislike_counts[missing]) / 4.0\n )\n log(f\"target users without qualifying catalog sequence={len(missing):,}\")\n\n prepared = SequenceData(\n item_ids=item_ids,\n item_features=base.build_item_features(catalog),\n user_ids=np.asarray(user_ids, dtype=np.int64),\n offsets=np.asarray(offsets, dtype=np.int64),\n sequence_timestamps=np.concatenate(timestamp_parts),\n sequence_items=np.concatenate(item_parts),\n event_features=np.concatenate(event_parts),\n activity_features=np.concatenate(activity_parts),\n state_ids=np.concatenate(state_parts),\n liked_states=np.stack(liked_states),\n disliked_states=np.stack(disliked_states),\n state_counts=np.asarray(state_counts, dtype=np.int32),\n inference_user_rows=inference_sequence_rows,\n inference_liked_items=inference_likes,\n inference_disliked_items=inference_dislikes,\n inference_user_features=inference_user_features,\n )\n log(\n f\"prepared stage={stage} sequence_users={len(prepared.user_ids):,} \"\n f\"tokens={len(prepared.sequence_items):,} \"\n f\"next_item_targets={prepared.target_count:,} \"\n f\"states={len(prepared.liked_states):,}\"\n )\n return prepared, eligibility\n\n\ndef cache_metadata(stage: str, config: Config) -> dict[str, object]:\n return {\n \"cache_version\": CACHE_VERSION,\n \"stage\": stage,\n \"positive_listen_threshold\": base.LISTEN_POSITIVE_THRESHOLD,\n \"profile_lookback_days\": base.PROFILE_LOOKBACK_DAYS,\n \"event_time_feature\": \"exp-negative-gap-over-60-days\",\n }\n\n\ndef save_cache(\n path: Path, prepared: SequenceData, stage: str, config: Config\n) -> None:\n path.parent.mkdir(parents=True, exist_ok=True)\n np.savez(\n path,\n metadata=np.asarray(json.dumps(cache_metadata(stage, config), sort_keys=True)),\n **prepared.__dict__,\n )\n log(f\"saved public-derived full-sequence cache: {path}\")\n\n\ndef load_cache(path: Path, stage: str, config: Config) -> SequenceData:\n with np.load(path, allow_pickle=False) as artifact:\n actual = json.loads(str(artifact[\"metadata\"].item()))\n expected = cache_metadata(stage, config)\n if actual != expected:\n raise RuntimeError(f\"sequence cache differs: {actual} != {expected}\")\n values = {\n field: artifact[field].copy()\n for field in SequenceData.__dataclass_fields__\n }\n prepared = SequenceData(**values)\n log(\n f\"loaded full-sequence cache: tokens={len(prepared.sequence_items):,} \"\n f\"targets={prepared.target_count:,}\"\n )\n return prepared\n\n\ndef load_or_prepare(\n public: Path,\n stage: str,\n config: Config,\n cache: Path | None,\n) -> tuple[SequenceData, CandidateSet]:\n candidate_name = (\n \"validation_eligible_candidates.npz\"\n if stage == \"validation\"\n else \"test_eligible_candidates.npz\"\n )\n eligibility = CandidateSet.read(public / \"data\" / candidate_name)\n if cache is not None and cache.is_file():\n prepared = load_cache(cache, stage, config)\n if not np.array_equal(prepared.item_ids, eligibility.item_ids):\n raise RuntimeError(\"cached catalog differs from candidate artifact\")\n return prepared, eligibility\n prepared, built_eligibility = prepare_data(public, stage, config)\n if cache is not None:\n save_cache(cache, prepared, stage, config)\n return prepared, built_eligibility\n\n\ndef build_item_correction(\n prepared: SequenceData,\n config: Config,\n target_starts: np.ndarray,\n target_counts: np.ndarray,\n) -> torch.Tensor:\n \"\"\"Estimate in-batch propensities from the selected training targets.\"\"\"\n counts = np.zeros(len(prepared.item_ids), dtype=np.float64)\n for user_row, count in enumerate(target_counts):\n if not count:\n continue\n offset = int(prepared.offsets[user_row])\n target_start = int(target_starts[user_row])\n counts += np.bincount(\n prepared.sequence_items[\n offset\n + target_start\n + 1 : offset\n + target_start\n + int(count)\n + 1\n ].astype(np.int64),\n minlength=len(prepared.item_ids),\n )\n total = float(counts.sum())\n probabilities = (counts + 1.0) / (total + len(counts))\n raw = np.power(\n (1.0 / len(counts)) / probabilities,\n config.inbatch_correction_power,\n )\n normalization = float(np.sum(counts * raw) / max(total, 1.0))\n correction = np.clip(\n raw / max(normalization, 1.0e-12),\n config.inbatch_correction_min,\n config.inbatch_correction_max,\n ).astype(np.float32)\n mean_on_targets = float(np.sum(counts * correction) / max(total, 1.0))\n log(\n f\"inbatch correction: min={correction.min():.3f} \"\n f\"mean_on_targets={mean_on_targets:.3f} max={correction.max():.3f}\"\n )\n return torch.from_numpy(correction)\n\n\ndef build_known_positive_filter(\n prepared: SequenceData, training_counts: np.ndarray\n) -> np.ndarray:\n \"\"\"Pack each user's complete public-training positive-item set.\"\"\"\n width = (len(prepared.item_ids) + 7) // 8\n packed = np.zeros((len(prepared.user_ids), width), dtype=np.uint8)\n retained = 0\n for user_row, target_count in enumerate(training_counts):\n if not target_count:\n continue\n offset = int(prepared.offsets[user_row])\n # T adjacent targets consume sequence positions 0 through T. The\n # chronological BCE suffix starts at T + 1 and is intentionally absent.\n items = np.unique(\n prepared.sequence_items[\n offset : offset + int(target_count) + 1\n ].astype(np.int64)\n )\n byte_columns = items >> 3\n bit_values = np.left_shift(\n np.uint8(1), (items & 7).astype(np.uint8)\n )\n np.bitwise_or.at(packed[user_row], byte_columns, bit_values)\n retained += len(items)\n log(\n f\"full-history false-negative filter: users={len(prepared.user_ids):,} \"\n f\"unique_user_items={retained:,} bytes={packed.nbytes:,}\"\n )\n return packed\n\n\ndef invalid_full_history_negatives(\n negatives: np.ndarray,\n targets: np.ndarray,\n user_rows: np.ndarray,\n known_positive_packed: np.ndarray,\n liked_items: np.ndarray,\n) -> np.ndarray:\n \"\"\"Return row-level invalid negatives using the complete training prefix.\"\"\"\n byte_columns = negatives >> 3\n bit_columns = negatives & 7\n known = known_positive_packed[user_rows[:, None], byte_columns]\n invalid = ((known >> bit_columns) & 1).astype(bool)\n invalid |= negatives == targets[:, None]\n invalid |= (negatives[:, :, None] == liked_items[:, None, :]).any(axis=2)\n for column in range(1, negatives.shape[1]):\n invalid[:, column] |= (\n negatives[:, column, None] == negatives[:, :column]\n ).any(axis=1)\n return invalid\n\n\ndef draw_full_history_filtered_negatives(\n targets: torch.Tensor,\n user_rows: np.ndarray,\n liked_items: torch.Tensor,\n known_positive_packed: np.ndarray,\n n_items: int,\n config: Config,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n \"\"\"Draw mixed negatives and reject all known public-training positives.\"\"\"\n targets_cpu = targets.detach().cpu().numpy().astype(np.int64, copy=False)\n liked_cpu = liked_items.detach().cpu().numpy().astype(np.int64, copy=False)\n batch_size = len(targets_cpu)\n uniform = torch.randint(\n n_items,\n (batch_size, config.uniform_negatives),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.empty((batch_size, 0), dtype=torch.long)\n corrected_count = 0\n if config.inbatch_negatives:\n unique_targets = torch.unique(targets.detach().cpu())\n corrected_count = min(\n config.inbatch_negatives, max(len(unique_targets) - 1, 0)\n )\n if corrected_count:\n priorities = torch.rand(\n (batch_size, len(unique_targets)), generator=generator\n )\n priorities.masked_fill_(\n unique_targets[None, :].eq(\n torch.from_numpy(targets_cpu)[:, None]\n ),\n -1.0,\n )\n columns = torch.topk(\n priorities, k=corrected_count, dim=1\n ).indices\n inbatch = unique_targets[columns]\n if corrected_count < config.inbatch_negatives:\n filler = torch.randint(\n n_items,\n (batch_size, config.inbatch_negatives - corrected_count),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.cat([inbatch, filler], dim=1)\n\n negatives = torch.cat([uniform, inbatch], dim=1).numpy()\n correction = np.ones_like(negatives, dtype=np.float32)\n if corrected_count:\n correction[\n :, config.uniform_negatives : config.uniform_negatives + corrected_count\n ] = item_correction[inbatch[:, :corrected_count]].numpy()\n for _ in range(32):\n invalid = invalid_full_history_negatives(\n negatives,\n targets_cpu,\n user_rows,\n known_positive_packed,\n liked_cpu,\n )\n if not np.any(invalid):\n return (\n torch.from_numpy(negatives).to(device),\n torch.from_numpy(correction).to(device),\n )\n replacements = torch.randint(\n n_items,\n (int(invalid.sum()),),\n generator=generator,\n device=\"cpu\",\n ).numpy()\n negatives[invalid] = replacements\n # Rejected in-batch items become ordinary uniform draws.\n correction[invalid] = 1.0\n invalid = invalid_full_history_negatives(\n negatives,\n targets_cpu,\n user_rows,\n known_positive_packed,\n liked_cpu,\n )\n if np.any(invalid):\n raise RuntimeError(\"full-history negative resampling failed\")\n return (\n torch.from_numpy(negatives).to(device),\n torch.from_numpy(correction).to(device),\n )\n\n\ndef draw_training_negatives(\n positive_items: torch.Tensor,\n known_history: torch.Tensor,\n liked_items: torch.Tensor,\n user_rows: np.ndarray,\n known_positive_packed: np.ndarray | None,\n prepared: SequenceData,\n config: Config,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n if config.negative_filter_scope == \"full-training-history\":\n if known_positive_packed is None:\n raise RuntimeError(\"full-history negative filter was not built\")\n return draw_full_history_filtered_negatives(\n positive_items,\n user_rows,\n liked_items,\n known_positive_packed,\n len(prepared.item_ids),\n config,\n item_correction,\n generator,\n device,\n )\n return base.draw_filtered_mixed_negatives(\n positive_items,\n known_history,\n liked_items,\n len(prepared.item_ids),\n config.uniform_negatives,\n config.inbatch_negatives,\n item_correction,\n generator,\n device,\n )\n\n\ndef event_sequence(\n model: base.SequentialTwoTower,\n items: torch.Tensor,\n features: torch.Tensor,\n hidden: torch.Tensor,\n) -> tuple[torch.Tensor, torch.Tensor]:\n inputs = torch.cat([model.context_embedding(items), features], dim=-1)\n projected = model.event_projection(inputs)\n return model.sequence_encoder(projected, hidden)\n\n\ndef pool_explicit_states(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n state_ids: np.ndarray,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor, torch.Tensor]:\n unique_states, inverse = np.unique(state_ids, return_inverse=True)\n liked_items = torch.from_numpy(\n prepared.liked_states[unique_states].astype(np.int64)\n ).to(device)\n disliked_items = torch.from_numpy(\n prepared.disliked_states[unique_states].astype(np.int64)\n ).to(device)\n liked_vectors = model._mean_context_channel(liked_items)\n disliked_vectors = model._mean_context_channel(disliked_items)\n inverse_rows = torch.from_numpy(inverse.astype(np.int64)).to(device)\n counts = prepared.state_counts[state_ids]\n return (\n liked_vectors[inverse_rows],\n disliked_vectors[inverse_rows],\n torch.from_numpy(counts.astype(np.float32)).to(device),\n )\n\n\ndef user_queries(\n model: base.SequentialTwoTower,\n sequence_vectors: torch.Tensor,\n liked_vectors: torch.Tensor,\n disliked_vectors: torch.Tensor,\n user_features: torch.Tensor,\n) -> torch.Tensor:\n return model.user_tower(\n torch.cat(\n [sequence_vectors, liked_vectors, disliked_vectors, user_features],\n dim=1,\n )\n )\n\n\ndef candidate_vectors(\n model: base.SequentialTwoTower,\n positives: torch.Tensor,\n negatives: torch.Tensor,\n) -> tuple[torch.Tensor, torch.Tensor]:\n all_items = torch.cat([positives, negatives.reshape(-1)])\n unique_items, inverse = torch.unique(all_items, return_inverse=True)\n encoded = model.encode_items(unique_items)\n positive_vectors = encoded[inverse[: len(positives)]]\n negative_vectors = encoded[inverse[len(positives) :]].reshape(\n len(positives), negatives.shape[1], -1\n )\n return positive_vectors, negative_vectors\n\n\ndef history_windows(\n sequence: np.ndarray,\n start: int,\n count: int,\n width: int,\n padding_item: int,\n) -> np.ndarray:\n base_start = max(0, start - width + 1)\n segment = sequence[base_start : start + count]\n padded = np.concatenate(\n [np.full(width - 1, padding_item, dtype=np.uint16), segment]\n )\n windows = np.lib.stride_tricks.sliding_window_view(padded, width)\n offset = start - base_start\n return windows[offset : offset + count].copy()\n\n\ndef corrected_negative_loss(\n logits: torch.Tensor,\n correction: torch.Tensor,\n config: Config,\n) -> torch.Tensor:\n raw = F.softplus(logits)\n total = config.uniform_negatives + config.inbatch_negatives\n result = torch.zeros(len(logits), device=logits.device)\n if config.uniform_negatives:\n result += config.uniform_negatives * raw[\n :, : config.uniform_negatives\n ].mean(dim=1)\n if config.inbatch_negatives:\n selected = slice(config.uniform_negatives, total)\n weights = correction[:, selected]\n result += config.inbatch_negatives * (\n (raw[:, selected] * weights).sum(dim=1)\n / weights.sum(dim=1).clamp_min(1.0e-6)\n )\n return result / total\n\n\ndef retained_target_window_starts(\n total_targets: np.ndarray,\n validation_counts: np.ndarray,\n training_counts: np.ndarray,\n) -> np.ndarray:\n \"\"\"Return the discarded prefix before the retained chronological suffix.\"\"\"\n training_region_ends = total_targets - validation_counts\n window_starts = training_region_ends - training_counts\n if (\n np.any(window_starts < 0)\n or np.any(training_counts < 0)\n or np.any(validation_counts < 0)\n ):\n raise RuntimeError(\"chronological target partition produced negative counts\")\n if not np.array_equal(\n window_starts + training_counts + validation_counts, total_targets\n ):\n raise RuntimeError(\"chronological target window is not a final contiguous suffix\")\n return window_starts\n\n\ndef chronological_target_partition(\n prepared: SequenceData,\n validation_fraction: float,\n max_train_targets_per_user: int,\n max_validation_targets_per_user: int,\n) -> TargetPartition:\n \"\"\"Select a recent contiguous window with a final out-of-time BCE suffix.\"\"\"\n totals = np.maximum(prepared.lengths - 1, 0).astype(np.int64)\n validation = np.floor(totals * validation_fraction).astype(np.int64)\n validation[(totals >= 2) & (validation == 0)] = 1\n if max_validation_targets_per_user:\n validation = np.minimum(validation, max_validation_targets_per_user)\n\n history_training = totals - validation\n training = history_training.copy()\n if max_train_targets_per_user:\n training = np.minimum(training, max_train_targets_per_user)\n window_starts = retained_target_window_starts(totals, validation, training)\n return TargetPartition(\n window_starts=window_starts,\n training_counts=training,\n validation_counts=validation,\n history_training_counts=history_training,\n )\n\n\ndef load_organic_like_outcomes(\n public: Path, stage: str, item_ids: np.ndarray\n) -> dict[int, dict[int, np.ndarray]]:\n \"\"\"Load mapped organic-like timestamps from public fitting history.\"\"\"\n prefixes = (\"train\",) if stage == \"validation\" else (\"train\", \"val\")\n frames = [\n pd.read_parquet(\n public / \"data\" / f\"{prefix}_likes.parquet\",\n columns=[\"uid\", \"timestamp\", \"item_id\", \"is_organic\"],\n )\n for prefix in prefixes\n ]\n likes = pd.concat(frames, ignore_index=True)\n likes = likes[likes.is_organic == 1].copy()\n mapped = pd.Index(item_ids).get_indexer(\n likes.item_id.to_numpy(dtype=np.int64)\n )\n likes = likes.loc[mapped >= 0, [\"uid\", \"timestamp\"]].copy()\n likes[\"item\"] = mapped[mapped >= 0].astype(np.int32, copy=False)\n likes = likes.sort_values([\"uid\", \"item\", \"timestamp\"], kind=\"stable\")\n nested: dict[int, dict[int, list[int]]] = {}\n for raw_uid, raw_timestamp, raw_item in likes.itertuples(\n index=False, name=None\n ):\n nested.setdefault(int(raw_uid), {}).setdefault(int(raw_item), []).append(\n int(raw_timestamp)\n )\n result = {\n uid: {\n item: np.asarray(timestamps, dtype=np.int64)\n for item, timestamps in items.items()\n }\n for uid, items in nested.items()\n }\n log(\n f\"loaded organic-like outcomes: rows={len(likes):,} \"\n f\"users={len(result):,}\"\n )\n return result\n\n\ndef build_target_weights(\n public: Path,\n stage: str,\n prepared: SequenceData,\n training_counts: np.ndarray,\n config: Config,\n) -> np.ndarray:\n \"\"\"Build nonzero per-token weights without dropping adjacent targets.\"\"\"\n if config.target_weighting == \"uniform\":\n weights = np.ones(len(prepared.sequence_items), dtype=np.float32)\n else:\n weights = np.full(\n len(prepared.sequence_items),\n config.ordinary_target_weight,\n dtype=np.float32,\n )\n organic_strong = (\n (prepared.event_features[:, 0] > 0.5)\n & (prepared.event_features[:, 1] >= 0.8)\n )\n weights[organic_strong] = config.organic_strong_target_weight\n if config.target_weighting == \"discovery-graded\":\n like_outcomes = load_organic_like_outcomes(\n public, stage, prepared.item_ids\n )\n history_end = (\n base.TRAIN_END if stage == \"validation\" else base.VALIDATION_END\n )\n horizon = config.discovery_horizon_days * base.DAY\n grade_three = 0\n for user_row, uid in enumerate(prepared.user_ids):\n start = int(prepared.offsets[user_row])\n end = int(prepared.offsets[user_row + 1])\n if end - start < 2:\n continue\n last_training_target = start + int(training_counts[user_row])\n first_validation_target = last_training_target + 1\n split_timestamp = (\n int(prepared.sequence_timestamps[first_validation_target])\n if first_validation_target < end\n else history_end\n )\n user_likes = like_outcomes.get(int(uid), {})\n seen = {int(prepared.sequence_items[start])}\n for position in range(start + 1, end):\n item = int(prepared.sequence_items[position])\n novel = item not in seen\n seen.add(item)\n if not novel:\n # Repeats remain training examples, but they cannot be\n # eligible discovery outcomes at the corresponding\n # cutoff and therefore keep only the ordinary weight.\n weights[position] = config.ordinary_target_weight\n continue\n timestamps = user_likes.get(item)\n if timestamps is None:\n continue\n timestamp = int(prepared.sequence_timestamps[position])\n upper = min(\n timestamp + horizon,\n split_timestamp\n if position <= last_training_target\n else history_end,\n )\n outcome = int(np.searchsorted(timestamps, timestamp, side=\"left\"))\n if outcome < len(timestamps) and int(timestamps[outcome]) < upper:\n weights[position] = config.organic_like_target_weight\n grade_three += 1\n log(\n f\"grade-3 first-encounter targets={grade_three:,} \"\n f\"horizon_days={config.discovery_horizon_days}\"\n )\n log(\n f\"target weighting={config.target_weighting} \"\n f\"min={weights.min():.3f} mean={weights.mean():.3f} \"\n f\"max={weights.max():.3f}\"\n )\n return weights\n\n\ndef build_dense_block(\n prepared: SequenceData,\n target_weights: np.ndarray,\n users: np.ndarray,\n window_starts: np.ndarray,\n process_counts: np.ndarray,\n score_starts: np.ndarray,\n start: int,\n steps: int,\n config: Config,\n) -> tuple[np.ndarray, ...]:\n \"\"\"Build one causal sequence block and mark the target rows to score.\"\"\"\n batch_size = len(users)\n padding_item = len(prepared.item_ids)\n input_items = np.full((batch_size, steps), padding_item, dtype=np.uint16)\n event_features = np.zeros((batch_size, steps, 3), dtype=np.float16)\n activity_features = np.zeros((batch_size, steps, 4), dtype=np.float16)\n state_ids = np.zeros((batch_size, steps), dtype=np.uint32)\n targets = np.zeros((batch_size, steps), dtype=np.uint16)\n weights = np.ones((batch_size, steps), dtype=np.float32)\n scored = np.zeros((batch_size, steps), dtype=bool)\n histories = np.full(\n (batch_size, steps, config.bptt_steps),\n padding_item,\n dtype=np.uint16,\n )\n processed = np.zeros(batch_size, dtype=np.int32)\n\n lengths = prepared.lengths\n for row, user_row in enumerate(users):\n count = min(steps, max(int(process_counts[row]) - start, 0))\n if not count:\n continue\n offset = int(prepared.offsets[user_row])\n window_start = int(window_starts[row])\n absolute_start = window_start + start\n positions = slice(\n offset + absolute_start,\n offset + absolute_start + count,\n )\n target_positions = slice(\n offset + absolute_start + 1,\n offset + absolute_start + count + 1,\n )\n input_items[row, :count] = prepared.sequence_items[positions]\n event_features[row, :count] = prepared.event_features[positions]\n activity_features[row, :count] = prepared.activity_features[positions]\n state_ids[row, :count] = prepared.state_ids[positions]\n targets[row, :count] = prepared.sequence_items[target_positions]\n weights[row, :count] = target_weights[target_positions]\n score_from = min(count, max(int(score_starts[row]) - start, 0))\n scored[row, score_from:count] = True\n processed[row] = count\n user_sequence = prepared.sequence_items[\n offset : offset + int(lengths[user_row])\n ]\n histories[row, :count] = history_windows(\n user_sequence,\n absolute_start,\n count,\n config.bptt_steps,\n padding_item,\n )\n return (\n input_items,\n event_features,\n activity_features,\n state_ids,\n targets,\n scored,\n histories,\n processed,\n weights,\n )\n\n\ndef score_target_microbatches(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n sequence_vectors: torch.Tensor,\n flat_states: np.ndarray,\n flat_activity: np.ndarray,\n positive_items: torch.Tensor,\n negatives: torch.Tensor,\n negative_weights: torch.Tensor,\n sample_weights: np.ndarray,\n device: torch.device,\n backward: bool,\n) -> tuple[dict[str, float], torch.Tensor | None]:\n \"\"\"Compute one logical block's exact BCE in bounded-memory slices.\"\"\"\n target_count = len(positive_items)\n total_weight = float(sample_weights.sum())\n if total_weight <= 0:\n raise RuntimeError(\"target block has no positive sample weight\")\n sequence_gradient = torch.zeros_like(sequence_vectors) if backward else None\n statistics = {\n \"loss_sum\": 0.0,\n \"positive_sum\": 0.0,\n \"negative_sum\": 0.0,\n \"pair_correct\": 0.0,\n \"weight_sum\": 0.0,\n }\n for begin in range(0, target_count, config.loss_microbatch_size):\n end = min(begin + config.loss_microbatch_size, target_count)\n selected_states = flat_states[begin:end]\n if backward:\n sequence_input = (\n sequence_vectors[begin:end].detach().requires_grad_(True)\n )\n else:\n sequence_input = sequence_vectors[begin:end]\n liked_vectors, disliked_vectors, _ = pool_explicit_states(\n model, prepared, selected_states, device\n )\n counts = prepared.state_counts[selected_states]\n numeric = np.zeros((end - begin, 6), dtype=np.float32)\n numeric[:, :4] = flat_activity[begin:end]\n numeric[:, 4] = np.log1p(counts[:, 0]) / 6.0\n numeric[:, 5] = np.log1p(counts[:, 1]) / 4.0\n query_vectors = user_queries(\n model,\n sequence_input,\n liked_vectors,\n disliked_vectors,\n torch.from_numpy(numeric).to(device),\n )\n positive_vectors, negative_vectors = candidate_vectors(\n model,\n positive_items[begin:end],\n negatives[begin:end],\n )\n positive_logits = (query_vectors * positive_vectors).sum(dim=1)\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n if config.objective == \"infonce\":\n cat_logits = torch.cat(\n [positive_logits.unsqueeze(1), negative_logits], dim=1\n ) / config.temperature\n labels = torch.zeros(\n len(positive_logits), dtype=torch.long, device=device\n )\n raw_loss = F.cross_entropy(cat_logits, labels, reduction=\"none\")\n loss_sum = (selected_weights * raw_loss).sum()\n else:\n positive_loss = F.softplus(-positive_logits)\n negative_loss = corrected_negative_loss(\n negative_logits, negative_weights[begin:end], config\n )\n loss_sum = (\n selected_weights * (positive_loss + negative_loss) / 2.0\n ).sum()\n if not bool(torch.isfinite(loss_sum)):\n raise RuntimeError(\"dense next-item training produced non-finite loss\")\n if backward:\n (loss_sum / total_weight).backward()\n if sequence_input.grad is None:\n raise RuntimeError(\"sequence microbatch did not produce a gradient\")\n sequence_gradient[begin:end] = sequence_input.grad\n statistics[\"loss_sum\"] += float(loss_sum.detach().cpu())\n statistics[\"positive_sum\"] += float(positive_logits.detach().sum().cpu())\n statistics[\"negative_sum\"] += float(negative_logits.detach().sum().cpu())\n statistics[\"pair_correct\"] += float(\n (positive_logits.detach()[:, None] > negative_logits.detach()).sum().cpu()\n )\n statistics[\"weight_sum\"] += float(selected_weights.sum().detach().cpu())\n return statistics, sequence_gradient\n\n\ndef release_accelerator_cache(device: torch.device) -> None:\n if device.type == \"mps\":\n torch.mps.empty_cache()\n\n\ndef run_training_epoch(\n model: base.SequentialTwoTower,\n optimizer: torch.optim.Optimizer,\n prepared: SequenceData,\n config: Config,\n target_weights: np.ndarray,\n window_starts: np.ndarray,\n target_counts: np.ndarray,\n item_correction: torch.Tensor,\n known_positive_packed: np.ndarray | None,\n generator: torch.Generator,\n device: torch.device,\n epoch: int,\n) -> dict[str, float]:\n model.train()\n train_users = np.flatnonzero(target_counts > 0)\n sorted_users = train_users[\n np.argsort(target_counts[train_users], kind=\"stable\")\n ]\n groups = [\n sorted_users[start : start + config.user_batch_size]\n for start in range(0, len(sorted_users), config.user_batch_size)\n ]\n group_order = torch.randperm(len(groups), generator=generator).numpy()\n negative_count = config.uniform_negatives + config.inbatch_negatives\n total_loss = 0.0\n total_targets = 0\n total_weight = 0.0\n total_positive = 0.0\n total_negative = 0.0\n total_pair_correct = 0.0\n batches = 0\n started = time.monotonic()\n\n for group_number in group_order:\n users = groups[int(group_number)]\n group_counts = target_counts[users]\n hidden = torch.zeros(\n 1, len(users), config.embedding_dim, device=device\n )\n max_targets = int(group_counts.max())\n for start in range(0, max_targets, config.bptt_steps):\n steps = min(config.bptt_steps, max_targets - start)\n block = build_dense_block(\n prepared,\n target_weights,\n users,\n window_starts[users],\n group_counts,\n np.zeros(len(users), dtype=np.int64),\n start,\n steps,\n config,\n )\n sequence_outputs, _ = event_sequence(\n model,\n torch.from_numpy(block[0].astype(np.int64)).to(device),\n torch.from_numpy(block[1].astype(np.float32)).to(device),\n hidden,\n )\n hidden = torch.stack(\n [\n sequence_outputs[row, count - 1]\n if count\n else hidden[0, row]\n for row, count in enumerate(block[7])\n ]\n ).unsqueeze(0).detach()\n flat_scored = block[5].reshape(-1)\n if not np.any(flat_scored):\n continue\n mask = torch.from_numpy(flat_scored).to(device)\n sequence_vectors = sequence_outputs.reshape(\n -1, config.embedding_dim\n )[mask]\n flat_states = block[3].reshape(-1)[flat_scored]\n positive_items = torch.from_numpy(\n block[4].reshape(-1)[flat_scored].astype(np.int64)\n ).to(device)\n known_history = torch.from_numpy(\n block[6].reshape(-1, config.bptt_steps)[flat_scored].astype(\n np.int64\n )\n ).to(device)\n liked_items = torch.from_numpy(\n prepared.liked_states[flat_states].astype(np.int64)\n ).to(device)\n flat_user_rows = np.broadcast_to(\n users[:, None], block[5].shape\n ).reshape(-1)[flat_scored]\n negatives, negative_weights = draw_training_negatives(\n positive_items,\n known_history,\n liked_items,\n flat_user_rows,\n known_positive_packed,\n prepared,\n config,\n item_correction,\n generator,\n device,\n )\n optimizer.zero_grad(set_to_none=True)\n statistics, sequence_gradient = score_target_microbatches(\n model,\n prepared,\n config,\n sequence_vectors,\n flat_states,\n block[2].reshape(-1, 4)[flat_scored].astype(np.float32),\n positive_items,\n negatives,\n negative_weights,\n block[8].reshape(-1)[flat_scored],\n device,\n backward=True,\n )\n if sequence_gradient is None:\n raise RuntimeError(\"training block omitted its sequence gradient\")\n sequence_vectors.backward(sequence_gradient)\n nn.utils.clip_grad_norm_(model.parameters(), config.gradient_clip)\n optimizer.step()\n\n count = len(positive_items)\n total_loss += statistics[\"loss_sum\"]\n total_targets += count\n total_weight += statistics[\"weight_sum\"]\n total_positive += statistics[\"positive_sum\"]\n total_negative += statistics[\"negative_sum\"]\n total_pair_correct += statistics[\"pair_correct\"]\n batches += 1\n if batches % 25 == 0:\n release_accelerator_cache(device)\n if batches % 250 == 0:\n batch_bce = statistics[\"loss_sum\"] / max(\n statistics[\"weight_sum\"], 1.0e-12\n )\n running_bce = total_loss / max(total_weight, 1.0e-12)\n log(\n f\"epoch={epoch}/{config.epochs} batches={batches:,} \"\n f\"targets={total_targets:,} \"\n f\"batch_bce={batch_bce:.6f} \"\n f\"running_bce={running_bce:.6f} \"\n f\"elapsed={time.monotonic() - started:.1f}s\"\n )\n\n return {\n \"epoch\": float(epoch),\n \"loss\": total_loss / max(total_weight, 1.0e-12),\n \"positive_logit\": total_positive / max(total_targets, 1),\n \"negative_logit\": total_negative / max(total_targets * negative_count, 1),\n \"pair_accuracy\": total_pair_correct\n / max(total_targets * negative_count, 1),\n \"targets\": float(total_targets),\n \"target_weight_sum\": total_weight,\n \"batches\": float(batches),\n \"seconds\": time.monotonic() - started,\n }\n\n\n@torch.no_grad()\ndef evaluate_validation_bce(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n target_weights: np.ndarray,\n window_starts: np.ndarray,\n training_counts: np.ndarray,\n validation_counts: np.ndarray,\n item_correction: torch.Tensor,\n known_positive_packed: np.ndarray | None,\n device: torch.device,\n) -> dict[str, float]:\n \"\"\"Measure deterministic BCE on each user's chronological target suffix.\"\"\"\n model.eval()\n totals = training_counts + validation_counts\n users_with_validation = np.flatnonzero(validation_counts > 0)\n ordered = users_with_validation[\n np.argsort(totals[users_with_validation], kind=\"stable\")\n ]\n groups = [\n ordered[start : start + config.user_batch_size]\n for start in range(0, len(ordered), config.user_batch_size)\n ]\n generator = torch.Generator(device=\"cpu\").manual_seed(config.seed + 20_000)\n total_loss = 0.0\n total_targets = 0\n total_weight = 0.0\n started = time.monotonic()\n\n for group_number, users in enumerate(groups, start=1):\n group_totals = totals[users]\n group_starts = training_counts[users]\n hidden = torch.zeros(\n 1, len(users), config.embedding_dim, device=device\n )\n max_targets = int(group_totals.max())\n for start in range(0, max_targets, config.bptt_steps):\n steps = min(config.bptt_steps, max_targets - start)\n block = build_dense_block(\n prepared,\n target_weights,\n users,\n window_starts[users],\n group_totals,\n group_starts,\n start,\n steps,\n config,\n )\n sequence_outputs, _ = event_sequence(\n model,\n torch.from_numpy(block[0].astype(np.int64)).to(device),\n torch.from_numpy(block[1].astype(np.float32)).to(device),\n hidden,\n )\n hidden = torch.stack(\n [\n sequence_outputs[row, count - 1]\n if count\n else hidden[0, row]\n for row, count in enumerate(block[7])\n ]\n ).unsqueeze(0)\n flat_scored = block[5].reshape(-1)\n if not np.any(flat_scored):\n continue\n mask = torch.from_numpy(flat_scored).to(device)\n sequence_vectors = sequence_outputs.reshape(\n -1, config.embedding_dim\n )[mask]\n flat_states = block[3].reshape(-1)[flat_scored]\n positive_items = torch.from_numpy(\n block[4].reshape(-1)[flat_scored].astype(np.int64)\n ).to(device)\n known_history = torch.from_numpy(\n block[6].reshape(-1, config.bptt_steps)[flat_scored].astype(\n np.int64\n )\n ).to(device)\n liked_items = torch.from_numpy(\n prepared.liked_states[flat_states].astype(np.int64)\n ).to(device)\n flat_user_rows = np.broadcast_to(\n users[:, None], block[5].shape\n ).reshape(-1)[flat_scored]\n negatives, negative_weights = draw_training_negatives(\n positive_items,\n known_history,\n liked_items,\n flat_user_rows,\n known_positive_packed,\n prepared,\n config,\n item_correction,\n generator,\n device,\n )\n statistics, _ = score_target_microbatches(\n model,\n prepared,\n config,\n sequence_vectors,\n flat_states,\n block[2].reshape(-1, 4)[flat_scored].astype(np.float32),\n positive_items,\n negatives,\n negative_weights,\n block[8].reshape(-1)[flat_scored],\n device,\n backward=False,\n )\n total_loss += statistics[\"loss_sum\"]\n total_targets += len(positive_items)\n total_weight += statistics[\"weight_sum\"]\n if group_number % 25 == 0:\n release_accelerator_cache(device)\n return {\n \"validation_loss\": total_loss / max(total_weight, 1.0e-12),\n \"validation_targets\": float(total_targets),\n \"validation_target_weight_sum\": total_weight,\n \"validation_seconds\": time.monotonic() - started,\n }\n\n\ndef snapshot_model(model: nn.Module) -> dict[str, torch.Tensor]:\n return {\n name: value.detach().cpu().clone()\n for name, value in model.state_dict().items()\n }\n\n\ndef save_epoch_checkpoint(\n checkpoint_dir: Path,\n epoch: int,\n model_state: dict[str, torch.Tensor],\n config: Config,\n diagnostics: dict[str, float],\n optimizer: torch.optim.Optimizer,\n generator: torch.Generator,\n) -> Path:\n checkpoint_dir.mkdir(parents=True, exist_ok=True)\n path = checkpoint_dir / f\"epoch-{epoch:02d}.pt\"\n torch.save(\n {\n \"epoch\": epoch,\n \"config\": asdict(config),\n \"diagnostics\": diagnostics,\n \"model_state\": model_state,\n \"optimizer_state\": optimizer.state_dict(),\n \"generator_state\": generator.get_state(),\n },\n path,\n )\n log(f\"saved epoch checkpoint: {path}\")\n return path\n\n\ndef train_model(\n public: Path,\n stage: str,\n prepared: SequenceData,\n config: Config,\n device: torch.device,\n checkpoint_dir: Path | None,\n resume_checkpoint: Path | None,\n) -> tuple[base.SequentialTwoTower, list[dict[str, float]]]:\n partition = chronological_target_partition(\n prepared,\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n training_counts = partition.training_counts\n validation_counts = partition.validation_counts\n log(\n f\"chronological BCE window: train={int(training_counts.sum()):,} \"\n f\"validation={int(validation_counts.sum()):,} \"\n f\"selected={partition.selected_count:,} \"\n f\"discarded_old_targets={int(partition.window_starts.sum()):,} \"\n f\"all={prepared.target_count:,}\"\n )\n model = base.SequentialTwoTower(prepared.item_features, model_config(config)).to(\n device\n )\n optimizer = torch.optim.AdamW(\n model.parameters(), lr=config.learning_rate, weight_decay=config.weight_decay\n )\n generator = torch.Generator(device=\"cpu\").manual_seed(config.seed + 10_000)\n item_correction = build_item_correction(\n prepared,\n config,\n partition.window_starts,\n training_counts,\n )\n target_weights = build_target_weights(\n public,\n stage,\n prepared,\n partition.history_training_counts,\n config,\n )\n known_positive_packed = (\n build_known_positive_filter(prepared, partition.history_training_counts)\n if config.negative_filter_scope == \"full-training-history\"\n else None\n )\n history: list[dict[str, float]] = []\n best_loss = float(\"inf\")\n best_epoch = 0\n best_state: dict[str, torch.Tensor] | None = None\n epochs_without_improvement = 0\n start_epoch = 0\n\n if resume_checkpoint is not None:\n payload = torch.load(\n resume_checkpoint, map_location=device, weights_only=False\n )\n stored_config = payload.get(\"config\", {})\n architecture_fields = (\n \"bptt_steps\",\n \"embedding_dim\",\n \"hidden_dim\",\n \"user_hidden_1\",\n \"user_hidden_2\",\n \"item_hidden_1\",\n \"item_hidden_2\",\n )\n mismatches = {\n name: (stored_config.get(name), getattr(config, name))\n for name in architecture_fields\n if stored_config.get(name) != getattr(config, name)\n }\n if mismatches:\n raise RuntimeError(\n f\"resume checkpoint architecture differs: {mismatches}\"\n )\n stored_filter_scope = stored_config.get(\n \"negative_filter_scope\", \"last-100\"\n )\n stored_target_weighting = stored_config.get(\n \"target_weighting\", \"uniform\"\n )\n stored_weight_values = (\n stored_config.get(\"ordinary_target_weight\", 1.0),\n stored_config.get(\"organic_strong_target_weight\", 1.0),\n stored_config.get(\"organic_like_target_weight\", 3.0),\n stored_config.get(\"discovery_horizon_days\", 7),\n )\n current_weight_values = (\n config.ordinary_target_weight,\n config.organic_strong_target_weight,\n config.organic_like_target_weight,\n config.discovery_horizon_days,\n )\n stored_window_values = (\n stored_config.get(\"validation_fraction\", 0.10),\n stored_config.get(\"max_train_targets_per_user\", 0),\n stored_config.get(\"max_validation_targets_per_user\", 0),\n )\n current_window_values = (\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n objective_changed = (\n stored_filter_scope != config.negative_filter_scope\n or stored_target_weighting != config.target_weighting\n or stored_weight_values != current_weight_values\n or stored_window_values != current_window_values\n )\n model.load_state_dict(payload[\"model_state\"])\n if \"optimizer_state\" in payload and not objective_changed:\n optimizer.load_state_dict(payload[\"optimizer_state\"])\n for group in optimizer.param_groups:\n group[\"lr\"] = config.learning_rate\n group[\"weight_decay\"] = config.weight_decay\n log(f\"restored optimizer state from {resume_checkpoint}\")\n else:\n log(\n f\"using a fresh AdamW optimizer after loading {resume_checkpoint}\"\n )\n if \"generator_state\" in payload and not objective_changed:\n generator.set_state(payload[\"generator_state\"])\n start_epoch = int(payload[\"epoch\"])\n if start_epoch >= config.epochs:\n raise RuntimeError(\n f\"resume epoch {start_epoch} must be below --epochs {config.epochs}\"\n )\n prior = {\n str(name): float(value)\n for name, value in payload.get(\"diagnostics\", {}).items()\n }\n if objective_changed:\n prior.pop(\"validation_loss\", None)\n log(\n \"BCE objective changed: filter \"\n f\"{stored_filter_scope}->{config.negative_filter_scope}, \"\n \"target weighting \"\n f\"{stored_target_weighting}->{config.target_weighting}, \"\n f\"target window {stored_window_values}->{current_window_values}; \"\n \"recomputing the resume checkpoint's validation BCE\"\n )\n if \"validation_loss\" not in prior:\n prior.update(\n evaluate_validation_bce(\n model,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n validation_counts,\n item_correction,\n known_positive_packed,\n device,\n )\n )\n history.append(prior)\n best_loss = float(prior[\"validation_loss\"])\n best_epoch = start_epoch\n best_state = snapshot_model(model)\n log(\n f\"resumed validation-BCE training: epoch={start_epoch} \"\n f\"loss={best_loss:.6f}\"\n )\n\n for epoch in range(start_epoch + 1, config.epochs + 1):\n diagnostics = run_training_epoch(\n model,\n optimizer,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n item_correction,\n known_positive_packed,\n generator,\n device,\n epoch,\n )\n validation = evaluate_validation_bce(\n model,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n validation_counts,\n item_correction,\n known_positive_packed,\n device,\n )\n diagnostics.update(validation)\n history.append(diagnostics)\n current_state = snapshot_model(model)\n if checkpoint_dir is not None:\n save_epoch_checkpoint(\n checkpoint_dir,\n epoch,\n current_state,\n config,\n diagnostics,\n optimizer,\n generator,\n )\n improved = (\n diagnostics[\"validation_loss\"]\n < best_loss - config.early_stopping_min_delta\n )\n if improved:\n best_loss = diagnostics[\"validation_loss\"]\n best_epoch = epoch\n best_state = current_state\n epochs_without_improvement = 0\n if checkpoint_dir is not None:\n torch.save(\n {\n \"epoch\": epoch,\n \"config\": asdict(config),\n \"diagnostics\": diagnostics,\n \"model_state\": current_state,\n \"optimizer_state\": optimizer.state_dict(),\n \"generator_state\": generator.get_state(),\n },\n checkpoint_dir / \"best.pt\",\n )\n else:\n epochs_without_improvement += 1\n log(\n f\"epoch={epoch}/{config.epochs} train_loss={diagnostics['loss']:.5f} \"\n f\"validation_loss={diagnostics['validation_loss']:.5f} \"\n f\"pair_accuracy={diagnostics['pair_accuracy']:.3f} \"\n f\"train_targets={int(diagnostics['targets']):,} \"\n f\"validation_targets={int(diagnostics['validation_targets']):,} \"\n f\"train_seconds={diagnostics['seconds']:.1f} \"\n f\"validation_seconds={diagnostics['validation_seconds']:.1f}\"\n )\n release_accelerator_cache(device)\n if epochs_without_improvement >= config.early_stopping_patience:\n log(\n f\"early stopping after epoch={epoch}; \"\n f\"best_epoch={best_epoch} best_validation_loss={best_loss:.6f}\"\n )\n break\n\n if best_state is None:\n raise RuntimeError(\"training did not produce a valid checkpoint\")\n model.load_state_dict(best_state)\n model.to(device).eval()\n log(\n f\"restored best validation-BCE checkpoint: epoch={best_epoch} \"\n f\"loss={best_loss:.6f}\"\n )\n return model, history\n\n\n@torch.inference_mode()\ndef inference_queries(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n device: torch.device,\n) -> torch.Tensor:\n sequence_rows = prepared.inference_user_rows\n lengths = np.where(\n sequence_rows >= 0, prepared.lengths[np.maximum(sequence_rows, 0)], 0\n )\n ordered = np.argsort(lengths, kind=\"stable\")\n result = torch.zeros(\n len(sequence_rows), config.embedding_dim, dtype=torch.float32\n )\n for group_start in range(0, len(ordered), config.user_batch_size):\n output_rows = ordered[group_start : group_start + config.user_batch_size]\n rows = sequence_rows[output_rows]\n group_lengths = lengths[output_rows]\n hidden = torch.zeros(\n 1, len(output_rows), config.embedding_dim, device=device\n )\n for start in range(0, int(group_lengths.max(initial=0)), config.bptt_steps):\n steps = min(\n config.bptt_steps,\n int(group_lengths.max(initial=0)) - start,\n )\n items = np.full(\n (len(output_rows), steps),\n len(prepared.item_ids),\n dtype=np.uint16,\n )\n features = np.zeros((len(output_rows), steps, 3), dtype=np.float16)\n counts = np.zeros(len(output_rows), dtype=np.int32)\n for local_row, sequence_row in enumerate(rows):\n if sequence_row < 0:\n continue\n count = min(steps, max(int(group_lengths[local_row]) - start, 0))\n if not count:\n continue\n offset = int(prepared.offsets[sequence_row])\n selected = slice(offset + start, offset + start + count)\n items[local_row, :count] = prepared.sequence_items[selected]\n features[local_row, :count] = prepared.event_features[selected]\n counts[local_row] = count\n outputs, _ = event_sequence(\n model,\n torch.from_numpy(items.astype(np.int64)).to(device),\n torch.from_numpy(features.astype(np.float32)).to(device),\n hidden,\n )\n next_rows: list[torch.Tensor] = []\n for local_row, count in enumerate(counts):\n next_rows.append(\n outputs[local_row, count - 1]\n if count\n else hidden[0, local_row]\n )\n hidden = torch.stack(next_rows).unsqueeze(0)\n\n liked_items = torch.from_numpy(\n prepared.inference_liked_items[output_rows].astype(np.int64)\n ).to(device)\n disliked_items = torch.from_numpy(\n prepared.inference_disliked_items[output_rows].astype(np.int64)\n ).to(device)\n queries = user_queries(\n model,\n hidden.squeeze(0),\n model._mean_context_channel(liked_items),\n model._mean_context_channel(disliked_items),\n torch.from_numpy(\n prepared.inference_user_features[output_rows].astype(np.float32)\n ).to(device),\n )\n result[torch.from_numpy(output_rows.astype(np.int64))] = queries.cpu()\n return result\n\n\n@torch.inference_mode()\ndef rank_candidates(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n eligibility: CandidateSet,\n config: Config,\n device: torch.device,\n ranking_batch_size: int,\n) -> tuple[pd.DataFrame, dict[str, float]]:\n model.eval()\n queries = inference_queries(model, prepared, config, device)\n item_parts: list[torch.Tensor] = []\n for start in range(0, len(prepared.item_ids), 2_048):\n rows = torch.arange(\n start,\n min(start + 2_048, len(prepared.item_ids)),\n dtype=torch.long,\n device=device,\n )\n item_parts.append(model.encode_items(rows).cpu())\n items = torch.cat(item_parts)\n rows: list[tuple[int, int, int]] = []\n for start in range(0, len(eligibility.uids), ranking_batch_size):\n end = min(start + ranking_batch_size, len(eligibility.uids))\n scores = queries[start:end] @ items.T\n eligible = np.unpackbits(\n eligibility.packed[start:end],\n axis=1,\n count=len(prepared.item_ids),\n bitorder=\"little\",\n ).astype(bool, copy=False)\n scores.masked_fill_(~torch.from_numpy(eligible), -torch.inf)\n selected = torch.topk(scores, k=base.TOP_N, dim=1).indices.numpy()\n for local_row, columns in enumerate(selected):\n uid = int(eligibility.uids[start + local_row])\n rows.extend(\n (uid, int(prepared.item_ids[column]), rank)\n for rank, column in enumerate(columns, start=1)\n )\n if end == len(eligibility.uids) or end % 1_024 == 0:\n log(f\"ranked users={end:,}/{len(eligibility.uids):,}\")\n query_norm = torch.linalg.vector_norm(queries, dim=1).numpy()\n item_norm = torch.linalg.vector_norm(items, dim=1).numpy()\n diagnostics = {\n \"query_norm_mean\": float(query_norm.mean()),\n \"query_norm_std\": float(query_norm.std()),\n \"item_norm_mean\": float(item_norm.mean()),\n \"item_norm_std\": float(item_norm.std()),\n }\n return pd.DataFrame(rows, columns=[\"uid\", \"item_id\", \"rank\"]), diagnostics\n\n\ndef write_ranking(frame: pd.DataFrame, output: Path) -> None:\n output.parent.mkdir(parents=True, exist_ok=True)\n if output.suffix == \".parquet\":\n frame.to_parquet(output, index=False, compression=\"zstd\")\n elif output.suffix == \".csv\":\n frame.to_csv(output, index=False)\n else:\n raise ValueError(\"--output must end in .parquet or .csv\")\n\n\ndef main() -> int:\n parser = argparse.ArgumentParser(description=__doc__)\n parser.add_argument(\n \"--stage\", choices=(\"validation\", \"submission\"), default=\"validation\"\n )\n parser.add_argument(\"--output\", type=Path, required=True)\n parser.add_argument(\"--public-dir\", type=Path, required=True)\n parser.add_argument(\"--prepared-cache\", type=Path)\n parser.add_argument(\"--device\", choices=(\"cpu\", \"mps\"), default=\"cpu\")\n parser.add_argument(\"--threads\", type=int, default=8)\n parser.add_argument(\"--bptt-steps\", type=int, default=Config.bptt_steps)\n parser.add_argument(\n \"--user-batch-size\", type=int, default=Config.user_batch_size\n )\n parser.add_argument(\"--epochs\", type=int, default=Config.epochs)\n parser.add_argument(\"--learning-rate\", type=float, default=Config.learning_rate)\n parser.add_argument(\n \"--loss-microbatch-size\",\n type=int,\n default=Config.loss_microbatch_size,\n )\n parser.add_argument(\n \"--negative-filter-scope\",\n choices=(\"full-training-history\", \"last-100\"),\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n \"--objective\",\n choices=(\"bce\", \"infonce\"),\n default=Config.objective,\n )\n parser.add_argument(\n \"--temperature\",\n type=float,\n default=Config.temperature,\n )\n parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n )\n parser.add_argument(\n \"--ordinary-target-weight\",\n type=float,\n default=Config.ordinary_target_weight,\n )\n parser.add_argument(\n \"--organic-strong-target-weight\",\n type=float,\n default=Config.organic_strong_target_weight,\n )\n parser.add_argument(\n \"--organic-like-target-weight\",\n type=float,\n default=Config.organic_like_target_weight,\n )\n parser.add_argument(\n \"--discovery-horizon-days\",\n type=int,\n default=Config.discovery_horizon_days,\n )\n parser.add_argument(\n \"--validation-fraction\",\n type=float,\n default=Config.validation_fraction,\n )\n parser.add_argument(\n \"--max-train-targets-per-user\",\n type=int,\n default=Config.max_train_targets_per_user,\n help=\"retain this many recent pre-validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--max-validation-targets-per-user\",\n type=int,\n default=Config.max_validation_targets_per_user,\n help=\"retain this many final validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--early-stopping-patience\",\n type=int,\n default=Config.early_stopping_patience,\n )\n parser.add_argument(\n \"--early-stopping-min-delta\",\n type=float,\n default=Config.early_stopping_min_delta,\n )\n parser.add_argument(\"--checkpoint-dir\", type=Path)\n parser.add_argument(\"--resume-checkpoint\", type=Path)\n parser.add_argument(\"--seed\", type=int, default=Config.seed)\n parser.add_argument(\"--ranking-batch-size\", type=int, default=64)\n parser.add_argument(\"--prepare-only\", action=\"store_true\")\n args = parser.parse_args()\n\n if args.threads < 1 or args.bptt_steps < 2 or args.user_batch_size < 1:\n parser.error(\"threads, BPTT steps, and user batch size must be positive\")\n if (\n args.epochs < 1\n or args.learning_rate <= 0\n or args.loss_microbatch_size < 1\n or args.early_stopping_patience < 1\n or args.early_stopping_min_delta < 0\n or args.ordinary_target_weight <= 0\n or args.organic_strong_target_weight <= 0\n or args.organic_like_target_weight <= 0\n or args.discovery_horizon_days < 1\n or args.max_train_targets_per_user < 0\n or args.max_validation_targets_per_user < 0\n ):\n parser.error(\"training counts, learning rate, and patience are invalid\")\n if not 0.0 < args.validation_fraction < 1.0:\n parser.error(\"validation fraction must be strictly between zero and one\")\n public = args.public_dir.resolve()\n if not (public / \"data\" / \"candidate_catalog.parquet\").is_file():\n parser.error(f\"public task data not found under {public}\")\n config = Config(\n bptt_steps=args.bptt_steps,\n user_batch_size=args.user_batch_size,\n epochs=args.epochs,\n learning_rate=args.learning_rate,\n loss_microbatch_size=args.loss_microbatch_size,\n negative_filter_scope=args.negative_filter_scope,\n target_weighting=args.target_weighting,\n ordinary_target_weight=args.ordinary_target_weight,\n organic_strong_target_weight=args.organic_strong_target_weight,\n organic_like_target_weight=args.organic_like_target_weight,\n discovery_horizon_days=args.discovery_horizon_days,\n validation_fraction=args.validation_fraction,\n max_train_targets_per_user=args.max_train_targets_per_user,\n max_validation_targets_per_user=args.max_validation_targets_per_user,\n early_stopping_patience=args.early_stopping_patience,\n early_stopping_min_delta=args.early_stopping_min_delta,\n seed=args.seed,\n )\n torch.set_num_threads(args.threads)\n torch.set_num_interop_threads(1)\n seed_everything(config.seed)\n device = torch.device(args.device)\n if device.type == \"mps\" and not torch.backends.mps.is_available():\n parser.error(\"MPS was requested but is unavailable\")\n\n started = time.monotonic()\n prepared, eligibility = load_or_prepare(\n public,\n args.stage,\n config,\n args.prepared_cache.resolve() if args.prepared_cache else None,\n )\n target_partition = chronological_target_partition(\n prepared,\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n if args.prepare_only:\n print(\n json.dumps(\n {\n \"stage\": args.stage,\n \"sequence_users\": len(prepared.user_ids),\n \"sequence_tokens\": len(prepared.sequence_items),\n \"next_item_targets\": prepared.target_count,\n \"selected_training_targets\": int(\n target_partition.training_counts.sum()\n ),\n \"selected_validation_targets\": int(\n target_partition.validation_counts.sum()\n ),\n \"discarded_old_targets\": int(\n target_partition.window_starts.sum()\n ),\n \"target_users\": len(eligibility.uids),\n \"runtime_seconds\": round(time.monotonic() - started, 3),\n \"config\": asdict(config),\n },\n indent=2,\n )\n )\n return 0\n\n model, training_history = train_model(\n public,\n args.stage,\n prepared,\n config,\n device,\n args.checkpoint_dir.resolve() if args.checkpoint_dir else None,\n args.resume_checkpoint.resolve() if args.resume_checkpoint else None,\n )\n ranking, embedding_diagnostics = rank_candidates(\n model,\n prepared,\n eligibility,\n config,\n device,\n args.ranking_batch_size,\n )\n write_ranking(ranking, args.output.resolve())\n print(\n json.dumps(\n {\n \"baseline\": \"yambda-windowed-sequence-two-tower-base\",\n \"stage\": args.stage,\n \"output\": str(args.output.resolve()),\n \"rows\": len(ranking),\n \"users\": int(ranking.uid.nunique()),\n \"runtime_seconds\": round(time.monotonic() - started, 3),\n \"public_inputs_only\": True,\n \"architecture\": {\n \"family\": \"pure-windowed-sequence-two-tower\",\n \"sequence_training\": (\n \"every-adjacent-target-within-recent-chronological-window\"\n ),\n \"sequence_encoder\": \"one-layer-GRU-dense-timestep-loss\",\n \"bptt_hidden_state\": \"carried-and-detached\",\n \"model_selection\": \"chronological-validation-BCE-early-stopping\",\n \"loss_execution\": \"gradient-accumulated-microbatches\",\n \"user_id_embedding\": False,\n \"user_mlp\": \"582-512-256-192\",\n \"item_mlp\": \"196-512-256-192\",\n \"normalization\": \"none\",\n \"score\": \"raw-dot-product\",\n \"objective\": \"sample-weighted-BCE-at-every-timestep\",\n \"negative_sampling\": (\n \"24-uniform-plus-8-propensity-corrected-inbatch\"\n ),\n \"false_negative_filter\": config.negative_filter_scope,\n \"target_weighting\": config.target_weighting,\n },\n \"config\": asdict(config),\n \"sequence_tokens\": len(prepared.sequence_items),\n \"all_next_item_targets\": prepared.target_count,\n \"selected_training_targets\": int(\n target_partition.training_counts.sum()\n ),\n \"selected_validation_targets\": int(\n target_partition.validation_counts.sum()\n ),\n \"discarded_old_targets\": int(target_partition.window_starts.sum()),\n \"training_history\": training_history,\n \"embedding_diagnostics\": embedding_diagnostics,\n \"torch_version\": torch.__version__,\n \"device\": str(device),\n },\n indent=2,\n )\n )\n return 0\n\n\nif __name__ == \"__main__\":\n raise SystemExit(main())\n",
"diffStat": {
"model_added_lines": 10,
"model_removed_lines": 0,
"model_added_chars": 226,
"model_removed_chars": 0,
"user_added_lines": 0,
"user_removed_lines": 0,
"user_added_chars": 0,
"user_removed_chars": 0
},
"isNewFile": false,
"isBuildFile": false
},
"description": "solution/train_current_model.py: parser.add_argument(... => parser.add_argument(...",
"displayName": "Edit",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T06:07:10.148Z",
"content": "Successfully modified file: /app/solution/train_current_model.py (1 replacements). Here is the updated code:\n...\n \"--negative-filter-scope\",\n choices=(\"full-training-history\", \"last-100\"),\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n \"--objective\",\n choices=(\"bce\", \"infonce\"),\n default=Config.objective,\n )\n parser.add_argument(\n \"--temperature\",\n type=float,\n default=Config.temperature,\n )\n parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n )\n parser.add_argument(\n...",
"isError": false
},
"responseMessageId": "58319f3a-e779-4cc0-8c4c-4a735ceb9c9d"
}
]
},
{
"id": 51,
"role": "agent",
"timestamp": "2026-09-29T06:07:13.096Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T06:07:13.096Z",
"callId": "read_file__call_4034917",
"name": "read_file",
"input": {
"start_line": 2045,
"end_line": 2075,
"file_path": "/app/solution/train_current_model.py"
},
"recorded": {
"id": "read_file__call_4034917",
"name": "read_file",
"args": {
"start_line": 2045,
"end_line": 2075,
"file_path": "/app/solution/train_current_model.py"
},
"result": [
{
"functionResponse": {
"id": "read_file__call_4034917",
"name": "read_file",
"response": {
"output": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 2045-2075 of 2187 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 2076.\n\n--- FILE CONTENT (truncated) ---\n if not 0.0 < args.validation_fraction < 1.0:\n parser.error(\"validation fraction must be strictly between zero and one\")\n public = args.public_dir.resolve()\n if not (public / \"data\" / \"candidate_catalog.parquet\").is_file():\n parser.error(f\"public task data not found under {public}\")\n config = Config(\n bptt_steps=args.bptt_steps,\n user_batch_size=args.user_batch_size,\n epochs=args.epochs,\n learning_rate=args.learning_rate,\n loss_microbatch_size=args.loss_microbatch_size,\n negative_filter_scope=args.negative_filter_scope,\n target_weighting=args.target_weighting,\n ordinary_target_weight=args.ordinary_target_weight,\n organic_strong_target_weight=args.organic_strong_target_weight,\n organic_like_target_weight=args.organic_like_target_weight,\n discovery_horizon_days=args.discovery_horizon_days,\n validation_fraction=args.validation_fraction,\n max_train_targets_per_user=args.max_train_targets_per_user,\n max_validation_targets_per_user=args.max_validation_targets_per_user,\n early_stopping_patience=args.early_stopping_patience,\n early_stopping_min_delta=args.early_stopping_min_delta,\n seed=args.seed,\n )\n torch.set_num_threads(args.threads)\n torch.set_num_interop_threads(1)\n seed_everything(config.seed)\n device = torch.device(args.device)\n if device.type == \"mps\" and not torch.backends.mps.is_available():\n parser.error(\"MPS was requested but is unavailable\")\n"
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T06:07:13.101Z",
"resultDisplay": "Read lines 2045-2075 of 2187 from solution/train_current_model.py",
"description": "solution/train_current_model.py",
"displayName": "ReadFile",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T06:07:13.103Z",
"content": "\nIMPORTANT: The file content has been truncated.\nStatus: Showing lines 2045-2075 of 2187 total lines.\nAction: To read more of the file, you can use the 'start_line' and 'end_line' parameters in a subsequent 'read_file' call. For example, to read the next section of the file, use start_line: 2076.\n\n--- FILE CONTENT (truncated) ---\n if not 0.0 < args.validation_fraction < 1.0:\n parser.error(\"validation fraction must be strictly between zero and one\")\n public = args.public_dir.resolve()\n if not (public / \"data\" / \"candidate_catalog.parquet\").is_file():\n parser.error(f\"public task data not found under {public}\")\n config = Config(\n bptt_steps=args.bptt_steps,\n user_batch_size=args.user_batch_size,\n epochs=args.epochs,\n learning_rate=args.learning_rate,\n loss_microbatch_size=args.loss_microbatch_size,\n negative_filter_scope=args.negative_filter_scope,\n target_weighting=args.target_weighting,\n ordinary_target_weight=args.ordinary_target_weight,\n organic_strong_target_weight=args.organic_strong_target_weight,\n organic_like_target_weight=args.organic_like_target_weight,\n discovery_horizon_days=args.discovery_horizon_days,\n validation_fraction=args.validation_fraction,\n max_train_targets_per_user=args.max_train_targets_per_user,\n max_validation_targets_per_user=args.max_validation_targets_per_user,\n early_stopping_patience=args.early_stopping_patience,\n early_stopping_min_delta=args.early_stopping_min_delta,\n seed=args.seed,\n )\n torch.set_num_threads(args.threads)\n torch.set_num_interop_threads(1)\n seed_everything(config.seed)\n device = torch.device(args.device)\n if device.type == \"mps\" and not torch.backends.mps.is_available():\n parser.error(\"MPS was requested but is unavailable\")\n",
"isError": false
},
"responseMessageId": "fd56e6a6-3e48-4c0b-af82-129042aa69f1"
}
]
},
{
"id": 52,
"role": "agent",
"timestamp": "2026-09-29T06:07:17.238Z",
"blocks": [
{
"kind": "tool",
"timestamp": "2026-09-29T06:07:17.238Z",
"callId": "replace__call_3934",
"name": "replace",
"input": {
"file_path": "/app/solution/train_current_model.py",
"instruction": "Pass objective and temperature into Config instantiation",
"allow_multiple": false,
"new_string": " negative_filter_scope=args.negative_filter_scope,\n objective=args.objective,\n temperature=args.temperature,\n target_weighting=args.target_weighting,",
"old_string": " negative_filter_scope=args.negative_filter_scope,\n target_weighting=args.target_weighting,"
},
"recorded": {
"id": "replace__call_3934",
"name": "replace",
"args": {
"file_path": "/app/solution/train_current_model.py",
"instruction": "Pass objective and temperature into Config instantiation",
"allow_multiple": false,
"new_string": " negative_filter_scope=args.negative_filter_scope,\n objective=args.objective,\n temperature=args.temperature,\n target_weighting=args.target_weighting,",
"old_string": " negative_filter_scope=args.negative_filter_scope,\n target_weighting=args.target_weighting,"
},
"result": [
{
"functionResponse": {
"id": "replace__call_3934",
"name": "replace",
"response": {
"output": "Successfully modified file: /app/solution/train_current_model.py (1 replacements). Here is the updated code:\n...\n user_batch_size=args.user_batch_size,\n epochs=args.epochs,\n learning_rate=args.learning_rate,\n loss_microbatch_size=args.loss_microbatch_size,\n negative_filter_scope=args.negative_filter_scope,\n objective=args.objective,\n temperature=args.temperature,\n target_weighting=args.target_weighting,\n ordinary_target_weight=args.ordinary_target_weight,\n organic_strong_target_weight=args.organic_strong_target_weight,\n organic_like_target_weight=args.organic_like_target_weight,\n discovery_horizon_days=args.discovery_horizon_days,\n..."
}
}
}
],
"status": "success",
"timestamp": "2026-09-29T06:07:17.251Z",
"resultDisplay": {
"fileDiff": "Index: train_current_model.py\n===================================================================\n--- train_current_model.py\tCurrent\n+++ train_current_model.py\tProposed\n@@ -2054,6 +2054,8 @@\n learning_rate=args.learning_rate,\n loss_microbatch_size=args.loss_microbatch_size,\n negative_filter_scope=args.negative_filter_scope,\n+ objective=args.objective,\n+ temperature=args.temperature,\n target_weighting=args.target_weighting,\n ordinary_target_weight=args.ordinary_target_weight,\n organic_strong_target_weight=args.organic_strong_target_weight,\n",
"fileName": "train_current_model.py",
"filePath": "/app/solution/train_current_model.py",
"originalContent": "#!/usr/bin/env python3\n\"\"\"Train the Yambda windowed two-tower baseline.\n\nEvery catalog listen with at least 50 percent completion is a sequence token.\nWithin each selected chronological window, every adjacent target contributes a\nloss. The fast default retains each user's most recent 128 training targets and\n16 held-out validation targets; passing zero for either cap restores the\nuncapped behavior for that side of the split. A 100-step truncated-BPTT block\ncomputes all of its timestep losses and carries the detached GRU hidden state\ninto the next block.\n\nThe model is a pure two-tower retriever: no user-ID embedding, no output\nnormalization, raw dot-product retrieval, sample-weighted binary cross entropy,\nand 24 uniform plus eight propensity-corrected unique in-batch negatives. Every\nselected adjacent target retains a nonzero weight. Only files under the supplied\npublic directory are read.\n\"\"\"\nfrom __future__ import annotations\n\nimport argparse\nimport json\nimport random\nimport sys\nimport time\nfrom dataclasses import asdict, dataclass\nfrom pathlib import Path\n\nimport numpy as np\nimport pandas as pd\nimport torch\nfrom torch import nn\nfrom torch.nn import functional as F\n\nimport two_tower_model as base\n\n\nCACHE_VERSION = 2\n\n\n@dataclass(frozen=True)\nclass Config:\n bptt_steps: int = 100\n user_batch_size: int = 64\n embedding_dim: int = 192\n hidden_dim: int = 192\n user_hidden_1: int = 512\n user_hidden_2: int = 256\n item_hidden_1: int = 512\n item_hidden_2: int = 256\n dropout: float = 0.05\n epochs: int = 1\n learning_rate: float = 1.0e-3\n weight_decay: float = 1.0e-5\n gradient_clip: float = 2.0\n loss_microbatch_size: int = 65_536\n uniform_negatives: int = 24\n inbatch_negatives: int = 8\n inbatch_correction_power: float = 0.5\n inbatch_correction_min: float = 0.25\n inbatch_correction_max: float = 4.0\n negative_filter_scope: str = \"full-training-history\"\n objective: str = \"bce\"\n temperature: float = 0.2\n target_weighting: str = \"organic-strong\"\n ordinary_target_weight: float = 0.25\n organic_strong_target_weight: float = 1.0\n organic_like_target_weight: float = 3.0\n discovery_horizon_days: int = 7\n validation_fraction: float = 0.10\n max_train_targets_per_user: int = 128\n max_validation_targets_per_user: int = 16\n early_stopping_patience: int = 1\n early_stopping_min_delta: float = 1.0e-4\n seed: int = 29\n\n\n@dataclass(frozen=True)\nclass CandidateSet:\n uids: np.ndarray\n item_ids: np.ndarray\n packed: np.ndarray\n\n @classmethod\n def read(cls, path: Path) -> \"CandidateSet\":\n with np.load(path, allow_pickle=False) as artifact:\n result = cls(\n uids=artifact[\"uids\"].astype(np.int64, copy=True),\n item_ids=artifact[\"item_ids\"].astype(np.int64, copy=True),\n packed=artifact[\"packed_eligible\"].astype(np.uint8, copy=True),\n )\n expected = (len(result.uids), (len(result.item_ids) + 7) // 8)\n if result.packed.shape != expected:\n raise ValueError(\n f\"invalid candidate artifact shape {result.packed.shape}; \"\n f\"expected {expected}\"\n )\n return result\n\n\n@dataclass(frozen=True)\nclass SequenceData:\n item_ids: np.ndarray\n item_features: np.ndarray\n user_ids: np.ndarray\n offsets: np.ndarray\n sequence_timestamps: np.ndarray\n sequence_items: np.ndarray\n event_features: np.ndarray\n activity_features: np.ndarray\n state_ids: np.ndarray\n liked_states: np.ndarray\n disliked_states: np.ndarray\n state_counts: np.ndarray\n inference_user_rows: np.ndarray\n inference_liked_items: np.ndarray\n inference_disliked_items: np.ndarray\n inference_user_features: np.ndarray\n\n @property\n def lengths(self) -> np.ndarray:\n return np.diff(self.offsets)\n\n @property\n def target_count(self) -> int:\n return int(np.maximum(self.lengths - 1, 0).sum())\n\n\n@dataclass(frozen=True)\nclass TargetPartition:\n \"\"\"A contiguous recent target window and its chronological BCE split.\"\"\"\n\n window_starts: np.ndarray\n training_counts: np.ndarray\n validation_counts: np.ndarray\n history_training_counts: np.ndarray\n\n @property\n def selected_count(self) -> int:\n return int(self.training_counts.sum() + self.validation_counts.sum())\n\n\ndef log(message: str) -> None:\n print(message, file=sys.stderr, flush=True)\n\n\ndef seed_everything(seed: int) -> None:\n random.seed(seed)\n np.random.seed(seed)\n torch.manual_seed(seed)\n\n\ndef model_config(config: Config) -> base.TowerConfig:\n \"\"\"Map the dense trainer configuration onto the unchanged tower.\"\"\"\n return base.TowerConfig(\n sequence_length=config.bptt_steps,\n embedding_dim=config.embedding_dim,\n hidden_dim=config.hidden_dim,\n user_hidden_1=config.user_hidden_1,\n user_hidden_2=config.user_hidden_2,\n item_hidden_1=config.item_hidden_1,\n item_hidden_2=config.item_hidden_2,\n dropout=config.dropout,\n epochs=config.epochs,\n learning_rate=config.learning_rate,\n weight_decay=config.weight_decay,\n gradient_clip=config.gradient_clip,\n uniform_negatives=config.uniform_negatives,\n inbatch_negatives=config.inbatch_negatives,\n inbatch_correction_power=config.inbatch_correction_power,\n inbatch_correction_min=config.inbatch_correction_min,\n inbatch_correction_max=config.inbatch_correction_max,\n seed=config.seed,\n )\n\n\ndef append_state_snapshot(\n active_likes: dict[int, int],\n active_dislikes: dict[int, int],\n n_items: int,\n liked_states: list[np.ndarray],\n disliked_states: list[np.ndarray],\n state_counts: list[tuple[int, int]],\n) -> int:\n liked_states.append(base.pack_active_items(active_likes, base.LIKE_ITEMS, n_items))\n disliked_states.append(\n base.pack_active_items(active_dislikes, base.DISLIKE_ITEMS, n_items)\n )\n state_counts.append((len(active_likes), len(active_dislikes)))\n return len(liked_states) - 1\n\n\ndef causal_activity_features(\n timestamps: np.ndarray,\n items: np.ndarray,\n organic: np.ndarray,\n completion: np.ndarray,\n) -> np.ndarray:\n \"\"\"Compute the first four reviewed user features at every event cutoff.\"\"\"\n length = len(items)\n result = np.zeros((length, 4), dtype=np.float16)\n if not length:\n return result\n left_rows = np.searchsorted(\n timestamps,\n timestamps - base.PROFILE_LOOKBACK_DAYS * base.DAY,\n side=\"left\",\n )\n rows = np.arange(length, dtype=np.int64)\n counts = rows - left_rows + 1\n organic_sum = np.cumsum(organic.astype(np.float64))\n completion_sum = np.cumsum(np.minimum(completion, 100).astype(np.float64))\n before_organic = np.where(\n left_rows > 0, organic_sum[np.maximum(left_rows - 1, 0)], 0.0\n )\n before_completion = np.where(\n left_rows > 0, completion_sum[np.maximum(left_rows - 1, 0)], 0.0\n )\n\n unique_counts = np.empty(length, dtype=np.int32)\n frequencies: dict[int, int] = {}\n left = 0\n for right, raw_item in enumerate(items):\n item = int(raw_item)\n frequencies[item] = frequencies.get(item, 0) + 1\n required_left = int(left_rows[right])\n while left < required_left:\n old_item = int(items[left])\n remaining = frequencies[old_item] - 1\n if remaining:\n frequencies[old_item] = remaining\n else:\n del frequencies[old_item]\n left += 1\n unique_counts[right] = len(frequencies)\n\n result[:, 0] = (np.log1p(counts) / 8.0).astype(np.float16)\n result[:, 1] = (np.log1p(unique_counts) / 6.0).astype(np.float16)\n result[:, 2] = (\n (organic_sum - before_organic) / np.maximum(counts, 1)\n ).astype(np.float16)\n result[:, 3] = (\n (completion_sum - before_completion)\n / np.maximum(counts, 1)\n / 100.0\n ).astype(np.float16)\n return result\n\n\ndef build_sequence_parts(\n listens: base.ListenEvents,\n explicit: base.ExplicitEvents,\n n_items: int,\n liked_states: list[np.ndarray],\n disliked_states: list[np.ndarray],\n state_counts: list[tuple[int, int]],\n) -> tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray, np.ndarray] | None:\n qualifying = listens.completion_pct >= base.LISTEN_POSITIVE_THRESHOLD\n timestamps = listens.timestamps[qualifying].astype(np.int64, copy=False)\n items = listens.items[qualifying].astype(np.uint16, copy=False)\n organic = listens.organic[qualifying].astype(np.float32, copy=False)\n completion = listens.completion_pct[qualifying].astype(np.float32, copy=False)\n if not len(items):\n return None\n\n event_features = np.zeros((len(items), 3), dtype=np.float16)\n event_features[:, 0] = organic.astype(np.float16)\n event_features[:, 1] = (\n np.minimum(completion, 100.0) / 100.0\n ).astype(np.float16)\n if len(items) > 1:\n gaps = np.maximum(0, timestamps[1:] - timestamps[:-1])\n event_features[1:, 2] = np.exp(\n -gaps.astype(np.float64) / (60.0 * base.DAY)\n ).astype(np.float16)\n\n activity = causal_activity_features(\n timestamps, items, organic, completion\n )\n states = np.zeros(len(items), dtype=np.uint32)\n active_likes: dict[int, int] = {}\n active_dislikes: dict[int, int] = {}\n explicit_position = 0\n current_state = 0\n for row, timestamp in enumerate(timestamps):\n changed = False\n while (\n explicit_position < len(explicit.timestamps)\n and int(explicit.timestamps[explicit_position]) < int(timestamp)\n ):\n item = int(explicit.items[explicit_position])\n event_timestamp = int(explicit.timestamps[explicit_position])\n destination = (\n active_likes\n if int(explicit.channel[explicit_position]) == 0\n else active_dislikes\n )\n if int(explicit.active[explicit_position]) == 1:\n destination[item] = event_timestamp\n else:\n destination.pop(item, None)\n explicit_position += 1\n changed = True\n if changed:\n current_state = append_state_snapshot(\n active_likes,\n active_dislikes,\n n_items,\n liked_states,\n disliked_states,\n state_counts,\n )\n states[row] = np.uint32(current_state)\n return timestamps, items.copy(), event_features, activity, states\n\n\ndef final_user_features(\n timestamps: np.ndarray,\n items: np.ndarray,\n event_features: np.ndarray,\n history_end: int,\n like_count: int,\n dislike_count: int,\n) -> np.ndarray:\n start = int(\n np.searchsorted(\n timestamps,\n history_end - base.PROFILE_LOOKBACK_DAYS * base.DAY,\n side=\"left\",\n )\n )\n result = np.zeros(6, dtype=np.float32)\n recent_items = items[start:]\n if len(recent_items):\n result[0] = np.log1p(len(recent_items)) / 8.0\n result[1] = np.log1p(len(np.unique(recent_items))) / 6.0\n result[2] = float(np.mean(event_features[start:, 0]))\n result[3] = float(np.mean(event_features[start:, 1]))\n result[4] = np.log1p(like_count) / 6.0\n result[5] = np.log1p(dislike_count) / 4.0\n return result\n\n\ndef prepare_data(\n public: Path, stage: str, config: Config\n) -> tuple[SequenceData, CandidateSet]:\n data = public / \"data\"\n catalog = pd.read_parquet(data / \"candidate_catalog.parquet\").sort_values(\n \"popularity_rank\"\n )\n item_ids = catalog.item_id.to_numpy(dtype=np.int64, copy=True)\n item_index = pd.Index(item_ids)\n n_items = len(item_ids)\n if n_items >= np.iinfo(np.uint16).max:\n raise RuntimeError(\"uint16 item storage is too small for this catalog\")\n\n if stage == \"validation\":\n prefixes = (\"train\",)\n history_end = base.TRAIN_END\n candidate_name = \"validation_eligible_candidates.npz\"\n state_name = \"validation_preference_state.parquet\"\n else:\n prefixes = (\"train\", \"val\")\n history_end = base.VALIDATION_END\n candidate_name = \"test_eligible_candidates.npz\"\n state_name = \"test_preference_state.parquet\"\n eligibility = CandidateSet.read(data / candidate_name)\n if not np.array_equal(eligibility.item_ids, item_ids):\n raise RuntimeError(\"catalog and candidate artifact item order differ\")\n\n explicit_by_user = base.load_explicit_events(data, prefixes, item_index)\n inference_likes, inference_dislikes, like_counts, dislike_counts = (\n base.load_inference_state(\n data / state_name,\n item_index,\n eligibility.uids,\n n_items,\n )\n )\n inference_rows = {int(uid): row for row, uid in enumerate(eligibility.uids)}\n inference_user_features = np.zeros((len(eligibility.uids), 6), dtype=np.float32)\n inference_sequence_rows = np.full(len(eligibility.uids), -1, dtype=np.int32)\n\n # Snapshot zero is the shared empty explicit state.\n liked_states = [np.full(base.LIKE_ITEMS, n_items, dtype=np.uint16)]\n disliked_states = [\n np.full(base.DISLIKE_ITEMS, n_items, dtype=np.uint16)\n ]\n state_counts: list[tuple[int, int]] = [(0, 0)]\n user_ids: list[int] = []\n offsets = [0]\n timestamp_parts: list[np.ndarray] = []\n item_parts: list[np.ndarray] = []\n event_parts: list[np.ndarray] = []\n activity_parts: list[np.ndarray] = []\n state_parts: list[np.ndarray] = []\n\n listen_paths = [data / f\"{prefix}_listens.parquet\" for prefix in prefixes]\n started = time.monotonic()\n for seen_users, (uid, listens) in enumerate(\n base.merge_listen_streams(listen_paths, item_index), start=1\n ):\n parts = build_sequence_parts(\n listens,\n explicit_by_user.get(uid, base.empty_explicit_events()),\n n_items,\n liked_states,\n disliked_states,\n state_counts,\n )\n if parts is None:\n continue\n sequence_row = len(user_ids)\n user_ids.append(uid)\n timestamp_parts.append(parts[0].astype(np.uint32, copy=False))\n item_parts.append(parts[1])\n event_parts.append(parts[2])\n activity_parts.append(parts[3])\n state_parts.append(parts[4])\n offsets.append(offsets[-1] + len(parts[1]))\n inference_row = inference_rows.get(uid)\n if inference_row is not None:\n inference_sequence_rows[inference_row] = sequence_row\n inference_user_features[inference_row] = final_user_features(\n parts[0],\n parts[1],\n parts[2],\n history_end,\n int(like_counts[inference_row]),\n int(dislike_counts[inference_row]),\n )\n if seen_users % 1_000 == 0:\n log(\n f\"prepared users={seen_users:,} tokens={offsets[-1]:,} \"\n f\"states={len(liked_states):,} \"\n f\"elapsed={time.monotonic() - started:.1f}s\"\n )\n\n missing = np.flatnonzero(inference_sequence_rows < 0)\n if len(missing):\n inference_user_features[missing, 4] = np.log1p(like_counts[missing]) / 6.0\n inference_user_features[missing, 5] = (\n np.log1p(dislike_counts[missing]) / 4.0\n )\n log(f\"target users without qualifying catalog sequence={len(missing):,}\")\n\n prepared = SequenceData(\n item_ids=item_ids,\n item_features=base.build_item_features(catalog),\n user_ids=np.asarray(user_ids, dtype=np.int64),\n offsets=np.asarray(offsets, dtype=np.int64),\n sequence_timestamps=np.concatenate(timestamp_parts),\n sequence_items=np.concatenate(item_parts),\n event_features=np.concatenate(event_parts),\n activity_features=np.concatenate(activity_parts),\n state_ids=np.concatenate(state_parts),\n liked_states=np.stack(liked_states),\n disliked_states=np.stack(disliked_states),\n state_counts=np.asarray(state_counts, dtype=np.int32),\n inference_user_rows=inference_sequence_rows,\n inference_liked_items=inference_likes,\n inference_disliked_items=inference_dislikes,\n inference_user_features=inference_user_features,\n )\n log(\n f\"prepared stage={stage} sequence_users={len(prepared.user_ids):,} \"\n f\"tokens={len(prepared.sequence_items):,} \"\n f\"next_item_targets={prepared.target_count:,} \"\n f\"states={len(prepared.liked_states):,}\"\n )\n return prepared, eligibility\n\n\ndef cache_metadata(stage: str, config: Config) -> dict[str, object]:\n return {\n \"cache_version\": CACHE_VERSION,\n \"stage\": stage,\n \"positive_listen_threshold\": base.LISTEN_POSITIVE_THRESHOLD,\n \"profile_lookback_days\": base.PROFILE_LOOKBACK_DAYS,\n \"event_time_feature\": \"exp-negative-gap-over-60-days\",\n }\n\n\ndef save_cache(\n path: Path, prepared: SequenceData, stage: str, config: Config\n) -> None:\n path.parent.mkdir(parents=True, exist_ok=True)\n np.savez(\n path,\n metadata=np.asarray(json.dumps(cache_metadata(stage, config), sort_keys=True)),\n **prepared.__dict__,\n )\n log(f\"saved public-derived full-sequence cache: {path}\")\n\n\ndef load_cache(path: Path, stage: str, config: Config) -> SequenceData:\n with np.load(path, allow_pickle=False) as artifact:\n actual = json.loads(str(artifact[\"metadata\"].item()))\n expected = cache_metadata(stage, config)\n if actual != expected:\n raise RuntimeError(f\"sequence cache differs: {actual} != {expected}\")\n values = {\n field: artifact[field].copy()\n for field in SequenceData.__dataclass_fields__\n }\n prepared = SequenceData(**values)\n log(\n f\"loaded full-sequence cache: tokens={len(prepared.sequence_items):,} \"\n f\"targets={prepared.target_count:,}\"\n )\n return prepared\n\n\ndef load_or_prepare(\n public: Path,\n stage: str,\n config: Config,\n cache: Path | None,\n) -> tuple[SequenceData, CandidateSet]:\n candidate_name = (\n \"validation_eligible_candidates.npz\"\n if stage == \"validation\"\n else \"test_eligible_candidates.npz\"\n )\n eligibility = CandidateSet.read(public / \"data\" / candidate_name)\n if cache is not None and cache.is_file():\n prepared = load_cache(cache, stage, config)\n if not np.array_equal(prepared.item_ids, eligibility.item_ids):\n raise RuntimeError(\"cached catalog differs from candidate artifact\")\n return prepared, eligibility\n prepared, built_eligibility = prepare_data(public, stage, config)\n if cache is not None:\n save_cache(cache, prepared, stage, config)\n return prepared, built_eligibility\n\n\ndef build_item_correction(\n prepared: SequenceData,\n config: Config,\n target_starts: np.ndarray,\n target_counts: np.ndarray,\n) -> torch.Tensor:\n \"\"\"Estimate in-batch propensities from the selected training targets.\"\"\"\n counts = np.zeros(len(prepared.item_ids), dtype=np.float64)\n for user_row, count in enumerate(target_counts):\n if not count:\n continue\n offset = int(prepared.offsets[user_row])\n target_start = int(target_starts[user_row])\n counts += np.bincount(\n prepared.sequence_items[\n offset\n + target_start\n + 1 : offset\n + target_start\n + int(count)\n + 1\n ].astype(np.int64),\n minlength=len(prepared.item_ids),\n )\n total = float(counts.sum())\n probabilities = (counts + 1.0) / (total + len(counts))\n raw = np.power(\n (1.0 / len(counts)) / probabilities,\n config.inbatch_correction_power,\n )\n normalization = float(np.sum(counts * raw) / max(total, 1.0))\n correction = np.clip(\n raw / max(normalization, 1.0e-12),\n config.inbatch_correction_min,\n config.inbatch_correction_max,\n ).astype(np.float32)\n mean_on_targets = float(np.sum(counts * correction) / max(total, 1.0))\n log(\n f\"inbatch correction: min={correction.min():.3f} \"\n f\"mean_on_targets={mean_on_targets:.3f} max={correction.max():.3f}\"\n )\n return torch.from_numpy(correction)\n\n\ndef build_known_positive_filter(\n prepared: SequenceData, training_counts: np.ndarray\n) -> np.ndarray:\n \"\"\"Pack each user's complete public-training positive-item set.\"\"\"\n width = (len(prepared.item_ids) + 7) // 8\n packed = np.zeros((len(prepared.user_ids), width), dtype=np.uint8)\n retained = 0\n for user_row, target_count in enumerate(training_counts):\n if not target_count:\n continue\n offset = int(prepared.offsets[user_row])\n # T adjacent targets consume sequence positions 0 through T. The\n # chronological BCE suffix starts at T + 1 and is intentionally absent.\n items = np.unique(\n prepared.sequence_items[\n offset : offset + int(target_count) + 1\n ].astype(np.int64)\n )\n byte_columns = items >> 3\n bit_values = np.left_shift(\n np.uint8(1), (items & 7).astype(np.uint8)\n )\n np.bitwise_or.at(packed[user_row], byte_columns, bit_values)\n retained += len(items)\n log(\n f\"full-history false-negative filter: users={len(prepared.user_ids):,} \"\n f\"unique_user_items={retained:,} bytes={packed.nbytes:,}\"\n )\n return packed\n\n\ndef invalid_full_history_negatives(\n negatives: np.ndarray,\n targets: np.ndarray,\n user_rows: np.ndarray,\n known_positive_packed: np.ndarray,\n liked_items: np.ndarray,\n) -> np.ndarray:\n \"\"\"Return row-level invalid negatives using the complete training prefix.\"\"\"\n byte_columns = negatives >> 3\n bit_columns = negatives & 7\n known = known_positive_packed[user_rows[:, None], byte_columns]\n invalid = ((known >> bit_columns) & 1).astype(bool)\n invalid |= negatives == targets[:, None]\n invalid |= (negatives[:, :, None] == liked_items[:, None, :]).any(axis=2)\n for column in range(1, negatives.shape[1]):\n invalid[:, column] |= (\n negatives[:, column, None] == negatives[:, :column]\n ).any(axis=1)\n return invalid\n\n\ndef draw_full_history_filtered_negatives(\n targets: torch.Tensor,\n user_rows: np.ndarray,\n liked_items: torch.Tensor,\n known_positive_packed: np.ndarray,\n n_items: int,\n config: Config,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n \"\"\"Draw mixed negatives and reject all known public-training positives.\"\"\"\n targets_cpu = targets.detach().cpu().numpy().astype(np.int64, copy=False)\n liked_cpu = liked_items.detach().cpu().numpy().astype(np.int64, copy=False)\n batch_size = len(targets_cpu)\n uniform = torch.randint(\n n_items,\n (batch_size, config.uniform_negatives),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.empty((batch_size, 0), dtype=torch.long)\n corrected_count = 0\n if config.inbatch_negatives:\n unique_targets = torch.unique(targets.detach().cpu())\n corrected_count = min(\n config.inbatch_negatives, max(len(unique_targets) - 1, 0)\n )\n if corrected_count:\n priorities = torch.rand(\n (batch_size, len(unique_targets)), generator=generator\n )\n priorities.masked_fill_(\n unique_targets[None, :].eq(\n torch.from_numpy(targets_cpu)[:, None]\n ),\n -1.0,\n )\n columns = torch.topk(\n priorities, k=corrected_count, dim=1\n ).indices\n inbatch = unique_targets[columns]\n if corrected_count < config.inbatch_negatives:\n filler = torch.randint(\n n_items,\n (batch_size, config.inbatch_negatives - corrected_count),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.cat([inbatch, filler], dim=1)\n\n negatives = torch.cat([uniform, inbatch], dim=1).numpy()\n correction = np.ones_like(negatives, dtype=np.float32)\n if corrected_count:\n correction[\n :, config.uniform_negatives : config.uniform_negatives + corrected_count\n ] = item_correction[inbatch[:, :corrected_count]].numpy()\n for _ in range(32):\n invalid = invalid_full_history_negatives(\n negatives,\n targets_cpu,\n user_rows,\n known_positive_packed,\n liked_cpu,\n )\n if not np.any(invalid):\n return (\n torch.from_numpy(negatives).to(device),\n torch.from_numpy(correction).to(device),\n )\n replacements = torch.randint(\n n_items,\n (int(invalid.sum()),),\n generator=generator,\n device=\"cpu\",\n ).numpy()\n negatives[invalid] = replacements\n # Rejected in-batch items become ordinary uniform draws.\n correction[invalid] = 1.0\n invalid = invalid_full_history_negatives(\n negatives,\n targets_cpu,\n user_rows,\n known_positive_packed,\n liked_cpu,\n )\n if np.any(invalid):\n raise RuntimeError(\"full-history negative resampling failed\")\n return (\n torch.from_numpy(negatives).to(device),\n torch.from_numpy(correction).to(device),\n )\n\n\ndef draw_training_negatives(\n positive_items: torch.Tensor,\n known_history: torch.Tensor,\n liked_items: torch.Tensor,\n user_rows: np.ndarray,\n known_positive_packed: np.ndarray | None,\n prepared: SequenceData,\n config: Config,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n if config.negative_filter_scope == \"full-training-history\":\n if known_positive_packed is None:\n raise RuntimeError(\"full-history negative filter was not built\")\n return draw_full_history_filtered_negatives(\n positive_items,\n user_rows,\n liked_items,\n known_positive_packed,\n len(prepared.item_ids),\n config,\n item_correction,\n generator,\n device,\n )\n return base.draw_filtered_mixed_negatives(\n positive_items,\n known_history,\n liked_items,\n len(prepared.item_ids),\n config.uniform_negatives,\n config.inbatch_negatives,\n item_correction,\n generator,\n device,\n )\n\n\ndef event_sequence(\n model: base.SequentialTwoTower,\n items: torch.Tensor,\n features: torch.Tensor,\n hidden: torch.Tensor,\n) -> tuple[torch.Tensor, torch.Tensor]:\n inputs = torch.cat([model.context_embedding(items), features], dim=-1)\n projected = model.event_projection(inputs)\n return model.sequence_encoder(projected, hidden)\n\n\ndef pool_explicit_states(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n state_ids: np.ndarray,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor, torch.Tensor]:\n unique_states, inverse = np.unique(state_ids, return_inverse=True)\n liked_items = torch.from_numpy(\n prepared.liked_states[unique_states].astype(np.int64)\n ).to(device)\n disliked_items = torch.from_numpy(\n prepared.disliked_states[unique_states].astype(np.int64)\n ).to(device)\n liked_vectors = model._mean_context_channel(liked_items)\n disliked_vectors = model._mean_context_channel(disliked_items)\n inverse_rows = torch.from_numpy(inverse.astype(np.int64)).to(device)\n counts = prepared.state_counts[state_ids]\n return (\n liked_vectors[inverse_rows],\n disliked_vectors[inverse_rows],\n torch.from_numpy(counts.astype(np.float32)).to(device),\n )\n\n\ndef user_queries(\n model: base.SequentialTwoTower,\n sequence_vectors: torch.Tensor,\n liked_vectors: torch.Tensor,\n disliked_vectors: torch.Tensor,\n user_features: torch.Tensor,\n) -> torch.Tensor:\n return model.user_tower(\n torch.cat(\n [sequence_vectors, liked_vectors, disliked_vectors, user_features],\n dim=1,\n )\n )\n\n\ndef candidate_vectors(\n model: base.SequentialTwoTower,\n positives: torch.Tensor,\n negatives: torch.Tensor,\n) -> tuple[torch.Tensor, torch.Tensor]:\n all_items = torch.cat([positives, negatives.reshape(-1)])\n unique_items, inverse = torch.unique(all_items, return_inverse=True)\n encoded = model.encode_items(unique_items)\n positive_vectors = encoded[inverse[: len(positives)]]\n negative_vectors = encoded[inverse[len(positives) :]].reshape(\n len(positives), negatives.shape[1], -1\n )\n return positive_vectors, negative_vectors\n\n\ndef history_windows(\n sequence: np.ndarray,\n start: int,\n count: int,\n width: int,\n padding_item: int,\n) -> np.ndarray:\n base_start = max(0, start - width + 1)\n segment = sequence[base_start : start + count]\n padded = np.concatenate(\n [np.full(width - 1, padding_item, dtype=np.uint16), segment]\n )\n windows = np.lib.stride_tricks.sliding_window_view(padded, width)\n offset = start - base_start\n return windows[offset : offset + count].copy()\n\n\ndef corrected_negative_loss(\n logits: torch.Tensor,\n correction: torch.Tensor,\n config: Config,\n) -> torch.Tensor:\n raw = F.softplus(logits)\n total = config.uniform_negatives + config.inbatch_negatives\n result = torch.zeros(len(logits), device=logits.device)\n if config.uniform_negatives:\n result += config.uniform_negatives * raw[\n :, : config.uniform_negatives\n ].mean(dim=1)\n if config.inbatch_negatives:\n selected = slice(config.uniform_negatives, total)\n weights = correction[:, selected]\n result += config.inbatch_negatives * (\n (raw[:, selected] * weights).sum(dim=1)\n / weights.sum(dim=1).clamp_min(1.0e-6)\n )\n return result / total\n\n\ndef retained_target_window_starts(\n total_targets: np.ndarray,\n validation_counts: np.ndarray,\n training_counts: np.ndarray,\n) -> np.ndarray:\n \"\"\"Return the discarded prefix before the retained chronological suffix.\"\"\"\n training_region_ends = total_targets - validation_counts\n window_starts = training_region_ends - training_counts\n if (\n np.any(window_starts < 0)\n or np.any(training_counts < 0)\n or np.any(validation_counts < 0)\n ):\n raise RuntimeError(\"chronological target partition produced negative counts\")\n if not np.array_equal(\n window_starts + training_counts + validation_counts, total_targets\n ):\n raise RuntimeError(\"chronological target window is not a final contiguous suffix\")\n return window_starts\n\n\ndef chronological_target_partition(\n prepared: SequenceData,\n validation_fraction: float,\n max_train_targets_per_user: int,\n max_validation_targets_per_user: int,\n) -> TargetPartition:\n \"\"\"Select a recent contiguous window with a final out-of-time BCE suffix.\"\"\"\n totals = np.maximum(prepared.lengths - 1, 0).astype(np.int64)\n validation = np.floor(totals * validation_fraction).astype(np.int64)\n validation[(totals >= 2) & (validation == 0)] = 1\n if max_validation_targets_per_user:\n validation = np.minimum(validation, max_validation_targets_per_user)\n\n history_training = totals - validation\n training = history_training.copy()\n if max_train_targets_per_user:\n training = np.minimum(training, max_train_targets_per_user)\n window_starts = retained_target_window_starts(totals, validation, training)\n return TargetPartition(\n window_starts=window_starts,\n training_counts=training,\n validation_counts=validation,\n history_training_counts=history_training,\n )\n\n\ndef load_organic_like_outcomes(\n public: Path, stage: str, item_ids: np.ndarray\n) -> dict[int, dict[int, np.ndarray]]:\n \"\"\"Load mapped organic-like timestamps from public fitting history.\"\"\"\n prefixes = (\"train\",) if stage == \"validation\" else (\"train\", \"val\")\n frames = [\n pd.read_parquet(\n public / \"data\" / f\"{prefix}_likes.parquet\",\n columns=[\"uid\", \"timestamp\", \"item_id\", \"is_organic\"],\n )\n for prefix in prefixes\n ]\n likes = pd.concat(frames, ignore_index=True)\n likes = likes[likes.is_organic == 1].copy()\n mapped = pd.Index(item_ids).get_indexer(\n likes.item_id.to_numpy(dtype=np.int64)\n )\n likes = likes.loc[mapped >= 0, [\"uid\", \"timestamp\"]].copy()\n likes[\"item\"] = mapped[mapped >= 0].astype(np.int32, copy=False)\n likes = likes.sort_values([\"uid\", \"item\", \"timestamp\"], kind=\"stable\")\n nested: dict[int, dict[int, list[int]]] = {}\n for raw_uid, raw_timestamp, raw_item in likes.itertuples(\n index=False, name=None\n ):\n nested.setdefault(int(raw_uid), {}).setdefault(int(raw_item), []).append(\n int(raw_timestamp)\n )\n result = {\n uid: {\n item: np.asarray(timestamps, dtype=np.int64)\n for item, timestamps in items.items()\n }\n for uid, items in nested.items()\n }\n log(\n f\"loaded organic-like outcomes: rows={len(likes):,} \"\n f\"users={len(result):,}\"\n )\n return result\n\n\ndef build_target_weights(\n public: Path,\n stage: str,\n prepared: SequenceData,\n training_counts: np.ndarray,\n config: Config,\n) -> np.ndarray:\n \"\"\"Build nonzero per-token weights without dropping adjacent targets.\"\"\"\n if config.target_weighting == \"uniform\":\n weights = np.ones(len(prepared.sequence_items), dtype=np.float32)\n else:\n weights = np.full(\n len(prepared.sequence_items),\n config.ordinary_target_weight,\n dtype=np.float32,\n )\n organic_strong = (\n (prepared.event_features[:, 0] > 0.5)\n & (prepared.event_features[:, 1] >= 0.8)\n )\n weights[organic_strong] = config.organic_strong_target_weight\n if config.target_weighting == \"discovery-graded\":\n like_outcomes = load_organic_like_outcomes(\n public, stage, prepared.item_ids\n )\n history_end = (\n base.TRAIN_END if stage == \"validation\" else base.VALIDATION_END\n )\n horizon = config.discovery_horizon_days * base.DAY\n grade_three = 0\n for user_row, uid in enumerate(prepared.user_ids):\n start = int(prepared.offsets[user_row])\n end = int(prepared.offsets[user_row + 1])\n if end - start < 2:\n continue\n last_training_target = start + int(training_counts[user_row])\n first_validation_target = last_training_target + 1\n split_timestamp = (\n int(prepared.sequence_timestamps[first_validation_target])\n if first_validation_target < end\n else history_end\n )\n user_likes = like_outcomes.get(int(uid), {})\n seen = {int(prepared.sequence_items[start])}\n for position in range(start + 1, end):\n item = int(prepared.sequence_items[position])\n novel = item not in seen\n seen.add(item)\n if not novel:\n # Repeats remain training examples, but they cannot be\n # eligible discovery outcomes at the corresponding\n # cutoff and therefore keep only the ordinary weight.\n weights[position] = config.ordinary_target_weight\n continue\n timestamps = user_likes.get(item)\n if timestamps is None:\n continue\n timestamp = int(prepared.sequence_timestamps[position])\n upper = min(\n timestamp + horizon,\n split_timestamp\n if position <= last_training_target\n else history_end,\n )\n outcome = int(np.searchsorted(timestamps, timestamp, side=\"left\"))\n if outcome < len(timestamps) and int(timestamps[outcome]) < upper:\n weights[position] = config.organic_like_target_weight\n grade_three += 1\n log(\n f\"grade-3 first-encounter targets={grade_three:,} \"\n f\"horizon_days={config.discovery_horizon_days}\"\n )\n log(\n f\"target weighting={config.target_weighting} \"\n f\"min={weights.min():.3f} mean={weights.mean():.3f} \"\n f\"max={weights.max():.3f}\"\n )\n return weights\n\n\ndef build_dense_block(\n prepared: SequenceData,\n target_weights: np.ndarray,\n users: np.ndarray,\n window_starts: np.ndarray,\n process_counts: np.ndarray,\n score_starts: np.ndarray,\n start: int,\n steps: int,\n config: Config,\n) -> tuple[np.ndarray, ...]:\n \"\"\"Build one causal sequence block and mark the target rows to score.\"\"\"\n batch_size = len(users)\n padding_item = len(prepared.item_ids)\n input_items = np.full((batch_size, steps), padding_item, dtype=np.uint16)\n event_features = np.zeros((batch_size, steps, 3), dtype=np.float16)\n activity_features = np.zeros((batch_size, steps, 4), dtype=np.float16)\n state_ids = np.zeros((batch_size, steps), dtype=np.uint32)\n targets = np.zeros((batch_size, steps), dtype=np.uint16)\n weights = np.ones((batch_size, steps), dtype=np.float32)\n scored = np.zeros((batch_size, steps), dtype=bool)\n histories = np.full(\n (batch_size, steps, config.bptt_steps),\n padding_item,\n dtype=np.uint16,\n )\n processed = np.zeros(batch_size, dtype=np.int32)\n\n lengths = prepared.lengths\n for row, user_row in enumerate(users):\n count = min(steps, max(int(process_counts[row]) - start, 0))\n if not count:\n continue\n offset = int(prepared.offsets[user_row])\n window_start = int(window_starts[row])\n absolute_start = window_start + start\n positions = slice(\n offset + absolute_start,\n offset + absolute_start + count,\n )\n target_positions = slice(\n offset + absolute_start + 1,\n offset + absolute_start + count + 1,\n )\n input_items[row, :count] = prepared.sequence_items[positions]\n event_features[row, :count] = prepared.event_features[positions]\n activity_features[row, :count] = prepared.activity_features[positions]\n state_ids[row, :count] = prepared.state_ids[positions]\n targets[row, :count] = prepared.sequence_items[target_positions]\n weights[row, :count] = target_weights[target_positions]\n score_from = min(count, max(int(score_starts[row]) - start, 0))\n scored[row, score_from:count] = True\n processed[row] = count\n user_sequence = prepared.sequence_items[\n offset : offset + int(lengths[user_row])\n ]\n histories[row, :count] = history_windows(\n user_sequence,\n absolute_start,\n count,\n config.bptt_steps,\n padding_item,\n )\n return (\n input_items,\n event_features,\n activity_features,\n state_ids,\n targets,\n scored,\n histories,\n processed,\n weights,\n )\n\n\ndef score_target_microbatches(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n sequence_vectors: torch.Tensor,\n flat_states: np.ndarray,\n flat_activity: np.ndarray,\n positive_items: torch.Tensor,\n negatives: torch.Tensor,\n negative_weights: torch.Tensor,\n sample_weights: np.ndarray,\n device: torch.device,\n backward: bool,\n) -> tuple[dict[str, float], torch.Tensor | None]:\n \"\"\"Compute one logical block's exact BCE in bounded-memory slices.\"\"\"\n target_count = len(positive_items)\n total_weight = float(sample_weights.sum())\n if total_weight <= 0:\n raise RuntimeError(\"target block has no positive sample weight\")\n sequence_gradient = torch.zeros_like(sequence_vectors) if backward else None\n statistics = {\n \"loss_sum\": 0.0,\n \"positive_sum\": 0.0,\n \"negative_sum\": 0.0,\n \"pair_correct\": 0.0,\n \"weight_sum\": 0.0,\n }\n for begin in range(0, target_count, config.loss_microbatch_size):\n end = min(begin + config.loss_microbatch_size, target_count)\n selected_states = flat_states[begin:end]\n if backward:\n sequence_input = (\n sequence_vectors[begin:end].detach().requires_grad_(True)\n )\n else:\n sequence_input = sequence_vectors[begin:end]\n liked_vectors, disliked_vectors, _ = pool_explicit_states(\n model, prepared, selected_states, device\n )\n counts = prepared.state_counts[selected_states]\n numeric = np.zeros((end - begin, 6), dtype=np.float32)\n numeric[:, :4] = flat_activity[begin:end]\n numeric[:, 4] = np.log1p(counts[:, 0]) / 6.0\n numeric[:, 5] = np.log1p(counts[:, 1]) / 4.0\n query_vectors = user_queries(\n model,\n sequence_input,\n liked_vectors,\n disliked_vectors,\n torch.from_numpy(numeric).to(device),\n )\n positive_vectors, negative_vectors = candidate_vectors(\n model,\n positive_items[begin:end],\n negatives[begin:end],\n )\n positive_logits = (query_vectors * positive_vectors).sum(dim=1)\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n if config.objective == \"infonce\":\n cat_logits = torch.cat(\n [positive_logits.unsqueeze(1), negative_logits], dim=1\n ) / config.temperature\n labels = torch.zeros(\n len(positive_logits), dtype=torch.long, device=device\n )\n raw_loss = F.cross_entropy(cat_logits, labels, reduction=\"none\")\n loss_sum = (selected_weights * raw_loss).sum()\n else:\n positive_loss = F.softplus(-positive_logits)\n negative_loss = corrected_negative_loss(\n negative_logits, negative_weights[begin:end], config\n )\n loss_sum = (\n selected_weights * (positive_loss + negative_loss) / 2.0\n ).sum()\n if not bool(torch.isfinite(loss_sum)):\n raise RuntimeError(\"dense next-item training produced non-finite loss\")\n if backward:\n (loss_sum / total_weight).backward()\n if sequence_input.grad is None:\n raise RuntimeError(\"sequence microbatch did not produce a gradient\")\n sequence_gradient[begin:end] = sequence_input.grad\n statistics[\"loss_sum\"] += float(loss_sum.detach().cpu())\n statistics[\"positive_sum\"] += float(positive_logits.detach().sum().cpu())\n statistics[\"negative_sum\"] += float(negative_logits.detach().sum().cpu())\n statistics[\"pair_correct\"] += float(\n (positive_logits.detach()[:, None] > negative_logits.detach()).sum().cpu()\n )\n statistics[\"weight_sum\"] += float(selected_weights.sum().detach().cpu())\n return statistics, sequence_gradient\n\n\ndef release_accelerator_cache(device: torch.device) -> None:\n if device.type == \"mps\":\n torch.mps.empty_cache()\n\n\ndef run_training_epoch(\n model: base.SequentialTwoTower,\n optimizer: torch.optim.Optimizer,\n prepared: SequenceData,\n config: Config,\n target_weights: np.ndarray,\n window_starts: np.ndarray,\n target_counts: np.ndarray,\n item_correction: torch.Tensor,\n known_positive_packed: np.ndarray | None,\n generator: torch.Generator,\n device: torch.device,\n epoch: int,\n) -> dict[str, float]:\n model.train()\n train_users = np.flatnonzero(target_counts > 0)\n sorted_users = train_users[\n np.argsort(target_counts[train_users], kind=\"stable\")\n ]\n groups = [\n sorted_users[start : start + config.user_batch_size]\n for start in range(0, len(sorted_users), config.user_batch_size)\n ]\n group_order = torch.randperm(len(groups), generator=generator).numpy()\n negative_count = config.uniform_negatives + config.inbatch_negatives\n total_loss = 0.0\n total_targets = 0\n total_weight = 0.0\n total_positive = 0.0\n total_negative = 0.0\n total_pair_correct = 0.0\n batches = 0\n started = time.monotonic()\n\n for group_number in group_order:\n users = groups[int(group_number)]\n group_counts = target_counts[users]\n hidden = torch.zeros(\n 1, len(users), config.embedding_dim, device=device\n )\n max_targets = int(group_counts.max())\n for start in range(0, max_targets, config.bptt_steps):\n steps = min(config.bptt_steps, max_targets - start)\n block = build_dense_block(\n prepared,\n target_weights,\n users,\n window_starts[users],\n group_counts,\n np.zeros(len(users), dtype=np.int64),\n start,\n steps,\n config,\n )\n sequence_outputs, _ = event_sequence(\n model,\n torch.from_numpy(block[0].astype(np.int64)).to(device),\n torch.from_numpy(block[1].astype(np.float32)).to(device),\n hidden,\n )\n hidden = torch.stack(\n [\n sequence_outputs[row, count - 1]\n if count\n else hidden[0, row]\n for row, count in enumerate(block[7])\n ]\n ).unsqueeze(0).detach()\n flat_scored = block[5].reshape(-1)\n if not np.any(flat_scored):\n continue\n mask = torch.from_numpy(flat_scored).to(device)\n sequence_vectors = sequence_outputs.reshape(\n -1, config.embedding_dim\n )[mask]\n flat_states = block[3].reshape(-1)[flat_scored]\n positive_items = torch.from_numpy(\n block[4].reshape(-1)[flat_scored].astype(np.int64)\n ).to(device)\n known_history = torch.from_numpy(\n block[6].reshape(-1, config.bptt_steps)[flat_scored].astype(\n np.int64\n )\n ).to(device)\n liked_items = torch.from_numpy(\n prepared.liked_states[flat_states].astype(np.int64)\n ).to(device)\n flat_user_rows = np.broadcast_to(\n users[:, None], block[5].shape\n ).reshape(-1)[flat_scored]\n negatives, negative_weights = draw_training_negatives(\n positive_items,\n known_history,\n liked_items,\n flat_user_rows,\n known_positive_packed,\n prepared,\n config,\n item_correction,\n generator,\n device,\n )\n optimizer.zero_grad(set_to_none=True)\n statistics, sequence_gradient = score_target_microbatches(\n model,\n prepared,\n config,\n sequence_vectors,\n flat_states,\n block[2].reshape(-1, 4)[flat_scored].astype(np.float32),\n positive_items,\n negatives,\n negative_weights,\n block[8].reshape(-1)[flat_scored],\n device,\n backward=True,\n )\n if sequence_gradient is None:\n raise RuntimeError(\"training block omitted its sequence gradient\")\n sequence_vectors.backward(sequence_gradient)\n nn.utils.clip_grad_norm_(model.parameters(), config.gradient_clip)\n optimizer.step()\n\n count = len(positive_items)\n total_loss += statistics[\"loss_sum\"]\n total_targets += count\n total_weight += statistics[\"weight_sum\"]\n total_positive += statistics[\"positive_sum\"]\n total_negative += statistics[\"negative_sum\"]\n total_pair_correct += statistics[\"pair_correct\"]\n batches += 1\n if batches % 25 == 0:\n release_accelerator_cache(device)\n if batches % 250 == 0:\n batch_bce = statistics[\"loss_sum\"] / max(\n statistics[\"weight_sum\"], 1.0e-12\n )\n running_bce = total_loss / max(total_weight, 1.0e-12)\n log(\n f\"epoch={epoch}/{config.epochs} batches={batches:,} \"\n f\"targets={total_targets:,} \"\n f\"batch_bce={batch_bce:.6f} \"\n f\"running_bce={running_bce:.6f} \"\n f\"elapsed={time.monotonic() - started:.1f}s\"\n )\n\n return {\n \"epoch\": float(epoch),\n \"loss\": total_loss / max(total_weight, 1.0e-12),\n \"positive_logit\": total_positive / max(total_targets, 1),\n \"negative_logit\": total_negative / max(total_targets * negative_count, 1),\n \"pair_accuracy\": total_pair_correct\n / max(total_targets * negative_count, 1),\n \"targets\": float(total_targets),\n \"target_weight_sum\": total_weight,\n \"batches\": float(batches),\n \"seconds\": time.monotonic() - started,\n }\n\n\n@torch.no_grad()\ndef evaluate_validation_bce(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n target_weights: np.ndarray,\n window_starts: np.ndarray,\n training_counts: np.ndarray,\n validation_counts: np.ndarray,\n item_correction: torch.Tensor,\n known_positive_packed: np.ndarray | None,\n device: torch.device,\n) -> dict[str, float]:\n \"\"\"Measure deterministic BCE on each user's chronological target suffix.\"\"\"\n model.eval()\n totals = training_counts + validation_counts\n users_with_validation = np.flatnonzero(validation_counts > 0)\n ordered = users_with_validation[\n np.argsort(totals[users_with_validation], kind=\"stable\")\n ]\n groups = [\n ordered[start : start + config.user_batch_size]\n for start in range(0, len(ordered), config.user_batch_size)\n ]\n generator = torch.Generator(device=\"cpu\").manual_seed(config.seed + 20_000)\n total_loss = 0.0\n total_targets = 0\n total_weight = 0.0\n started = time.monotonic()\n\n for group_number, users in enumerate(groups, start=1):\n group_totals = totals[users]\n group_starts = training_counts[users]\n hidden = torch.zeros(\n 1, len(users), config.embedding_dim, device=device\n )\n max_targets = int(group_totals.max())\n for start in range(0, max_targets, config.bptt_steps):\n steps = min(config.bptt_steps, max_targets - start)\n block = build_dense_block(\n prepared,\n target_weights,\n users,\n window_starts[users],\n group_totals,\n group_starts,\n start,\n steps,\n config,\n )\n sequence_outputs, _ = event_sequence(\n model,\n torch.from_numpy(block[0].astype(np.int64)).to(device),\n torch.from_numpy(block[1].astype(np.float32)).to(device),\n hidden,\n )\n hidden = torch.stack(\n [\n sequence_outputs[row, count - 1]\n if count\n else hidden[0, row]\n for row, count in enumerate(block[7])\n ]\n ).unsqueeze(0)\n flat_scored = block[5].reshape(-1)\n if not np.any(flat_scored):\n continue\n mask = torch.from_numpy(flat_scored).to(device)\n sequence_vectors = sequence_outputs.reshape(\n -1, config.embedding_dim\n )[mask]\n flat_states = block[3].reshape(-1)[flat_scored]\n positive_items = torch.from_numpy(\n block[4].reshape(-1)[flat_scored].astype(np.int64)\n ).to(device)\n known_history = torch.from_numpy(\n block[6].reshape(-1, config.bptt_steps)[flat_scored].astype(\n np.int64\n )\n ).to(device)\n liked_items = torch.from_numpy(\n prepared.liked_states[flat_states].astype(np.int64)\n ).to(device)\n flat_user_rows = np.broadcast_to(\n users[:, None], block[5].shape\n ).reshape(-1)[flat_scored]\n negatives, negative_weights = draw_training_negatives(\n positive_items,\n known_history,\n liked_items,\n flat_user_rows,\n known_positive_packed,\n prepared,\n config,\n item_correction,\n generator,\n device,\n )\n statistics, _ = score_target_microbatches(\n model,\n prepared,\n config,\n sequence_vectors,\n flat_states,\n block[2].reshape(-1, 4)[flat_scored].astype(np.float32),\n positive_items,\n negatives,\n negative_weights,\n block[8].reshape(-1)[flat_scored],\n device,\n backward=False,\n )\n total_loss += statistics[\"loss_sum\"]\n total_targets += len(positive_items)\n total_weight += statistics[\"weight_sum\"]\n if group_number % 25 == 0:\n release_accelerator_cache(device)\n return {\n \"validation_loss\": total_loss / max(total_weight, 1.0e-12),\n \"validation_targets\": float(total_targets),\n \"validation_target_weight_sum\": total_weight,\n \"validation_seconds\": time.monotonic() - started,\n }\n\n\ndef snapshot_model(model: nn.Module) -> dict[str, torch.Tensor]:\n return {\n name: value.detach().cpu().clone()\n for name, value in model.state_dict().items()\n }\n\n\ndef save_epoch_checkpoint(\n checkpoint_dir: Path,\n epoch: int,\n model_state: dict[str, torch.Tensor],\n config: Config,\n diagnostics: dict[str, float],\n optimizer: torch.optim.Optimizer,\n generator: torch.Generator,\n) -> Path:\n checkpoint_dir.mkdir(parents=True, exist_ok=True)\n path = checkpoint_dir / f\"epoch-{epoch:02d}.pt\"\n torch.save(\n {\n \"epoch\": epoch,\n \"config\": asdict(config),\n \"diagnostics\": diagnostics,\n \"model_state\": model_state,\n \"optimizer_state\": optimizer.state_dict(),\n \"generator_state\": generator.get_state(),\n },\n path,\n )\n log(f\"saved epoch checkpoint: {path}\")\n return path\n\n\ndef train_model(\n public: Path,\n stage: str,\n prepared: SequenceData,\n config: Config,\n device: torch.device,\n checkpoint_dir: Path | None,\n resume_checkpoint: Path | None,\n) -> tuple[base.SequentialTwoTower, list[dict[str, float]]]:\n partition = chronological_target_partition(\n prepared,\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n training_counts = partition.training_counts\n validation_counts = partition.validation_counts\n log(\n f\"chronological BCE window: train={int(training_counts.sum()):,} \"\n f\"validation={int(validation_counts.sum()):,} \"\n f\"selected={partition.selected_count:,} \"\n f\"discarded_old_targets={int(partition.window_starts.sum()):,} \"\n f\"all={prepared.target_count:,}\"\n )\n model = base.SequentialTwoTower(prepared.item_features, model_config(config)).to(\n device\n )\n optimizer = torch.optim.AdamW(\n model.parameters(), lr=config.learning_rate, weight_decay=config.weight_decay\n )\n generator = torch.Generator(device=\"cpu\").manual_seed(config.seed + 10_000)\n item_correction = build_item_correction(\n prepared,\n config,\n partition.window_starts,\n training_counts,\n )\n target_weights = build_target_weights(\n public,\n stage,\n prepared,\n partition.history_training_counts,\n config,\n )\n known_positive_packed = (\n build_known_positive_filter(prepared, partition.history_training_counts)\n if config.negative_filter_scope == \"full-training-history\"\n else None\n )\n history: list[dict[str, float]] = []\n best_loss = float(\"inf\")\n best_epoch = 0\n best_state: dict[str, torch.Tensor] | None = None\n epochs_without_improvement = 0\n start_epoch = 0\n\n if resume_checkpoint is not None:\n payload = torch.load(\n resume_checkpoint, map_location=device, weights_only=False\n )\n stored_config = payload.get(\"config\", {})\n architecture_fields = (\n \"bptt_steps\",\n \"embedding_dim\",\n \"hidden_dim\",\n \"user_hidden_1\",\n \"user_hidden_2\",\n \"item_hidden_1\",\n \"item_hidden_2\",\n )\n mismatches = {\n name: (stored_config.get(name), getattr(config, name))\n for name in architecture_fields\n if stored_config.get(name) != getattr(config, name)\n }\n if mismatches:\n raise RuntimeError(\n f\"resume checkpoint architecture differs: {mismatches}\"\n )\n stored_filter_scope = stored_config.get(\n \"negative_filter_scope\", \"last-100\"\n )\n stored_target_weighting = stored_config.get(\n \"target_weighting\", \"uniform\"\n )\n stored_weight_values = (\n stored_config.get(\"ordinary_target_weight\", 1.0),\n stored_config.get(\"organic_strong_target_weight\", 1.0),\n stored_config.get(\"organic_like_target_weight\", 3.0),\n stored_config.get(\"discovery_horizon_days\", 7),\n )\n current_weight_values = (\n config.ordinary_target_weight,\n config.organic_strong_target_weight,\n config.organic_like_target_weight,\n config.discovery_horizon_days,\n )\n stored_window_values = (\n stored_config.get(\"validation_fraction\", 0.10),\n stored_config.get(\"max_train_targets_per_user\", 0),\n stored_config.get(\"max_validation_targets_per_user\", 0),\n )\n current_window_values = (\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n objective_changed = (\n stored_filter_scope != config.negative_filter_scope\n or stored_target_weighting != config.target_weighting\n or stored_weight_values != current_weight_values\n or stored_window_values != current_window_values\n )\n model.load_state_dict(payload[\"model_state\"])\n if \"optimizer_state\" in payload and not objective_changed:\n optimizer.load_state_dict(payload[\"optimizer_state\"])\n for group in optimizer.param_groups:\n group[\"lr\"] = config.learning_rate\n group[\"weight_decay\"] = config.weight_decay\n log(f\"restored optimizer state from {resume_checkpoint}\")\n else:\n log(\n f\"using a fresh AdamW optimizer after loading {resume_checkpoint}\"\n )\n if \"generator_state\" in payload and not objective_changed:\n generator.set_state(payload[\"generator_state\"])\n start_epoch = int(payload[\"epoch\"])\n if start_epoch >= config.epochs:\n raise RuntimeError(\n f\"resume epoch {start_epoch} must be below --epochs {config.epochs}\"\n )\n prior = {\n str(name): float(value)\n for name, value in payload.get(\"diagnostics\", {}).items()\n }\n if objective_changed:\n prior.pop(\"validation_loss\", None)\n log(\n \"BCE objective changed: filter \"\n f\"{stored_filter_scope}->{config.negative_filter_scope}, \"\n \"target weighting \"\n f\"{stored_target_weighting}->{config.target_weighting}, \"\n f\"target window {stored_window_values}->{current_window_values}; \"\n \"recomputing the resume checkpoint's validation BCE\"\n )\n if \"validation_loss\" not in prior:\n prior.update(\n evaluate_validation_bce(\n model,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n validation_counts,\n item_correction,\n known_positive_packed,\n device,\n )\n )\n history.append(prior)\n best_loss = float(prior[\"validation_loss\"])\n best_epoch = start_epoch\n best_state = snapshot_model(model)\n log(\n f\"resumed validation-BCE training: epoch={start_epoch} \"\n f\"loss={best_loss:.6f}\"\n )\n\n for epoch in range(start_epoch + 1, config.epochs + 1):\n diagnostics = run_training_epoch(\n model,\n optimizer,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n item_correction,\n known_positive_packed,\n generator,\n device,\n epoch,\n )\n validation = evaluate_validation_bce(\n model,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n validation_counts,\n item_correction,\n known_positive_packed,\n device,\n )\n diagnostics.update(validation)\n history.append(diagnostics)\n current_state = snapshot_model(model)\n if checkpoint_dir is not None:\n save_epoch_checkpoint(\n checkpoint_dir,\n epoch,\n current_state,\n config,\n diagnostics,\n optimizer,\n generator,\n )\n improved = (\n diagnostics[\"validation_loss\"]\n < best_loss - config.early_stopping_min_delta\n )\n if improved:\n best_loss = diagnostics[\"validation_loss\"]\n best_epoch = epoch\n best_state = current_state\n epochs_without_improvement = 0\n if checkpoint_dir is not None:\n torch.save(\n {\n \"epoch\": epoch,\n \"config\": asdict(config),\n \"diagnostics\": diagnostics,\n \"model_state\": current_state,\n \"optimizer_state\": optimizer.state_dict(),\n \"generator_state\": generator.get_state(),\n },\n checkpoint_dir / \"best.pt\",\n )\n else:\n epochs_without_improvement += 1\n log(\n f\"epoch={epoch}/{config.epochs} train_loss={diagnostics['loss']:.5f} \"\n f\"validation_loss={diagnostics['validation_loss']:.5f} \"\n f\"pair_accuracy={diagnostics['pair_accuracy']:.3f} \"\n f\"train_targets={int(diagnostics['targets']):,} \"\n f\"validation_targets={int(diagnostics['validation_targets']):,} \"\n f\"train_seconds={diagnostics['seconds']:.1f} \"\n f\"validation_seconds={diagnostics['validation_seconds']:.1f}\"\n )\n release_accelerator_cache(device)\n if epochs_without_improvement >= config.early_stopping_patience:\n log(\n f\"early stopping after epoch={epoch}; \"\n f\"best_epoch={best_epoch} best_validation_loss={best_loss:.6f}\"\n )\n break\n\n if best_state is None:\n raise RuntimeError(\"training did not produce a valid checkpoint\")\n model.load_state_dict(best_state)\n model.to(device).eval()\n log(\n f\"restored best validation-BCE checkpoint: epoch={best_epoch} \"\n f\"loss={best_loss:.6f}\"\n )\n return model, history\n\n\n@torch.inference_mode()\ndef inference_queries(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n device: torch.device,\n) -> torch.Tensor:\n sequence_rows = prepared.inference_user_rows\n lengths = np.where(\n sequence_rows >= 0, prepared.lengths[np.maximum(sequence_rows, 0)], 0\n )\n ordered = np.argsort(lengths, kind=\"stable\")\n result = torch.zeros(\n len(sequence_rows), config.embedding_dim, dtype=torch.float32\n )\n for group_start in range(0, len(ordered), config.user_batch_size):\n output_rows = ordered[group_start : group_start + config.user_batch_size]\n rows = sequence_rows[output_rows]\n group_lengths = lengths[output_rows]\n hidden = torch.zeros(\n 1, len(output_rows), config.embedding_dim, device=device\n )\n for start in range(0, int(group_lengths.max(initial=0)), config.bptt_steps):\n steps = min(\n config.bptt_steps,\n int(group_lengths.max(initial=0)) - start,\n )\n items = np.full(\n (len(output_rows), steps),\n len(prepared.item_ids),\n dtype=np.uint16,\n )\n features = np.zeros((len(output_rows), steps, 3), dtype=np.float16)\n counts = np.zeros(len(output_rows), dtype=np.int32)\n for local_row, sequence_row in enumerate(rows):\n if sequence_row < 0:\n continue\n count = min(steps, max(int(group_lengths[local_row]) - start, 0))\n if not count:\n continue\n offset = int(prepared.offsets[sequence_row])\n selected = slice(offset + start, offset + start + count)\n items[local_row, :count] = prepared.sequence_items[selected]\n features[local_row, :count] = prepared.event_features[selected]\n counts[local_row] = count\n outputs, _ = event_sequence(\n model,\n torch.from_numpy(items.astype(np.int64)).to(device),\n torch.from_numpy(features.astype(np.float32)).to(device),\n hidden,\n )\n next_rows: list[torch.Tensor] = []\n for local_row, count in enumerate(counts):\n next_rows.append(\n outputs[local_row, count - 1]\n if count\n else hidden[0, local_row]\n )\n hidden = torch.stack(next_rows).unsqueeze(0)\n\n liked_items = torch.from_numpy(\n prepared.inference_liked_items[output_rows].astype(np.int64)\n ).to(device)\n disliked_items = torch.from_numpy(\n prepared.inference_disliked_items[output_rows].astype(np.int64)\n ).to(device)\n queries = user_queries(\n model,\n hidden.squeeze(0),\n model._mean_context_channel(liked_items),\n model._mean_context_channel(disliked_items),\n torch.from_numpy(\n prepared.inference_user_features[output_rows].astype(np.float32)\n ).to(device),\n )\n result[torch.from_numpy(output_rows.astype(np.int64))] = queries.cpu()\n return result\n\n\n@torch.inference_mode()\ndef rank_candidates(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n eligibility: CandidateSet,\n config: Config,\n device: torch.device,\n ranking_batch_size: int,\n) -> tuple[pd.DataFrame, dict[str, float]]:\n model.eval()\n queries = inference_queries(model, prepared, config, device)\n item_parts: list[torch.Tensor] = []\n for start in range(0, len(prepared.item_ids), 2_048):\n rows = torch.arange(\n start,\n min(start + 2_048, len(prepared.item_ids)),\n dtype=torch.long,\n device=device,\n )\n item_parts.append(model.encode_items(rows).cpu())\n items = torch.cat(item_parts)\n rows: list[tuple[int, int, int]] = []\n for start in range(0, len(eligibility.uids), ranking_batch_size):\n end = min(start + ranking_batch_size, len(eligibility.uids))\n scores = queries[start:end] @ items.T\n eligible = np.unpackbits(\n eligibility.packed[start:end],\n axis=1,\n count=len(prepared.item_ids),\n bitorder=\"little\",\n ).astype(bool, copy=False)\n scores.masked_fill_(~torch.from_numpy(eligible), -torch.inf)\n selected = torch.topk(scores, k=base.TOP_N, dim=1).indices.numpy()\n for local_row, columns in enumerate(selected):\n uid = int(eligibility.uids[start + local_row])\n rows.extend(\n (uid, int(prepared.item_ids[column]), rank)\n for rank, column in enumerate(columns, start=1)\n )\n if end == len(eligibility.uids) or end % 1_024 == 0:\n log(f\"ranked users={end:,}/{len(eligibility.uids):,}\")\n query_norm = torch.linalg.vector_norm(queries, dim=1).numpy()\n item_norm = torch.linalg.vector_norm(items, dim=1).numpy()\n diagnostics = {\n \"query_norm_mean\": float(query_norm.mean()),\n \"query_norm_std\": float(query_norm.std()),\n \"item_norm_mean\": float(item_norm.mean()),\n \"item_norm_std\": float(item_norm.std()),\n }\n return pd.DataFrame(rows, columns=[\"uid\", \"item_id\", \"rank\"]), diagnostics\n\n\ndef write_ranking(frame: pd.DataFrame, output: Path) -> None:\n output.parent.mkdir(parents=True, exist_ok=True)\n if output.suffix == \".parquet\":\n frame.to_parquet(output, index=False, compression=\"zstd\")\n elif output.suffix == \".csv\":\n frame.to_csv(output, index=False)\n else:\n raise ValueError(\"--output must end in .parquet or .csv\")\n\n\ndef main() -> int:\n parser = argparse.ArgumentParser(description=__doc__)\n parser.add_argument(\n \"--stage\", choices=(\"validation\", \"submission\"), default=\"validation\"\n )\n parser.add_argument(\"--output\", type=Path, required=True)\n parser.add_argument(\"--public-dir\", type=Path, required=True)\n parser.add_argument(\"--prepared-cache\", type=Path)\n parser.add_argument(\"--device\", choices=(\"cpu\", \"mps\"), default=\"cpu\")\n parser.add_argument(\"--threads\", type=int, default=8)\n parser.add_argument(\"--bptt-steps\", type=int, default=Config.bptt_steps)\n parser.add_argument(\n \"--user-batch-size\", type=int, default=Config.user_batch_size\n )\n parser.add_argument(\"--epochs\", type=int, default=Config.epochs)\n parser.add_argument(\"--learning-rate\", type=float, default=Config.learning_rate)\n parser.add_argument(\n \"--loss-microbatch-size\",\n type=int,\n default=Config.loss_microbatch_size,\n )\n parser.add_argument(\n \"--negative-filter-scope\",\n choices=(\"full-training-history\", \"last-100\"),\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n \"--objective\",\n choices=(\"bce\", \"infonce\"),\n default=Config.objective,\n )\n parser.add_argument(\n \"--temperature\",\n type=float,\n default=Config.temperature,\n )\n parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n )\n parser.add_argument(\n \"--ordinary-target-weight\",\n type=float,\n default=Config.ordinary_target_weight,\n )\n parser.add_argument(\n \"--organic-strong-target-weight\",\n type=float,\n default=Config.organic_strong_target_weight,\n )\n parser.add_argument(\n \"--organic-like-target-weight\",\n type=float,\n default=Config.organic_like_target_weight,\n )\n parser.add_argument(\n \"--discovery-horizon-days\",\n type=int,\n default=Config.discovery_horizon_days,\n )\n parser.add_argument(\n \"--validation-fraction\",\n type=float,\n default=Config.validation_fraction,\n )\n parser.add_argument(\n \"--max-train-targets-per-user\",\n type=int,\n default=Config.max_train_targets_per_user,\n help=\"retain this many recent pre-validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--max-validation-targets-per-user\",\n type=int,\n default=Config.max_validation_targets_per_user,\n help=\"retain this many final validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--early-stopping-patience\",\n type=int,\n default=Config.early_stopping_patience,\n )\n parser.add_argument(\n \"--early-stopping-min-delta\",\n type=float,\n default=Config.early_stopping_min_delta,\n )\n parser.add_argument(\"--checkpoint-dir\", type=Path)\n parser.add_argument(\"--resume-checkpoint\", type=Path)\n parser.add_argument(\"--seed\", type=int, default=Config.seed)\n parser.add_argument(\"--ranking-batch-size\", type=int, default=64)\n parser.add_argument(\"--prepare-only\", action=\"store_true\")\n args = parser.parse_args()\n\n if args.threads < 1 or args.bptt_steps < 2 or args.user_batch_size < 1:\n parser.error(\"threads, BPTT steps, and user batch size must be positive\")\n if (\n args.epochs < 1\n or args.learning_rate <= 0\n or args.loss_microbatch_size < 1\n or args.early_stopping_patience < 1\n or args.early_stopping_min_delta < 0\n or args.ordinary_target_weight <= 0\n or args.organic_strong_target_weight <= 0\n or args.organic_like_target_weight <= 0\n or args.discovery_horizon_days < 1\n or args.max_train_targets_per_user < 0\n or args.max_validation_targets_per_user < 0\n ):\n parser.error(\"training counts, learning rate, and patience are invalid\")\n if not 0.0 < args.validation_fraction < 1.0:\n parser.error(\"validation fraction must be strictly between zero and one\")\n public = args.public_dir.resolve()\n if not (public / \"data\" / \"candidate_catalog.parquet\").is_file():\n parser.error(f\"public task data not found under {public}\")\n config = Config(\n bptt_steps=args.bptt_steps,\n user_batch_size=args.user_batch_size,\n epochs=args.epochs,\n learning_rate=args.learning_rate,\n loss_microbatch_size=args.loss_microbatch_size,\n negative_filter_scope=args.negative_filter_scope,\n target_weighting=args.target_weighting,\n ordinary_target_weight=args.ordinary_target_weight,\n organic_strong_target_weight=args.organic_strong_target_weight,\n organic_like_target_weight=args.organic_like_target_weight,\n discovery_horizon_days=args.discovery_horizon_days,\n validation_fraction=args.validation_fraction,\n max_train_targets_per_user=args.max_train_targets_per_user,\n max_validation_targets_per_user=args.max_validation_targets_per_user,\n early_stopping_patience=args.early_stopping_patience,\n early_stopping_min_delta=args.early_stopping_min_delta,\n seed=args.seed,\n )\n torch.set_num_threads(args.threads)\n torch.set_num_interop_threads(1)\n seed_everything(config.seed)\n device = torch.device(args.device)\n if device.type == \"mps\" and not torch.backends.mps.is_available():\n parser.error(\"MPS was requested but is unavailable\")\n\n started = time.monotonic()\n prepared, eligibility = load_or_prepare(\n public,\n args.stage,\n config,\n args.prepared_cache.resolve() if args.prepared_cache else None,\n )\n target_partition = chronological_target_partition(\n prepared,\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n if args.prepare_only:\n print(\n json.dumps(\n {\n \"stage\": args.stage,\n \"sequence_users\": len(prepared.user_ids),\n \"sequence_tokens\": len(prepared.sequence_items),\n \"next_item_targets\": prepared.target_count,\n \"selected_training_targets\": int(\n target_partition.training_counts.sum()\n ),\n \"selected_validation_targets\": int(\n target_partition.validation_counts.sum()\n ),\n \"discarded_old_targets\": int(\n target_partition.window_starts.sum()\n ),\n \"target_users\": len(eligibility.uids),\n \"runtime_seconds\": round(time.monotonic() - started, 3),\n \"config\": asdict(config),\n },\n indent=2,\n )\n )\n return 0\n\n model, training_history = train_model(\n public,\n args.stage,\n prepared,\n config,\n device,\n args.checkpoint_dir.resolve() if args.checkpoint_dir else None,\n args.resume_checkpoint.resolve() if args.resume_checkpoint else None,\n )\n ranking, embedding_diagnostics = rank_candidates(\n model,\n prepared,\n eligibility,\n config,\n device,\n args.ranking_batch_size,\n )\n write_ranking(ranking, args.output.resolve())\n print(\n json.dumps(\n {\n \"baseline\": \"yambda-windowed-sequence-two-tower-base\",\n \"stage\": args.stage,\n \"output\": str(args.output.resolve()),\n \"rows\": len(ranking),\n \"users\": int(ranking.uid.nunique()),\n \"runtime_seconds\": round(time.monotonic() - started, 3),\n \"public_inputs_only\": True,\n \"architecture\": {\n \"family\": \"pure-windowed-sequence-two-tower\",\n \"sequence_training\": (\n \"every-adjacent-target-within-recent-chronological-window\"\n ),\n \"sequence_encoder\": \"one-layer-GRU-dense-timestep-loss\",\n \"bptt_hidden_state\": \"carried-and-detached\",\n \"model_selection\": \"chronological-validation-BCE-early-stopping\",\n \"loss_execution\": \"gradient-accumulated-microbatches\",\n \"user_id_embedding\": False,\n \"user_mlp\": \"582-512-256-192\",\n \"item_mlp\": \"196-512-256-192\",\n \"normalization\": \"none\",\n \"score\": \"raw-dot-product\",\n \"objective\": \"sample-weighted-BCE-at-every-timestep\",\n \"negative_sampling\": (\n \"24-uniform-plus-8-propensity-corrected-inbatch\"\n ),\n \"false_negative_filter\": config.negative_filter_scope,\n \"target_weighting\": config.target_weighting,\n },\n \"config\": asdict(config),\n \"sequence_tokens\": len(prepared.sequence_items),\n \"all_next_item_targets\": prepared.target_count,\n \"selected_training_targets\": int(\n target_partition.training_counts.sum()\n ),\n \"selected_validation_targets\": int(\n target_partition.validation_counts.sum()\n ),\n \"discarded_old_targets\": int(target_partition.window_starts.sum()),\n \"training_history\": training_history,\n \"embedding_diagnostics\": embedding_diagnostics,\n \"torch_version\": torch.__version__,\n \"device\": str(device),\n },\n indent=2,\n )\n )\n return 0\n\n\nif __name__ == \"__main__\":\n raise SystemExit(main())\n",
"newContent": "#!/usr/bin/env python3\n\"\"\"Train the Yambda windowed two-tower baseline.\n\nEvery catalog listen with at least 50 percent completion is a sequence token.\nWithin each selected chronological window, every adjacent target contributes a\nloss. The fast default retains each user's most recent 128 training targets and\n16 held-out validation targets; passing zero for either cap restores the\nuncapped behavior for that side of the split. A 100-step truncated-BPTT block\ncomputes all of its timestep losses and carries the detached GRU hidden state\ninto the next block.\n\nThe model is a pure two-tower retriever: no user-ID embedding, no output\nnormalization, raw dot-product retrieval, sample-weighted binary cross entropy,\nand 24 uniform plus eight propensity-corrected unique in-batch negatives. Every\nselected adjacent target retains a nonzero weight. Only files under the supplied\npublic directory are read.\n\"\"\"\nfrom __future__ import annotations\n\nimport argparse\nimport json\nimport random\nimport sys\nimport time\nfrom dataclasses import asdict, dataclass\nfrom pathlib import Path\n\nimport numpy as np\nimport pandas as pd\nimport torch\nfrom torch import nn\nfrom torch.nn import functional as F\n\nimport two_tower_model as base\n\n\nCACHE_VERSION = 2\n\n\n@dataclass(frozen=True)\nclass Config:\n bptt_steps: int = 100\n user_batch_size: int = 64\n embedding_dim: int = 192\n hidden_dim: int = 192\n user_hidden_1: int = 512\n user_hidden_2: int = 256\n item_hidden_1: int = 512\n item_hidden_2: int = 256\n dropout: float = 0.05\n epochs: int = 1\n learning_rate: float = 1.0e-3\n weight_decay: float = 1.0e-5\n gradient_clip: float = 2.0\n loss_microbatch_size: int = 65_536\n uniform_negatives: int = 24\n inbatch_negatives: int = 8\n inbatch_correction_power: float = 0.5\n inbatch_correction_min: float = 0.25\n inbatch_correction_max: float = 4.0\n negative_filter_scope: str = \"full-training-history\"\n objective: str = \"bce\"\n temperature: float = 0.2\n target_weighting: str = \"organic-strong\"\n ordinary_target_weight: float = 0.25\n organic_strong_target_weight: float = 1.0\n organic_like_target_weight: float = 3.0\n discovery_horizon_days: int = 7\n validation_fraction: float = 0.10\n max_train_targets_per_user: int = 128\n max_validation_targets_per_user: int = 16\n early_stopping_patience: int = 1\n early_stopping_min_delta: float = 1.0e-4\n seed: int = 29\n\n\n@dataclass(frozen=True)\nclass CandidateSet:\n uids: np.ndarray\n item_ids: np.ndarray\n packed: np.ndarray\n\n @classmethod\n def read(cls, path: Path) -> \"CandidateSet\":\n with np.load(path, allow_pickle=False) as artifact:\n result = cls(\n uids=artifact[\"uids\"].astype(np.int64, copy=True),\n item_ids=artifact[\"item_ids\"].astype(np.int64, copy=True),\n packed=artifact[\"packed_eligible\"].astype(np.uint8, copy=True),\n )\n expected = (len(result.uids), (len(result.item_ids) + 7) // 8)\n if result.packed.shape != expected:\n raise ValueError(\n f\"invalid candidate artifact shape {result.packed.shape}; \"\n f\"expected {expected}\"\n )\n return result\n\n\n@dataclass(frozen=True)\nclass SequenceData:\n item_ids: np.ndarray\n item_features: np.ndarray\n user_ids: np.ndarray\n offsets: np.ndarray\n sequence_timestamps: np.ndarray\n sequence_items: np.ndarray\n event_features: np.ndarray\n activity_features: np.ndarray\n state_ids: np.ndarray\n liked_states: np.ndarray\n disliked_states: np.ndarray\n state_counts: np.ndarray\n inference_user_rows: np.ndarray\n inference_liked_items: np.ndarray\n inference_disliked_items: np.ndarray\n inference_user_features: np.ndarray\n\n @property\n def lengths(self) -> np.ndarray:\n return np.diff(self.offsets)\n\n @property\n def target_count(self) -> int:\n return int(np.maximum(self.lengths - 1, 0).sum())\n\n\n@dataclass(frozen=True)\nclass TargetPartition:\n \"\"\"A contiguous recent target window and its chronological BCE split.\"\"\"\n\n window_starts: np.ndarray\n training_counts: np.ndarray\n validation_counts: np.ndarray\n history_training_counts: np.ndarray\n\n @property\n def selected_count(self) -> int:\n return int(self.training_counts.sum() + self.validation_counts.sum())\n\n\ndef log(message: str) -> None:\n print(message, file=sys.stderr, flush=True)\n\n\ndef seed_everything(seed: int) -> None:\n random.seed(seed)\n np.random.seed(seed)\n torch.manual_seed(seed)\n\n\ndef model_config(config: Config) -> base.TowerConfig:\n \"\"\"Map the dense trainer configuration onto the unchanged tower.\"\"\"\n return base.TowerConfig(\n sequence_length=config.bptt_steps,\n embedding_dim=config.embedding_dim,\n hidden_dim=config.hidden_dim,\n user_hidden_1=config.user_hidden_1,\n user_hidden_2=config.user_hidden_2,\n item_hidden_1=config.item_hidden_1,\n item_hidden_2=config.item_hidden_2,\n dropout=config.dropout,\n epochs=config.epochs,\n learning_rate=config.learning_rate,\n weight_decay=config.weight_decay,\n gradient_clip=config.gradient_clip,\n uniform_negatives=config.uniform_negatives,\n inbatch_negatives=config.inbatch_negatives,\n inbatch_correction_power=config.inbatch_correction_power,\n inbatch_correction_min=config.inbatch_correction_min,\n inbatch_correction_max=config.inbatch_correction_max,\n seed=config.seed,\n )\n\n\ndef append_state_snapshot(\n active_likes: dict[int, int],\n active_dislikes: dict[int, int],\n n_items: int,\n liked_states: list[np.ndarray],\n disliked_states: list[np.ndarray],\n state_counts: list[tuple[int, int]],\n) -> int:\n liked_states.append(base.pack_active_items(active_likes, base.LIKE_ITEMS, n_items))\n disliked_states.append(\n base.pack_active_items(active_dislikes, base.DISLIKE_ITEMS, n_items)\n )\n state_counts.append((len(active_likes), len(active_dislikes)))\n return len(liked_states) - 1\n\n\ndef causal_activity_features(\n timestamps: np.ndarray,\n items: np.ndarray,\n organic: np.ndarray,\n completion: np.ndarray,\n) -> np.ndarray:\n \"\"\"Compute the first four reviewed user features at every event cutoff.\"\"\"\n length = len(items)\n result = np.zeros((length, 4), dtype=np.float16)\n if not length:\n return result\n left_rows = np.searchsorted(\n timestamps,\n timestamps - base.PROFILE_LOOKBACK_DAYS * base.DAY,\n side=\"left\",\n )\n rows = np.arange(length, dtype=np.int64)\n counts = rows - left_rows + 1\n organic_sum = np.cumsum(organic.astype(np.float64))\n completion_sum = np.cumsum(np.minimum(completion, 100).astype(np.float64))\n before_organic = np.where(\n left_rows > 0, organic_sum[np.maximum(left_rows - 1, 0)], 0.0\n )\n before_completion = np.where(\n left_rows > 0, completion_sum[np.maximum(left_rows - 1, 0)], 0.0\n )\n\n unique_counts = np.empty(length, dtype=np.int32)\n frequencies: dict[int, int] = {}\n left = 0\n for right, raw_item in enumerate(items):\n item = int(raw_item)\n frequencies[item] = frequencies.get(item, 0) + 1\n required_left = int(left_rows[right])\n while left < required_left:\n old_item = int(items[left])\n remaining = frequencies[old_item] - 1\n if remaining:\n frequencies[old_item] = remaining\n else:\n del frequencies[old_item]\n left += 1\n unique_counts[right] = len(frequencies)\n\n result[:, 0] = (np.log1p(counts) / 8.0).astype(np.float16)\n result[:, 1] = (np.log1p(unique_counts) / 6.0).astype(np.float16)\n result[:, 2] = (\n (organic_sum - before_organic) / np.maximum(counts, 1)\n ).astype(np.float16)\n result[:, 3] = (\n (completion_sum - before_completion)\n / np.maximum(counts, 1)\n / 100.0\n ).astype(np.float16)\n return result\n\n\ndef build_sequence_parts(\n listens: base.ListenEvents,\n explicit: base.ExplicitEvents,\n n_items: int,\n liked_states: list[np.ndarray],\n disliked_states: list[np.ndarray],\n state_counts: list[tuple[int, int]],\n) -> tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray, np.ndarray] | None:\n qualifying = listens.completion_pct >= base.LISTEN_POSITIVE_THRESHOLD\n timestamps = listens.timestamps[qualifying].astype(np.int64, copy=False)\n items = listens.items[qualifying].astype(np.uint16, copy=False)\n organic = listens.organic[qualifying].astype(np.float32, copy=False)\n completion = listens.completion_pct[qualifying].astype(np.float32, copy=False)\n if not len(items):\n return None\n\n event_features = np.zeros((len(items), 3), dtype=np.float16)\n event_features[:, 0] = organic.astype(np.float16)\n event_features[:, 1] = (\n np.minimum(completion, 100.0) / 100.0\n ).astype(np.float16)\n if len(items) > 1:\n gaps = np.maximum(0, timestamps[1:] - timestamps[:-1])\n event_features[1:, 2] = np.exp(\n -gaps.astype(np.float64) / (60.0 * base.DAY)\n ).astype(np.float16)\n\n activity = causal_activity_features(\n timestamps, items, organic, completion\n )\n states = np.zeros(len(items), dtype=np.uint32)\n active_likes: dict[int, int] = {}\n active_dislikes: dict[int, int] = {}\n explicit_position = 0\n current_state = 0\n for row, timestamp in enumerate(timestamps):\n changed = False\n while (\n explicit_position < len(explicit.timestamps)\n and int(explicit.timestamps[explicit_position]) < int(timestamp)\n ):\n item = int(explicit.items[explicit_position])\n event_timestamp = int(explicit.timestamps[explicit_position])\n destination = (\n active_likes\n if int(explicit.channel[explicit_position]) == 0\n else active_dislikes\n )\n if int(explicit.active[explicit_position]) == 1:\n destination[item] = event_timestamp\n else:\n destination.pop(item, None)\n explicit_position += 1\n changed = True\n if changed:\n current_state = append_state_snapshot(\n active_likes,\n active_dislikes,\n n_items,\n liked_states,\n disliked_states,\n state_counts,\n )\n states[row] = np.uint32(current_state)\n return timestamps, items.copy(), event_features, activity, states\n\n\ndef final_user_features(\n timestamps: np.ndarray,\n items: np.ndarray,\n event_features: np.ndarray,\n history_end: int,\n like_count: int,\n dislike_count: int,\n) -> np.ndarray:\n start = int(\n np.searchsorted(\n timestamps,\n history_end - base.PROFILE_LOOKBACK_DAYS * base.DAY,\n side=\"left\",\n )\n )\n result = np.zeros(6, dtype=np.float32)\n recent_items = items[start:]\n if len(recent_items):\n result[0] = np.log1p(len(recent_items)) / 8.0\n result[1] = np.log1p(len(np.unique(recent_items))) / 6.0\n result[2] = float(np.mean(event_features[start:, 0]))\n result[3] = float(np.mean(event_features[start:, 1]))\n result[4] = np.log1p(like_count) / 6.0\n result[5] = np.log1p(dislike_count) / 4.0\n return result\n\n\ndef prepare_data(\n public: Path, stage: str, config: Config\n) -> tuple[SequenceData, CandidateSet]:\n data = public / \"data\"\n catalog = pd.read_parquet(data / \"candidate_catalog.parquet\").sort_values(\n \"popularity_rank\"\n )\n item_ids = catalog.item_id.to_numpy(dtype=np.int64, copy=True)\n item_index = pd.Index(item_ids)\n n_items = len(item_ids)\n if n_items >= np.iinfo(np.uint16).max:\n raise RuntimeError(\"uint16 item storage is too small for this catalog\")\n\n if stage == \"validation\":\n prefixes = (\"train\",)\n history_end = base.TRAIN_END\n candidate_name = \"validation_eligible_candidates.npz\"\n state_name = \"validation_preference_state.parquet\"\n else:\n prefixes = (\"train\", \"val\")\n history_end = base.VALIDATION_END\n candidate_name = \"test_eligible_candidates.npz\"\n state_name = \"test_preference_state.parquet\"\n eligibility = CandidateSet.read(data / candidate_name)\n if not np.array_equal(eligibility.item_ids, item_ids):\n raise RuntimeError(\"catalog and candidate artifact item order differ\")\n\n explicit_by_user = base.load_explicit_events(data, prefixes, item_index)\n inference_likes, inference_dislikes, like_counts, dislike_counts = (\n base.load_inference_state(\n data / state_name,\n item_index,\n eligibility.uids,\n n_items,\n )\n )\n inference_rows = {int(uid): row for row, uid in enumerate(eligibility.uids)}\n inference_user_features = np.zeros((len(eligibility.uids), 6), dtype=np.float32)\n inference_sequence_rows = np.full(len(eligibility.uids), -1, dtype=np.int32)\n\n # Snapshot zero is the shared empty explicit state.\n liked_states = [np.full(base.LIKE_ITEMS, n_items, dtype=np.uint16)]\n disliked_states = [\n np.full(base.DISLIKE_ITEMS, n_items, dtype=np.uint16)\n ]\n state_counts: list[tuple[int, int]] = [(0, 0)]\n user_ids: list[int] = []\n offsets = [0]\n timestamp_parts: list[np.ndarray] = []\n item_parts: list[np.ndarray] = []\n event_parts: list[np.ndarray] = []\n activity_parts: list[np.ndarray] = []\n state_parts: list[np.ndarray] = []\n\n listen_paths = [data / f\"{prefix}_listens.parquet\" for prefix in prefixes]\n started = time.monotonic()\n for seen_users, (uid, listens) in enumerate(\n base.merge_listen_streams(listen_paths, item_index), start=1\n ):\n parts = build_sequence_parts(\n listens,\n explicit_by_user.get(uid, base.empty_explicit_events()),\n n_items,\n liked_states,\n disliked_states,\n state_counts,\n )\n if parts is None:\n continue\n sequence_row = len(user_ids)\n user_ids.append(uid)\n timestamp_parts.append(parts[0].astype(np.uint32, copy=False))\n item_parts.append(parts[1])\n event_parts.append(parts[2])\n activity_parts.append(parts[3])\n state_parts.append(parts[4])\n offsets.append(offsets[-1] + len(parts[1]))\n inference_row = inference_rows.get(uid)\n if inference_row is not None:\n inference_sequence_rows[inference_row] = sequence_row\n inference_user_features[inference_row] = final_user_features(\n parts[0],\n parts[1],\n parts[2],\n history_end,\n int(like_counts[inference_row]),\n int(dislike_counts[inference_row]),\n )\n if seen_users % 1_000 == 0:\n log(\n f\"prepared users={seen_users:,} tokens={offsets[-1]:,} \"\n f\"states={len(liked_states):,} \"\n f\"elapsed={time.monotonic() - started:.1f}s\"\n )\n\n missing = np.flatnonzero(inference_sequence_rows < 0)\n if len(missing):\n inference_user_features[missing, 4] = np.log1p(like_counts[missing]) / 6.0\n inference_user_features[missing, 5] = (\n np.log1p(dislike_counts[missing]) / 4.0\n )\n log(f\"target users without qualifying catalog sequence={len(missing):,}\")\n\n prepared = SequenceData(\n item_ids=item_ids,\n item_features=base.build_item_features(catalog),\n user_ids=np.asarray(user_ids, dtype=np.int64),\n offsets=np.asarray(offsets, dtype=np.int64),\n sequence_timestamps=np.concatenate(timestamp_parts),\n sequence_items=np.concatenate(item_parts),\n event_features=np.concatenate(event_parts),\n activity_features=np.concatenate(activity_parts),\n state_ids=np.concatenate(state_parts),\n liked_states=np.stack(liked_states),\n disliked_states=np.stack(disliked_states),\n state_counts=np.asarray(state_counts, dtype=np.int32),\n inference_user_rows=inference_sequence_rows,\n inference_liked_items=inference_likes,\n inference_disliked_items=inference_dislikes,\n inference_user_features=inference_user_features,\n )\n log(\n f\"prepared stage={stage} sequence_users={len(prepared.user_ids):,} \"\n f\"tokens={len(prepared.sequence_items):,} \"\n f\"next_item_targets={prepared.target_count:,} \"\n f\"states={len(prepared.liked_states):,}\"\n )\n return prepared, eligibility\n\n\ndef cache_metadata(stage: str, config: Config) -> dict[str, object]:\n return {\n \"cache_version\": CACHE_VERSION,\n \"stage\": stage,\n \"positive_listen_threshold\": base.LISTEN_POSITIVE_THRESHOLD,\n \"profile_lookback_days\": base.PROFILE_LOOKBACK_DAYS,\n \"event_time_feature\": \"exp-negative-gap-over-60-days\",\n }\n\n\ndef save_cache(\n path: Path, prepared: SequenceData, stage: str, config: Config\n) -> None:\n path.parent.mkdir(parents=True, exist_ok=True)\n np.savez(\n path,\n metadata=np.asarray(json.dumps(cache_metadata(stage, config), sort_keys=True)),\n **prepared.__dict__,\n )\n log(f\"saved public-derived full-sequence cache: {path}\")\n\n\ndef load_cache(path: Path, stage: str, config: Config) -> SequenceData:\n with np.load(path, allow_pickle=False) as artifact:\n actual = json.loads(str(artifact[\"metadata\"].item()))\n expected = cache_metadata(stage, config)\n if actual != expected:\n raise RuntimeError(f\"sequence cache differs: {actual} != {expected}\")\n values = {\n field: artifact[field].copy()\n for field in SequenceData.__dataclass_fields__\n }\n prepared = SequenceData(**values)\n log(\n f\"loaded full-sequence cache: tokens={len(prepared.sequence_items):,} \"\n f\"targets={prepared.target_count:,}\"\n )\n return prepared\n\n\ndef load_or_prepare(\n public: Path,\n stage: str,\n config: Config,\n cache: Path | None,\n) -> tuple[SequenceData, CandidateSet]:\n candidate_name = (\n \"validation_eligible_candidates.npz\"\n if stage == \"validation\"\n else \"test_eligible_candidates.npz\"\n )\n eligibility = CandidateSet.read(public / \"data\" / candidate_name)\n if cache is not None and cache.is_file():\n prepared = load_cache(cache, stage, config)\n if not np.array_equal(prepared.item_ids, eligibility.item_ids):\n raise RuntimeError(\"cached catalog differs from candidate artifact\")\n return prepared, eligibility\n prepared, built_eligibility = prepare_data(public, stage, config)\n if cache is not None:\n save_cache(cache, prepared, stage, config)\n return prepared, built_eligibility\n\n\ndef build_item_correction(\n prepared: SequenceData,\n config: Config,\n target_starts: np.ndarray,\n target_counts: np.ndarray,\n) -> torch.Tensor:\n \"\"\"Estimate in-batch propensities from the selected training targets.\"\"\"\n counts = np.zeros(len(prepared.item_ids), dtype=np.float64)\n for user_row, count in enumerate(target_counts):\n if not count:\n continue\n offset = int(prepared.offsets[user_row])\n target_start = int(target_starts[user_row])\n counts += np.bincount(\n prepared.sequence_items[\n offset\n + target_start\n + 1 : offset\n + target_start\n + int(count)\n + 1\n ].astype(np.int64),\n minlength=len(prepared.item_ids),\n )\n total = float(counts.sum())\n probabilities = (counts + 1.0) / (total + len(counts))\n raw = np.power(\n (1.0 / len(counts)) / probabilities,\n config.inbatch_correction_power,\n )\n normalization = float(np.sum(counts * raw) / max(total, 1.0))\n correction = np.clip(\n raw / max(normalization, 1.0e-12),\n config.inbatch_correction_min,\n config.inbatch_correction_max,\n ).astype(np.float32)\n mean_on_targets = float(np.sum(counts * correction) / max(total, 1.0))\n log(\n f\"inbatch correction: min={correction.min():.3f} \"\n f\"mean_on_targets={mean_on_targets:.3f} max={correction.max():.3f}\"\n )\n return torch.from_numpy(correction)\n\n\ndef build_known_positive_filter(\n prepared: SequenceData, training_counts: np.ndarray\n) -> np.ndarray:\n \"\"\"Pack each user's complete public-training positive-item set.\"\"\"\n width = (len(prepared.item_ids) + 7) // 8\n packed = np.zeros((len(prepared.user_ids), width), dtype=np.uint8)\n retained = 0\n for user_row, target_count in enumerate(training_counts):\n if not target_count:\n continue\n offset = int(prepared.offsets[user_row])\n # T adjacent targets consume sequence positions 0 through T. The\n # chronological BCE suffix starts at T + 1 and is intentionally absent.\n items = np.unique(\n prepared.sequence_items[\n offset : offset + int(target_count) + 1\n ].astype(np.int64)\n )\n byte_columns = items >> 3\n bit_values = np.left_shift(\n np.uint8(1), (items & 7).astype(np.uint8)\n )\n np.bitwise_or.at(packed[user_row], byte_columns, bit_values)\n retained += len(items)\n log(\n f\"full-history false-negative filter: users={len(prepared.user_ids):,} \"\n f\"unique_user_items={retained:,} bytes={packed.nbytes:,}\"\n )\n return packed\n\n\ndef invalid_full_history_negatives(\n negatives: np.ndarray,\n targets: np.ndarray,\n user_rows: np.ndarray,\n known_positive_packed: np.ndarray,\n liked_items: np.ndarray,\n) -> np.ndarray:\n \"\"\"Return row-level invalid negatives using the complete training prefix.\"\"\"\n byte_columns = negatives >> 3\n bit_columns = negatives & 7\n known = known_positive_packed[user_rows[:, None], byte_columns]\n invalid = ((known >> bit_columns) & 1).astype(bool)\n invalid |= negatives == targets[:, None]\n invalid |= (negatives[:, :, None] == liked_items[:, None, :]).any(axis=2)\n for column in range(1, negatives.shape[1]):\n invalid[:, column] |= (\n negatives[:, column, None] == negatives[:, :column]\n ).any(axis=1)\n return invalid\n\n\ndef draw_full_history_filtered_negatives(\n targets: torch.Tensor,\n user_rows: np.ndarray,\n liked_items: torch.Tensor,\n known_positive_packed: np.ndarray,\n n_items: int,\n config: Config,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n \"\"\"Draw mixed negatives and reject all known public-training positives.\"\"\"\n targets_cpu = targets.detach().cpu().numpy().astype(np.int64, copy=False)\n liked_cpu = liked_items.detach().cpu().numpy().astype(np.int64, copy=False)\n batch_size = len(targets_cpu)\n uniform = torch.randint(\n n_items,\n (batch_size, config.uniform_negatives),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.empty((batch_size, 0), dtype=torch.long)\n corrected_count = 0\n if config.inbatch_negatives:\n unique_targets = torch.unique(targets.detach().cpu())\n corrected_count = min(\n config.inbatch_negatives, max(len(unique_targets) - 1, 0)\n )\n if corrected_count:\n priorities = torch.rand(\n (batch_size, len(unique_targets)), generator=generator\n )\n priorities.masked_fill_(\n unique_targets[None, :].eq(\n torch.from_numpy(targets_cpu)[:, None]\n ),\n -1.0,\n )\n columns = torch.topk(\n priorities, k=corrected_count, dim=1\n ).indices\n inbatch = unique_targets[columns]\n if corrected_count < config.inbatch_negatives:\n filler = torch.randint(\n n_items,\n (batch_size, config.inbatch_negatives - corrected_count),\n generator=generator,\n device=\"cpu\",\n )\n inbatch = torch.cat([inbatch, filler], dim=1)\n\n negatives = torch.cat([uniform, inbatch], dim=1).numpy()\n correction = np.ones_like(negatives, dtype=np.float32)\n if corrected_count:\n correction[\n :, config.uniform_negatives : config.uniform_negatives + corrected_count\n ] = item_correction[inbatch[:, :corrected_count]].numpy()\n for _ in range(32):\n invalid = invalid_full_history_negatives(\n negatives,\n targets_cpu,\n user_rows,\n known_positive_packed,\n liked_cpu,\n )\n if not np.any(invalid):\n return (\n torch.from_numpy(negatives).to(device),\n torch.from_numpy(correction).to(device),\n )\n replacements = torch.randint(\n n_items,\n (int(invalid.sum()),),\n generator=generator,\n device=\"cpu\",\n ).numpy()\n negatives[invalid] = replacements\n # Rejected in-batch items become ordinary uniform draws.\n correction[invalid] = 1.0\n invalid = invalid_full_history_negatives(\n negatives,\n targets_cpu,\n user_rows,\n known_positive_packed,\n liked_cpu,\n )\n if np.any(invalid):\n raise RuntimeError(\"full-history negative resampling failed\")\n return (\n torch.from_numpy(negatives).to(device),\n torch.from_numpy(correction).to(device),\n )\n\n\ndef draw_training_negatives(\n positive_items: torch.Tensor,\n known_history: torch.Tensor,\n liked_items: torch.Tensor,\n user_rows: np.ndarray,\n known_positive_packed: np.ndarray | None,\n prepared: SequenceData,\n config: Config,\n item_correction: torch.Tensor,\n generator: torch.Generator,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor]:\n if config.negative_filter_scope == \"full-training-history\":\n if known_positive_packed is None:\n raise RuntimeError(\"full-history negative filter was not built\")\n return draw_full_history_filtered_negatives(\n positive_items,\n user_rows,\n liked_items,\n known_positive_packed,\n len(prepared.item_ids),\n config,\n item_correction,\n generator,\n device,\n )\n return base.draw_filtered_mixed_negatives(\n positive_items,\n known_history,\n liked_items,\n len(prepared.item_ids),\n config.uniform_negatives,\n config.inbatch_negatives,\n item_correction,\n generator,\n device,\n )\n\n\ndef event_sequence(\n model: base.SequentialTwoTower,\n items: torch.Tensor,\n features: torch.Tensor,\n hidden: torch.Tensor,\n) -> tuple[torch.Tensor, torch.Tensor]:\n inputs = torch.cat([model.context_embedding(items), features], dim=-1)\n projected = model.event_projection(inputs)\n return model.sequence_encoder(projected, hidden)\n\n\ndef pool_explicit_states(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n state_ids: np.ndarray,\n device: torch.device,\n) -> tuple[torch.Tensor, torch.Tensor, torch.Tensor]:\n unique_states, inverse = np.unique(state_ids, return_inverse=True)\n liked_items = torch.from_numpy(\n prepared.liked_states[unique_states].astype(np.int64)\n ).to(device)\n disliked_items = torch.from_numpy(\n prepared.disliked_states[unique_states].astype(np.int64)\n ).to(device)\n liked_vectors = model._mean_context_channel(liked_items)\n disliked_vectors = model._mean_context_channel(disliked_items)\n inverse_rows = torch.from_numpy(inverse.astype(np.int64)).to(device)\n counts = prepared.state_counts[state_ids]\n return (\n liked_vectors[inverse_rows],\n disliked_vectors[inverse_rows],\n torch.from_numpy(counts.astype(np.float32)).to(device),\n )\n\n\ndef user_queries(\n model: base.SequentialTwoTower,\n sequence_vectors: torch.Tensor,\n liked_vectors: torch.Tensor,\n disliked_vectors: torch.Tensor,\n user_features: torch.Tensor,\n) -> torch.Tensor:\n return model.user_tower(\n torch.cat(\n [sequence_vectors, liked_vectors, disliked_vectors, user_features],\n dim=1,\n )\n )\n\n\ndef candidate_vectors(\n model: base.SequentialTwoTower,\n positives: torch.Tensor,\n negatives: torch.Tensor,\n) -> tuple[torch.Tensor, torch.Tensor]:\n all_items = torch.cat([positives, negatives.reshape(-1)])\n unique_items, inverse = torch.unique(all_items, return_inverse=True)\n encoded = model.encode_items(unique_items)\n positive_vectors = encoded[inverse[: len(positives)]]\n negative_vectors = encoded[inverse[len(positives) :]].reshape(\n len(positives), negatives.shape[1], -1\n )\n return positive_vectors, negative_vectors\n\n\ndef history_windows(\n sequence: np.ndarray,\n start: int,\n count: int,\n width: int,\n padding_item: int,\n) -> np.ndarray:\n base_start = max(0, start - width + 1)\n segment = sequence[base_start : start + count]\n padded = np.concatenate(\n [np.full(width - 1, padding_item, dtype=np.uint16), segment]\n )\n windows = np.lib.stride_tricks.sliding_window_view(padded, width)\n offset = start - base_start\n return windows[offset : offset + count].copy()\n\n\ndef corrected_negative_loss(\n logits: torch.Tensor,\n correction: torch.Tensor,\n config: Config,\n) -> torch.Tensor:\n raw = F.softplus(logits)\n total = config.uniform_negatives + config.inbatch_negatives\n result = torch.zeros(len(logits), device=logits.device)\n if config.uniform_negatives:\n result += config.uniform_negatives * raw[\n :, : config.uniform_negatives\n ].mean(dim=1)\n if config.inbatch_negatives:\n selected = slice(config.uniform_negatives, total)\n weights = correction[:, selected]\n result += config.inbatch_negatives * (\n (raw[:, selected] * weights).sum(dim=1)\n / weights.sum(dim=1).clamp_min(1.0e-6)\n )\n return result / total\n\n\ndef retained_target_window_starts(\n total_targets: np.ndarray,\n validation_counts: np.ndarray,\n training_counts: np.ndarray,\n) -> np.ndarray:\n \"\"\"Return the discarded prefix before the retained chronological suffix.\"\"\"\n training_region_ends = total_targets - validation_counts\n window_starts = training_region_ends - training_counts\n if (\n np.any(window_starts < 0)\n or np.any(training_counts < 0)\n or np.any(validation_counts < 0)\n ):\n raise RuntimeError(\"chronological target partition produced negative counts\")\n if not np.array_equal(\n window_starts + training_counts + validation_counts, total_targets\n ):\n raise RuntimeError(\"chronological target window is not a final contiguous suffix\")\n return window_starts\n\n\ndef chronological_target_partition(\n prepared: SequenceData,\n validation_fraction: float,\n max_train_targets_per_user: int,\n max_validation_targets_per_user: int,\n) -> TargetPartition:\n \"\"\"Select a recent contiguous window with a final out-of-time BCE suffix.\"\"\"\n totals = np.maximum(prepared.lengths - 1, 0).astype(np.int64)\n validation = np.floor(totals * validation_fraction).astype(np.int64)\n validation[(totals >= 2) & (validation == 0)] = 1\n if max_validation_targets_per_user:\n validation = np.minimum(validation, max_validation_targets_per_user)\n\n history_training = totals - validation\n training = history_training.copy()\n if max_train_targets_per_user:\n training = np.minimum(training, max_train_targets_per_user)\n window_starts = retained_target_window_starts(totals, validation, training)\n return TargetPartition(\n window_starts=window_starts,\n training_counts=training,\n validation_counts=validation,\n history_training_counts=history_training,\n )\n\n\ndef load_organic_like_outcomes(\n public: Path, stage: str, item_ids: np.ndarray\n) -> dict[int, dict[int, np.ndarray]]:\n \"\"\"Load mapped organic-like timestamps from public fitting history.\"\"\"\n prefixes = (\"train\",) if stage == \"validation\" else (\"train\", \"val\")\n frames = [\n pd.read_parquet(\n public / \"data\" / f\"{prefix}_likes.parquet\",\n columns=[\"uid\", \"timestamp\", \"item_id\", \"is_organic\"],\n )\n for prefix in prefixes\n ]\n likes = pd.concat(frames, ignore_index=True)\n likes = likes[likes.is_organic == 1].copy()\n mapped = pd.Index(item_ids).get_indexer(\n likes.item_id.to_numpy(dtype=np.int64)\n )\n likes = likes.loc[mapped >= 0, [\"uid\", \"timestamp\"]].copy()\n likes[\"item\"] = mapped[mapped >= 0].astype(np.int32, copy=False)\n likes = likes.sort_values([\"uid\", \"item\", \"timestamp\"], kind=\"stable\")\n nested: dict[int, dict[int, list[int]]] = {}\n for raw_uid, raw_timestamp, raw_item in likes.itertuples(\n index=False, name=None\n ):\n nested.setdefault(int(raw_uid), {}).setdefault(int(raw_item), []).append(\n int(raw_timestamp)\n )\n result = {\n uid: {\n item: np.asarray(timestamps, dtype=np.int64)\n for item, timestamps in items.items()\n }\n for uid, items in nested.items()\n }\n log(\n f\"loaded organic-like outcomes: rows={len(likes):,} \"\n f\"users={len(result):,}\"\n )\n return result\n\n\ndef build_target_weights(\n public: Path,\n stage: str,\n prepared: SequenceData,\n training_counts: np.ndarray,\n config: Config,\n) -> np.ndarray:\n \"\"\"Build nonzero per-token weights without dropping adjacent targets.\"\"\"\n if config.target_weighting == \"uniform\":\n weights = np.ones(len(prepared.sequence_items), dtype=np.float32)\n else:\n weights = np.full(\n len(prepared.sequence_items),\n config.ordinary_target_weight,\n dtype=np.float32,\n )\n organic_strong = (\n (prepared.event_features[:, 0] > 0.5)\n & (prepared.event_features[:, 1] >= 0.8)\n )\n weights[organic_strong] = config.organic_strong_target_weight\n if config.target_weighting == \"discovery-graded\":\n like_outcomes = load_organic_like_outcomes(\n public, stage, prepared.item_ids\n )\n history_end = (\n base.TRAIN_END if stage == \"validation\" else base.VALIDATION_END\n )\n horizon = config.discovery_horizon_days * base.DAY\n grade_three = 0\n for user_row, uid in enumerate(prepared.user_ids):\n start = int(prepared.offsets[user_row])\n end = int(prepared.offsets[user_row + 1])\n if end - start < 2:\n continue\n last_training_target = start + int(training_counts[user_row])\n first_validation_target = last_training_target + 1\n split_timestamp = (\n int(prepared.sequence_timestamps[first_validation_target])\n if first_validation_target < end\n else history_end\n )\n user_likes = like_outcomes.get(int(uid), {})\n seen = {int(prepared.sequence_items[start])}\n for position in range(start + 1, end):\n item = int(prepared.sequence_items[position])\n novel = item not in seen\n seen.add(item)\n if not novel:\n # Repeats remain training examples, but they cannot be\n # eligible discovery outcomes at the corresponding\n # cutoff and therefore keep only the ordinary weight.\n weights[position] = config.ordinary_target_weight\n continue\n timestamps = user_likes.get(item)\n if timestamps is None:\n continue\n timestamp = int(prepared.sequence_timestamps[position])\n upper = min(\n timestamp + horizon,\n split_timestamp\n if position <= last_training_target\n else history_end,\n )\n outcome = int(np.searchsorted(timestamps, timestamp, side=\"left\"))\n if outcome < len(timestamps) and int(timestamps[outcome]) < upper:\n weights[position] = config.organic_like_target_weight\n grade_three += 1\n log(\n f\"grade-3 first-encounter targets={grade_three:,} \"\n f\"horizon_days={config.discovery_horizon_days}\"\n )\n log(\n f\"target weighting={config.target_weighting} \"\n f\"min={weights.min():.3f} mean={weights.mean():.3f} \"\n f\"max={weights.max():.3f}\"\n )\n return weights\n\n\ndef build_dense_block(\n prepared: SequenceData,\n target_weights: np.ndarray,\n users: np.ndarray,\n window_starts: np.ndarray,\n process_counts: np.ndarray,\n score_starts: np.ndarray,\n start: int,\n steps: int,\n config: Config,\n) -> tuple[np.ndarray, ...]:\n \"\"\"Build one causal sequence block and mark the target rows to score.\"\"\"\n batch_size = len(users)\n padding_item = len(prepared.item_ids)\n input_items = np.full((batch_size, steps), padding_item, dtype=np.uint16)\n event_features = np.zeros((batch_size, steps, 3), dtype=np.float16)\n activity_features = np.zeros((batch_size, steps, 4), dtype=np.float16)\n state_ids = np.zeros((batch_size, steps), dtype=np.uint32)\n targets = np.zeros((batch_size, steps), dtype=np.uint16)\n weights = np.ones((batch_size, steps), dtype=np.float32)\n scored = np.zeros((batch_size, steps), dtype=bool)\n histories = np.full(\n (batch_size, steps, config.bptt_steps),\n padding_item,\n dtype=np.uint16,\n )\n processed = np.zeros(batch_size, dtype=np.int32)\n\n lengths = prepared.lengths\n for row, user_row in enumerate(users):\n count = min(steps, max(int(process_counts[row]) - start, 0))\n if not count:\n continue\n offset = int(prepared.offsets[user_row])\n window_start = int(window_starts[row])\n absolute_start = window_start + start\n positions = slice(\n offset + absolute_start,\n offset + absolute_start + count,\n )\n target_positions = slice(\n offset + absolute_start + 1,\n offset + absolute_start + count + 1,\n )\n input_items[row, :count] = prepared.sequence_items[positions]\n event_features[row, :count] = prepared.event_features[positions]\n activity_features[row, :count] = prepared.activity_features[positions]\n state_ids[row, :count] = prepared.state_ids[positions]\n targets[row, :count] = prepared.sequence_items[target_positions]\n weights[row, :count] = target_weights[target_positions]\n score_from = min(count, max(int(score_starts[row]) - start, 0))\n scored[row, score_from:count] = True\n processed[row] = count\n user_sequence = prepared.sequence_items[\n offset : offset + int(lengths[user_row])\n ]\n histories[row, :count] = history_windows(\n user_sequence,\n absolute_start,\n count,\n config.bptt_steps,\n padding_item,\n )\n return (\n input_items,\n event_features,\n activity_features,\n state_ids,\n targets,\n scored,\n histories,\n processed,\n weights,\n )\n\n\ndef score_target_microbatches(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n sequence_vectors: torch.Tensor,\n flat_states: np.ndarray,\n flat_activity: np.ndarray,\n positive_items: torch.Tensor,\n negatives: torch.Tensor,\n negative_weights: torch.Tensor,\n sample_weights: np.ndarray,\n device: torch.device,\n backward: bool,\n) -> tuple[dict[str, float], torch.Tensor | None]:\n \"\"\"Compute one logical block's exact BCE in bounded-memory slices.\"\"\"\n target_count = len(positive_items)\n total_weight = float(sample_weights.sum())\n if total_weight <= 0:\n raise RuntimeError(\"target block has no positive sample weight\")\n sequence_gradient = torch.zeros_like(sequence_vectors) if backward else None\n statistics = {\n \"loss_sum\": 0.0,\n \"positive_sum\": 0.0,\n \"negative_sum\": 0.0,\n \"pair_correct\": 0.0,\n \"weight_sum\": 0.0,\n }\n for begin in range(0, target_count, config.loss_microbatch_size):\n end = min(begin + config.loss_microbatch_size, target_count)\n selected_states = flat_states[begin:end]\n if backward:\n sequence_input = (\n sequence_vectors[begin:end].detach().requires_grad_(True)\n )\n else:\n sequence_input = sequence_vectors[begin:end]\n liked_vectors, disliked_vectors, _ = pool_explicit_states(\n model, prepared, selected_states, device\n )\n counts = prepared.state_counts[selected_states]\n numeric = np.zeros((end - begin, 6), dtype=np.float32)\n numeric[:, :4] = flat_activity[begin:end]\n numeric[:, 4] = np.log1p(counts[:, 0]) / 6.0\n numeric[:, 5] = np.log1p(counts[:, 1]) / 4.0\n query_vectors = user_queries(\n model,\n sequence_input,\n liked_vectors,\n disliked_vectors,\n torch.from_numpy(numeric).to(device),\n )\n positive_vectors, negative_vectors = candidate_vectors(\n model,\n positive_items[begin:end],\n negatives[begin:end],\n )\n positive_logits = (query_vectors * positive_vectors).sum(dim=1)\n negative_logits = torch.einsum(\n \"bd,bkd->bk\", query_vectors, negative_vectors\n )\n selected_weights = torch.from_numpy(\n sample_weights[begin:end].astype(np.float32, copy=False)\n ).to(device)\n if config.objective == \"infonce\":\n cat_logits = torch.cat(\n [positive_logits.unsqueeze(1), negative_logits], dim=1\n ) / config.temperature\n labels = torch.zeros(\n len(positive_logits), dtype=torch.long, device=device\n )\n raw_loss = F.cross_entropy(cat_logits, labels, reduction=\"none\")\n loss_sum = (selected_weights * raw_loss).sum()\n else:\n positive_loss = F.softplus(-positive_logits)\n negative_loss = corrected_negative_loss(\n negative_logits, negative_weights[begin:end], config\n )\n loss_sum = (\n selected_weights * (positive_loss + negative_loss) / 2.0\n ).sum()\n if not bool(torch.isfinite(loss_sum)):\n raise RuntimeError(\"dense next-item training produced non-finite loss\")\n if backward:\n (loss_sum / total_weight).backward()\n if sequence_input.grad is None:\n raise RuntimeError(\"sequence microbatch did not produce a gradient\")\n sequence_gradient[begin:end] = sequence_input.grad\n statistics[\"loss_sum\"] += float(loss_sum.detach().cpu())\n statistics[\"positive_sum\"] += float(positive_logits.detach().sum().cpu())\n statistics[\"negative_sum\"] += float(negative_logits.detach().sum().cpu())\n statistics[\"pair_correct\"] += float(\n (positive_logits.detach()[:, None] > negative_logits.detach()).sum().cpu()\n )\n statistics[\"weight_sum\"] += float(selected_weights.sum().detach().cpu())\n return statistics, sequence_gradient\n\n\ndef release_accelerator_cache(device: torch.device) -> None:\n if device.type == \"mps\":\n torch.mps.empty_cache()\n\n\ndef run_training_epoch(\n model: base.SequentialTwoTower,\n optimizer: torch.optim.Optimizer,\n prepared: SequenceData,\n config: Config,\n target_weights: np.ndarray,\n window_starts: np.ndarray,\n target_counts: np.ndarray,\n item_correction: torch.Tensor,\n known_positive_packed: np.ndarray | None,\n generator: torch.Generator,\n device: torch.device,\n epoch: int,\n) -> dict[str, float]:\n model.train()\n train_users = np.flatnonzero(target_counts > 0)\n sorted_users = train_users[\n np.argsort(target_counts[train_users], kind=\"stable\")\n ]\n groups = [\n sorted_users[start : start + config.user_batch_size]\n for start in range(0, len(sorted_users), config.user_batch_size)\n ]\n group_order = torch.randperm(len(groups), generator=generator).numpy()\n negative_count = config.uniform_negatives + config.inbatch_negatives\n total_loss = 0.0\n total_targets = 0\n total_weight = 0.0\n total_positive = 0.0\n total_negative = 0.0\n total_pair_correct = 0.0\n batches = 0\n started = time.monotonic()\n\n for group_number in group_order:\n users = groups[int(group_number)]\n group_counts = target_counts[users]\n hidden = torch.zeros(\n 1, len(users), config.embedding_dim, device=device\n )\n max_targets = int(group_counts.max())\n for start in range(0, max_targets, config.bptt_steps):\n steps = min(config.bptt_steps, max_targets - start)\n block = build_dense_block(\n prepared,\n target_weights,\n users,\n window_starts[users],\n group_counts,\n np.zeros(len(users), dtype=np.int64),\n start,\n steps,\n config,\n )\n sequence_outputs, _ = event_sequence(\n model,\n torch.from_numpy(block[0].astype(np.int64)).to(device),\n torch.from_numpy(block[1].astype(np.float32)).to(device),\n hidden,\n )\n hidden = torch.stack(\n [\n sequence_outputs[row, count - 1]\n if count\n else hidden[0, row]\n for row, count in enumerate(block[7])\n ]\n ).unsqueeze(0).detach()\n flat_scored = block[5].reshape(-1)\n if not np.any(flat_scored):\n continue\n mask = torch.from_numpy(flat_scored).to(device)\n sequence_vectors = sequence_outputs.reshape(\n -1, config.embedding_dim\n )[mask]\n flat_states = block[3].reshape(-1)[flat_scored]\n positive_items = torch.from_numpy(\n block[4].reshape(-1)[flat_scored].astype(np.int64)\n ).to(device)\n known_history = torch.from_numpy(\n block[6].reshape(-1, config.bptt_steps)[flat_scored].astype(\n np.int64\n )\n ).to(device)\n liked_items = torch.from_numpy(\n prepared.liked_states[flat_states].astype(np.int64)\n ).to(device)\n flat_user_rows = np.broadcast_to(\n users[:, None], block[5].shape\n ).reshape(-1)[flat_scored]\n negatives, negative_weights = draw_training_negatives(\n positive_items,\n known_history,\n liked_items,\n flat_user_rows,\n known_positive_packed,\n prepared,\n config,\n item_correction,\n generator,\n device,\n )\n optimizer.zero_grad(set_to_none=True)\n statistics, sequence_gradient = score_target_microbatches(\n model,\n prepared,\n config,\n sequence_vectors,\n flat_states,\n block[2].reshape(-1, 4)[flat_scored].astype(np.float32),\n positive_items,\n negatives,\n negative_weights,\n block[8].reshape(-1)[flat_scored],\n device,\n backward=True,\n )\n if sequence_gradient is None:\n raise RuntimeError(\"training block omitted its sequence gradient\")\n sequence_vectors.backward(sequence_gradient)\n nn.utils.clip_grad_norm_(model.parameters(), config.gradient_clip)\n optimizer.step()\n\n count = len(positive_items)\n total_loss += statistics[\"loss_sum\"]\n total_targets += count\n total_weight += statistics[\"weight_sum\"]\n total_positive += statistics[\"positive_sum\"]\n total_negative += statistics[\"negative_sum\"]\n total_pair_correct += statistics[\"pair_correct\"]\n batches += 1\n if batches % 25 == 0:\n release_accelerator_cache(device)\n if batches % 250 == 0:\n batch_bce = statistics[\"loss_sum\"] / max(\n statistics[\"weight_sum\"], 1.0e-12\n )\n running_bce = total_loss / max(total_weight, 1.0e-12)\n log(\n f\"epoch={epoch}/{config.epochs} batches={batches:,} \"\n f\"targets={total_targets:,} \"\n f\"batch_bce={batch_bce:.6f} \"\n f\"running_bce={running_bce:.6f} \"\n f\"elapsed={time.monotonic() - started:.1f}s\"\n )\n\n return {\n \"epoch\": float(epoch),\n \"loss\": total_loss / max(total_weight, 1.0e-12),\n \"positive_logit\": total_positive / max(total_targets, 1),\n \"negative_logit\": total_negative / max(total_targets * negative_count, 1),\n \"pair_accuracy\": total_pair_correct\n / max(total_targets * negative_count, 1),\n \"targets\": float(total_targets),\n \"target_weight_sum\": total_weight,\n \"batches\": float(batches),\n \"seconds\": time.monotonic() - started,\n }\n\n\n@torch.no_grad()\ndef evaluate_validation_bce(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n target_weights: np.ndarray,\n window_starts: np.ndarray,\n training_counts: np.ndarray,\n validation_counts: np.ndarray,\n item_correction: torch.Tensor,\n known_positive_packed: np.ndarray | None,\n device: torch.device,\n) -> dict[str, float]:\n \"\"\"Measure deterministic BCE on each user's chronological target suffix.\"\"\"\n model.eval()\n totals = training_counts + validation_counts\n users_with_validation = np.flatnonzero(validation_counts > 0)\n ordered = users_with_validation[\n np.argsort(totals[users_with_validation], kind=\"stable\")\n ]\n groups = [\n ordered[start : start + config.user_batch_size]\n for start in range(0, len(ordered), config.user_batch_size)\n ]\n generator = torch.Generator(device=\"cpu\").manual_seed(config.seed + 20_000)\n total_loss = 0.0\n total_targets = 0\n total_weight = 0.0\n started = time.monotonic()\n\n for group_number, users in enumerate(groups, start=1):\n group_totals = totals[users]\n group_starts = training_counts[users]\n hidden = torch.zeros(\n 1, len(users), config.embedding_dim, device=device\n )\n max_targets = int(group_totals.max())\n for start in range(0, max_targets, config.bptt_steps):\n steps = min(config.bptt_steps, max_targets - start)\n block = build_dense_block(\n prepared,\n target_weights,\n users,\n window_starts[users],\n group_totals,\n group_starts,\n start,\n steps,\n config,\n )\n sequence_outputs, _ = event_sequence(\n model,\n torch.from_numpy(block[0].astype(np.int64)).to(device),\n torch.from_numpy(block[1].astype(np.float32)).to(device),\n hidden,\n )\n hidden = torch.stack(\n [\n sequence_outputs[row, count - 1]\n if count\n else hidden[0, row]\n for row, count in enumerate(block[7])\n ]\n ).unsqueeze(0)\n flat_scored = block[5].reshape(-1)\n if not np.any(flat_scored):\n continue\n mask = torch.from_numpy(flat_scored).to(device)\n sequence_vectors = sequence_outputs.reshape(\n -1, config.embedding_dim\n )[mask]\n flat_states = block[3].reshape(-1)[flat_scored]\n positive_items = torch.from_numpy(\n block[4].reshape(-1)[flat_scored].astype(np.int64)\n ).to(device)\n known_history = torch.from_numpy(\n block[6].reshape(-1, config.bptt_steps)[flat_scored].astype(\n np.int64\n )\n ).to(device)\n liked_items = torch.from_numpy(\n prepared.liked_states[flat_states].astype(np.int64)\n ).to(device)\n flat_user_rows = np.broadcast_to(\n users[:, None], block[5].shape\n ).reshape(-1)[flat_scored]\n negatives, negative_weights = draw_training_negatives(\n positive_items,\n known_history,\n liked_items,\n flat_user_rows,\n known_positive_packed,\n prepared,\n config,\n item_correction,\n generator,\n device,\n )\n statistics, _ = score_target_microbatches(\n model,\n prepared,\n config,\n sequence_vectors,\n flat_states,\n block[2].reshape(-1, 4)[flat_scored].astype(np.float32),\n positive_items,\n negatives,\n negative_weights,\n block[8].reshape(-1)[flat_scored],\n device,\n backward=False,\n )\n total_loss += statistics[\"loss_sum\"]\n total_targets += len(positive_items)\n total_weight += statistics[\"weight_sum\"]\n if group_number % 25 == 0:\n release_accelerator_cache(device)\n return {\n \"validation_loss\": total_loss / max(total_weight, 1.0e-12),\n \"validation_targets\": float(total_targets),\n \"validation_target_weight_sum\": total_weight,\n \"validation_seconds\": time.monotonic() - started,\n }\n\n\ndef snapshot_model(model: nn.Module) -> dict[str, torch.Tensor]:\n return {\n name: value.detach().cpu().clone()\n for name, value in model.state_dict().items()\n }\n\n\ndef save_epoch_checkpoint(\n checkpoint_dir: Path,\n epoch: int,\n model_state: dict[str, torch.Tensor],\n config: Config,\n diagnostics: dict[str, float],\n optimizer: torch.optim.Optimizer,\n generator: torch.Generator,\n) -> Path:\n checkpoint_dir.mkdir(parents=True, exist_ok=True)\n path = checkpoint_dir / f\"epoch-{epoch:02d}.pt\"\n torch.save(\n {\n \"epoch\": epoch,\n \"config\": asdict(config),\n \"diagnostics\": diagnostics,\n \"model_state\": model_state,\n \"optimizer_state\": optimizer.state_dict(),\n \"generator_state\": generator.get_state(),\n },\n path,\n )\n log(f\"saved epoch checkpoint: {path}\")\n return path\n\n\ndef train_model(\n public: Path,\n stage: str,\n prepared: SequenceData,\n config: Config,\n device: torch.device,\n checkpoint_dir: Path | None,\n resume_checkpoint: Path | None,\n) -> tuple[base.SequentialTwoTower, list[dict[str, float]]]:\n partition = chronological_target_partition(\n prepared,\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n training_counts = partition.training_counts\n validation_counts = partition.validation_counts\n log(\n f\"chronological BCE window: train={int(training_counts.sum()):,} \"\n f\"validation={int(validation_counts.sum()):,} \"\n f\"selected={partition.selected_count:,} \"\n f\"discarded_old_targets={int(partition.window_starts.sum()):,} \"\n f\"all={prepared.target_count:,}\"\n )\n model = base.SequentialTwoTower(prepared.item_features, model_config(config)).to(\n device\n )\n optimizer = torch.optim.AdamW(\n model.parameters(), lr=config.learning_rate, weight_decay=config.weight_decay\n )\n generator = torch.Generator(device=\"cpu\").manual_seed(config.seed + 10_000)\n item_correction = build_item_correction(\n prepared,\n config,\n partition.window_starts,\n training_counts,\n )\n target_weights = build_target_weights(\n public,\n stage,\n prepared,\n partition.history_training_counts,\n config,\n )\n known_positive_packed = (\n build_known_positive_filter(prepared, partition.history_training_counts)\n if config.negative_filter_scope == \"full-training-history\"\n else None\n )\n history: list[dict[str, float]] = []\n best_loss = float(\"inf\")\n best_epoch = 0\n best_state: dict[str, torch.Tensor] | None = None\n epochs_without_improvement = 0\n start_epoch = 0\n\n if resume_checkpoint is not None:\n payload = torch.load(\n resume_checkpoint, map_location=device, weights_only=False\n )\n stored_config = payload.get(\"config\", {})\n architecture_fields = (\n \"bptt_steps\",\n \"embedding_dim\",\n \"hidden_dim\",\n \"user_hidden_1\",\n \"user_hidden_2\",\n \"item_hidden_1\",\n \"item_hidden_2\",\n )\n mismatches = {\n name: (stored_config.get(name), getattr(config, name))\n for name in architecture_fields\n if stored_config.get(name) != getattr(config, name)\n }\n if mismatches:\n raise RuntimeError(\n f\"resume checkpoint architecture differs: {mismatches}\"\n )\n stored_filter_scope = stored_config.get(\n \"negative_filter_scope\", \"last-100\"\n )\n stored_target_weighting = stored_config.get(\n \"target_weighting\", \"uniform\"\n )\n stored_weight_values = (\n stored_config.get(\"ordinary_target_weight\", 1.0),\n stored_config.get(\"organic_strong_target_weight\", 1.0),\n stored_config.get(\"organic_like_target_weight\", 3.0),\n stored_config.get(\"discovery_horizon_days\", 7),\n )\n current_weight_values = (\n config.ordinary_target_weight,\n config.organic_strong_target_weight,\n config.organic_like_target_weight,\n config.discovery_horizon_days,\n )\n stored_window_values = (\n stored_config.get(\"validation_fraction\", 0.10),\n stored_config.get(\"max_train_targets_per_user\", 0),\n stored_config.get(\"max_validation_targets_per_user\", 0),\n )\n current_window_values = (\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n objective_changed = (\n stored_filter_scope != config.negative_filter_scope\n or stored_target_weighting != config.target_weighting\n or stored_weight_values != current_weight_values\n or stored_window_values != current_window_values\n )\n model.load_state_dict(payload[\"model_state\"])\n if \"optimizer_state\" in payload and not objective_changed:\n optimizer.load_state_dict(payload[\"optimizer_state\"])\n for group in optimizer.param_groups:\n group[\"lr\"] = config.learning_rate\n group[\"weight_decay\"] = config.weight_decay\n log(f\"restored optimizer state from {resume_checkpoint}\")\n else:\n log(\n f\"using a fresh AdamW optimizer after loading {resume_checkpoint}\"\n )\n if \"generator_state\" in payload and not objective_changed:\n generator.set_state(payload[\"generator_state\"])\n start_epoch = int(payload[\"epoch\"])\n if start_epoch >= config.epochs:\n raise RuntimeError(\n f\"resume epoch {start_epoch} must be below --epochs {config.epochs}\"\n )\n prior = {\n str(name): float(value)\n for name, value in payload.get(\"diagnostics\", {}).items()\n }\n if objective_changed:\n prior.pop(\"validation_loss\", None)\n log(\n \"BCE objective changed: filter \"\n f\"{stored_filter_scope}->{config.negative_filter_scope}, \"\n \"target weighting \"\n f\"{stored_target_weighting}->{config.target_weighting}, \"\n f\"target window {stored_window_values}->{current_window_values}; \"\n \"recomputing the resume checkpoint's validation BCE\"\n )\n if \"validation_loss\" not in prior:\n prior.update(\n evaluate_validation_bce(\n model,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n validation_counts,\n item_correction,\n known_positive_packed,\n device,\n )\n )\n history.append(prior)\n best_loss = float(prior[\"validation_loss\"])\n best_epoch = start_epoch\n best_state = snapshot_model(model)\n log(\n f\"resumed validation-BCE training: epoch={start_epoch} \"\n f\"loss={best_loss:.6f}\"\n )\n\n for epoch in range(start_epoch + 1, config.epochs + 1):\n diagnostics = run_training_epoch(\n model,\n optimizer,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n item_correction,\n known_positive_packed,\n generator,\n device,\n epoch,\n )\n validation = evaluate_validation_bce(\n model,\n prepared,\n config,\n target_weights,\n partition.window_starts,\n training_counts,\n validation_counts,\n item_correction,\n known_positive_packed,\n device,\n )\n diagnostics.update(validation)\n history.append(diagnostics)\n current_state = snapshot_model(model)\n if checkpoint_dir is not None:\n save_epoch_checkpoint(\n checkpoint_dir,\n epoch,\n current_state,\n config,\n diagnostics,\n optimizer,\n generator,\n )\n improved = (\n diagnostics[\"validation_loss\"]\n < best_loss - config.early_stopping_min_delta\n )\n if improved:\n best_loss = diagnostics[\"validation_loss\"]\n best_epoch = epoch\n best_state = current_state\n epochs_without_improvement = 0\n if checkpoint_dir is not None:\n torch.save(\n {\n \"epoch\": epoch,\n \"config\": asdict(config),\n \"diagnostics\": diagnostics,\n \"model_state\": current_state,\n \"optimizer_state\": optimizer.state_dict(),\n \"generator_state\": generator.get_state(),\n },\n checkpoint_dir / \"best.pt\",\n )\n else:\n epochs_without_improvement += 1\n log(\n f\"epoch={epoch}/{config.epochs} train_loss={diagnostics['loss']:.5f} \"\n f\"validation_loss={diagnostics['validation_loss']:.5f} \"\n f\"pair_accuracy={diagnostics['pair_accuracy']:.3f} \"\n f\"train_targets={int(diagnostics['targets']):,} \"\n f\"validation_targets={int(diagnostics['validation_targets']):,} \"\n f\"train_seconds={diagnostics['seconds']:.1f} \"\n f\"validation_seconds={diagnostics['validation_seconds']:.1f}\"\n )\n release_accelerator_cache(device)\n if epochs_without_improvement >= config.early_stopping_patience:\n log(\n f\"early stopping after epoch={epoch}; \"\n f\"best_epoch={best_epoch} best_validation_loss={best_loss:.6f}\"\n )\n break\n\n if best_state is None:\n raise RuntimeError(\"training did not produce a valid checkpoint\")\n model.load_state_dict(best_state)\n model.to(device).eval()\n log(\n f\"restored best validation-BCE checkpoint: epoch={best_epoch} \"\n f\"loss={best_loss:.6f}\"\n )\n return model, history\n\n\n@torch.inference_mode()\ndef inference_queries(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n config: Config,\n device: torch.device,\n) -> torch.Tensor:\n sequence_rows = prepared.inference_user_rows\n lengths = np.where(\n sequence_rows >= 0, prepared.lengths[np.maximum(sequence_rows, 0)], 0\n )\n ordered = np.argsort(lengths, kind=\"stable\")\n result = torch.zeros(\n len(sequence_rows), config.embedding_dim, dtype=torch.float32\n )\n for group_start in range(0, len(ordered), config.user_batch_size):\n output_rows = ordered[group_start : group_start + config.user_batch_size]\n rows = sequence_rows[output_rows]\n group_lengths = lengths[output_rows]\n hidden = torch.zeros(\n 1, len(output_rows), config.embedding_dim, device=device\n )\n for start in range(0, int(group_lengths.max(initial=0)), config.bptt_steps):\n steps = min(\n config.bptt_steps,\n int(group_lengths.max(initial=0)) - start,\n )\n items = np.full(\n (len(output_rows), steps),\n len(prepared.item_ids),\n dtype=np.uint16,\n )\n features = np.zeros((len(output_rows), steps, 3), dtype=np.float16)\n counts = np.zeros(len(output_rows), dtype=np.int32)\n for local_row, sequence_row in enumerate(rows):\n if sequence_row < 0:\n continue\n count = min(steps, max(int(group_lengths[local_row]) - start, 0))\n if not count:\n continue\n offset = int(prepared.offsets[sequence_row])\n selected = slice(offset + start, offset + start + count)\n items[local_row, :count] = prepared.sequence_items[selected]\n features[local_row, :count] = prepared.event_features[selected]\n counts[local_row] = count\n outputs, _ = event_sequence(\n model,\n torch.from_numpy(items.astype(np.int64)).to(device),\n torch.from_numpy(features.astype(np.float32)).to(device),\n hidden,\n )\n next_rows: list[torch.Tensor] = []\n for local_row, count in enumerate(counts):\n next_rows.append(\n outputs[local_row, count - 1]\n if count\n else hidden[0, local_row]\n )\n hidden = torch.stack(next_rows).unsqueeze(0)\n\n liked_items = torch.from_numpy(\n prepared.inference_liked_items[output_rows].astype(np.int64)\n ).to(device)\n disliked_items = torch.from_numpy(\n prepared.inference_disliked_items[output_rows].astype(np.int64)\n ).to(device)\n queries = user_queries(\n model,\n hidden.squeeze(0),\n model._mean_context_channel(liked_items),\n model._mean_context_channel(disliked_items),\n torch.from_numpy(\n prepared.inference_user_features[output_rows].astype(np.float32)\n ).to(device),\n )\n result[torch.from_numpy(output_rows.astype(np.int64))] = queries.cpu()\n return result\n\n\n@torch.inference_mode()\ndef rank_candidates(\n model: base.SequentialTwoTower,\n prepared: SequenceData,\n eligibility: CandidateSet,\n config: Config,\n device: torch.device,\n ranking_batch_size: int,\n) -> tuple[pd.DataFrame, dict[str, float]]:\n model.eval()\n queries = inference_queries(model, prepared, config, device)\n item_parts: list[torch.Tensor] = []\n for start in range(0, len(prepared.item_ids), 2_048):\n rows = torch.arange(\n start,\n min(start + 2_048, len(prepared.item_ids)),\n dtype=torch.long,\n device=device,\n )\n item_parts.append(model.encode_items(rows).cpu())\n items = torch.cat(item_parts)\n rows: list[tuple[int, int, int]] = []\n for start in range(0, len(eligibility.uids), ranking_batch_size):\n end = min(start + ranking_batch_size, len(eligibility.uids))\n scores = queries[start:end] @ items.T\n eligible = np.unpackbits(\n eligibility.packed[start:end],\n axis=1,\n count=len(prepared.item_ids),\n bitorder=\"little\",\n ).astype(bool, copy=False)\n scores.masked_fill_(~torch.from_numpy(eligible), -torch.inf)\n selected = torch.topk(scores, k=base.TOP_N, dim=1).indices.numpy()\n for local_row, columns in enumerate(selected):\n uid = int(eligibility.uids[start + local_row])\n rows.extend(\n (uid, int(prepared.item_ids[column]), rank)\n for rank, column in enumerate(columns, start=1)\n )\n if end == len(eligibility.uids) or end % 1_024 == 0:\n log(f\"ranked users={end:,}/{len(eligibility.uids):,}\")\n query_norm = torch.linalg.vector_norm(queries, dim=1).numpy()\n item_norm = torch.linalg.vector_norm(items, dim=1).numpy()\n diagnostics = {\n \"query_norm_mean\": float(query_norm.mean()),\n \"query_norm_std\": float(query_norm.std()),\n \"item_norm_mean\": float(item_norm.mean()),\n \"item_norm_std\": float(item_norm.std()),\n }\n return pd.DataFrame(rows, columns=[\"uid\", \"item_id\", \"rank\"]), diagnostics\n\n\ndef write_ranking(frame: pd.DataFrame, output: Path) -> None:\n output.parent.mkdir(parents=True, exist_ok=True)\n if output.suffix == \".parquet\":\n frame.to_parquet(output, index=False, compression=\"zstd\")\n elif output.suffix == \".csv\":\n frame.to_csv(output, index=False)\n else:\n raise ValueError(\"--output must end in .parquet or .csv\")\n\n\ndef main() -> int:\n parser = argparse.ArgumentParser(description=__doc__)\n parser.add_argument(\n \"--stage\", choices=(\"validation\", \"submission\"), default=\"validation\"\n )\n parser.add_argument(\"--output\", type=Path, required=True)\n parser.add_argument(\"--public-dir\", type=Path, required=True)\n parser.add_argument(\"--prepared-cache\", type=Path)\n parser.add_argument(\"--device\", choices=(\"cpu\", \"mps\"), default=\"cpu\")\n parser.add_argument(\"--threads\", type=int, default=8)\n parser.add_argument(\"--bptt-steps\", type=int, default=Config.bptt_steps)\n parser.add_argument(\n \"--user-batch-size\", type=int, default=Config.user_batch_size\n )\n parser.add_argument(\"--epochs\", type=int, default=Config.epochs)\n parser.add_argument(\"--learning-rate\", type=float, default=Config.learning_rate)\n parser.add_argument(\n \"--loss-microbatch-size\",\n type=int,\n default=Config.loss_microbatch_size,\n )\n parser.add_argument(\n \"--negative-filter-scope\",\n choices=(\"full-training-history\", \"last-100\"),\n default=Config.negative_filter_scope,\n )\n parser.add_argument(\n \"--objective\",\n choices=(\"bce\", \"infonce\"),\n default=Config.objective,\n )\n parser.add_argument(\n \"--temperature\",\n type=float,\n default=Config.temperature,\n )\n parser.add_argument(\n \"--target-weighting\",\n choices=(\"discovery-graded\", \"organic-strong\", \"uniform\"),\n default=Config.target_weighting,\n )\n parser.add_argument(\n \"--ordinary-target-weight\",\n type=float,\n default=Config.ordinary_target_weight,\n )\n parser.add_argument(\n \"--organic-strong-target-weight\",\n type=float,\n default=Config.organic_strong_target_weight,\n )\n parser.add_argument(\n \"--organic-like-target-weight\",\n type=float,\n default=Config.organic_like_target_weight,\n )\n parser.add_argument(\n \"--discovery-horizon-days\",\n type=int,\n default=Config.discovery_horizon_days,\n )\n parser.add_argument(\n \"--validation-fraction\",\n type=float,\n default=Config.validation_fraction,\n )\n parser.add_argument(\n \"--max-train-targets-per-user\",\n type=int,\n default=Config.max_train_targets_per_user,\n help=\"retain this many recent pre-validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--max-validation-targets-per-user\",\n type=int,\n default=Config.max_validation_targets_per_user,\n help=\"retain this many final validation targets per user; 0 is uncapped\",\n )\n parser.add_argument(\n \"--early-stopping-patience\",\n type=int,\n default=Config.early_stopping_patience,\n )\n parser.add_argument(\n \"--early-stopping-min-delta\",\n type=float,\n default=Config.early_stopping_min_delta,\n )\n parser.add_argument(\"--checkpoint-dir\", type=Path)\n parser.add_argument(\"--resume-checkpoint\", type=Path)\n parser.add_argument(\"--seed\", type=int, default=Config.seed)\n parser.add_argument(\"--ranking-batch-size\", type=int, default=64)\n parser.add_argument(\"--prepare-only\", action=\"store_true\")\n args = parser.parse_args()\n\n if args.threads < 1 or args.bptt_steps < 2 or args.user_batch_size < 1:\n parser.error(\"threads, BPTT steps, and user batch size must be positive\")\n if (\n args.epochs < 1\n or args.learning_rate <= 0\n or args.loss_microbatch_size < 1\n or args.early_stopping_patience < 1\n or args.early_stopping_min_delta < 0\n or args.ordinary_target_weight <= 0\n or args.organic_strong_target_weight <= 0\n or args.organic_like_target_weight <= 0\n or args.discovery_horizon_days < 1\n or args.max_train_targets_per_user < 0\n or args.max_validation_targets_per_user < 0\n ):\n parser.error(\"training counts, learning rate, and patience are invalid\")\n if not 0.0 < args.validation_fraction < 1.0:\n parser.error(\"validation fraction must be strictly between zero and one\")\n public = args.public_dir.resolve()\n if not (public / \"data\" / \"candidate_catalog.parquet\").is_file():\n parser.error(f\"public task data not found under {public}\")\n config = Config(\n bptt_steps=args.bptt_steps,\n user_batch_size=args.user_batch_size,\n epochs=args.epochs,\n learning_rate=args.learning_rate,\n loss_microbatch_size=args.loss_microbatch_size,\n negative_filter_scope=args.negative_filter_scope,\n objective=args.objective,\n temperature=args.temperature,\n target_weighting=args.target_weighting,\n ordinary_target_weight=args.ordinary_target_weight,\n organic_strong_target_weight=args.organic_strong_target_weight,\n organic_like_target_weight=args.organic_like_target_weight,\n discovery_horizon_days=args.discovery_horizon_days,\n validation_fraction=args.validation_fraction,\n max_train_targets_per_user=args.max_train_targets_per_user,\n max_validation_targets_per_user=args.max_validation_targets_per_user,\n early_stopping_patience=args.early_stopping_patience,\n early_stopping_min_delta=args.early_stopping_min_delta,\n seed=args.seed,\n )\n torch.set_num_threads(args.threads)\n torch.set_num_interop_threads(1)\n seed_everything(config.seed)\n device = torch.device(args.device)\n if device.type == \"mps\" and not torch.backends.mps.is_available():\n parser.error(\"MPS was requested but is unavailable\")\n\n started = time.monotonic()\n prepared, eligibility = load_or_prepare(\n public,\n args.stage,\n config,\n args.prepared_cache.resolve() if args.prepared_cache else None,\n )\n target_partition = chronological_target_partition(\n prepared,\n config.validation_fraction,\n config.max_train_targets_per_user,\n config.max_validation_targets_per_user,\n )\n if args.prepare_only:\n print(\n json.dumps(\n {\n \"stage\": args.stage,\n \"sequence_users\": len(prepared.user_ids),\n \"sequence_tokens\": len(prepared.sequence_items),\n \"next_item_targets\": prepared.target_count,\n \"selected_training_targets\": int(\n target_partition.training_counts.sum()\n ),\n \"selected_validation_targets\": int(\n target_partition.validation_counts.sum()\n ),\n \"discarded_old_targets\": int(\n target_partition.window_starts.sum()\n ),\n \"target_users\": len(eligibility.uids),\n \"runtime_seconds\": round(time.monotonic() - started, 3),\n \"config\": asdict(config),\n },\n indent=2,\n )\n )\n return 0\n\n model, training_history = train_model(\n public,\n args.stage,\n prepared,\n config,\n device,\n args.checkpoint_dir.resolve() if args.checkpoint_dir else None,\n args.resume_checkpoint.resolve() if args.resume_checkpoint else None,\n )\n ranking, embedding_diagnostics = rank_candidates(\n model,\n prepared,\n eligibility,\n config,\n device,\n args.ranking_batch_size,\n )\n write_ranking(ranking, args.output.resolve())\n print(\n json.dumps(\n {\n \"baseline\": \"yambda-windowed-sequence-two-tower-base\",\n \"stage\": args.stage,\n \"output\": str(args.output.resolve()),\n \"rows\": len(ranking),\n \"users\": int(ranking.uid.nunique()),\n \"runtime_seconds\": round(time.monotonic() - started, 3),\n \"public_inputs_only\": True,\n \"architecture\": {\n \"family\": \"pure-windowed-sequence-two-tower\",\n \"sequence_training\": (\n \"every-adjacent-target-within-recent-chronological-window\"\n ),\n \"sequence_encoder\": \"one-layer-GRU-dense-timestep-loss\",\n \"bptt_hidden_state\": \"carried-and-detached\",\n \"model_selection\": \"chronological-validation-BCE-early-stopping\",\n \"loss_execution\": \"gradient-accumulated-microbatches\",\n \"user_id_embedding\": False,\n \"user_mlp\": \"582-512-256-192\",\n \"item_mlp\": \"196-512-256-192\",\n \"normalization\": \"none\",\n \"score\": \"raw-dot-product\",\n \"objective\": \"sample-weighted-BCE-at-every-timestep\",\n \"negative_sampling\": (\n \"24-uniform-plus-8-propensity-corrected-inbatch\"\n ),\n \"false_negative_filter\": config.negative_filter_scope,\n \"target_weighting\": config.target_weighting,\n },\n \"config\": asdict(config),\n \"sequence_tokens\": len(prepared.sequence_items),\n \"all_next_item_targets\": prepared.target_count,\n \"selected_training_targets\": int(\n target_partition.training_counts.sum()\n ),\n \"selected_validation_targets\": int(\n target_partition.validation_counts.sum()\n ),\n \"discarded_old_targets\": int(target_partition.window_starts.sum()),\n \"training_history\": training_history,\n \"embedding_diagnostics\": embedding_diagnostics,\n \"torch_version\": torch.__version__,\n \"device\": str(device),\n },\n indent=2,\n )\n )\n return 0\n\n\nif __name__ == \"__main__\":\n raise SystemExit(main())\n",
"diffStat": {
"model_added_lines": 2,
"model_removed_lines": 0,
"model_added_chars": 70,
"model_removed_chars": 0,
"user_added_lines": 0,
"user_removed_lines": 0,
"user_added_chars": 0,
"user_removed_chars": 0
},
"isNewFile": false,
"isBuildFile": false
},
"description": "solution/train_current_model.py: negative_filter_scope=... => negative_filter_scope=...",
"displayName": "Edit",
"renderOutputAsMarkdown": true
},
"result": {
"timestamp": "2026-09-29T06:07:17.253Z",
"content": "Successfully modified file: /app/solution/train_current_model.py (1 replacements). Here is the updated code:\n...\n user_batch_size=args.user_batch_size,\n epochs=args.epochs,\n learning_rate=args.learning_rate,\n loss_microbatch_size=args.loss_microbatch_size,\n negative_filter_scope=args.negative_filter_scope,\n objective=args.objective,\n temperature=args.temperature,\n target_weighting=args.target_weighting,\n ordinary_target_weight=args.ordinary_target_weight,\n organic_strong_target_weight=args.organic_strong_target_weight,\n organic_like_target_weight=args.organic_like_target_weight,\n discovery_horizon_days=args.discovery_horizon_days,\n...",
"isError": false
},
"responseMessageId": "f1ddf3fa-5440-44d3-8d6d-290ae0b66ec9"
}
]
}
],
"source": {
"format": "gemini-cli-session",
"sha256": "b7dfa438085b20765b12a086f362987b46a4d756a5ab6b196f22d0fb85a0e843",
"records": 254,
"sessionMetadata": {
"sessionId": "333a7290-91e9-45e9-97c9-d83fb9219278",
"projectHash": "f53b52ad6d21cceb72dfa78fb67614fe14f110c58e68412b01508d6a485501c3",
"startTime": "2026-09-29T05:36:41.212Z",
"lastUpdated": "2026-09-29T05:36:41.212Z",
"kind": "main"
},
"sessionUpdates": 102,
"usageRecords": [
{
"messageId": "09e41ada-9955-4b68-b1be-3afd95fd04a1",
"timestamp": "2026-09-29T05:36:47.736Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 10563,
"output": 260,
"cached": 7676,
"thoughts": 570,
"tool": 0,
"total": 11393
}
},
{
"messageId": "90712173-ba1c-4e8d-82de-6e805a958c56",
"timestamp": "2026-09-29T05:36:50.225Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 17637,
"output": 34,
"cached": 7825,
"thoughts": 57,
"tool": 0,
"total": 17728
}
},
{
"messageId": "8a6c721f-1531-4a9a-82e1-a7f5231bf1a9",
"timestamp": "2026-09-29T05:36:52.365Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 17781,
"output": 81,
"cached": 15633,
"thoughts": 18,
"tool": 0,
"total": 17880
}
},
{
"messageId": "9c77f518-89dc-4899-b3f1-62150a56ea22",
"timestamp": "2026-09-29T05:37:05.566Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 17945,
"output": 79,
"cached": 15618,
"thoughts": 11,
"tool": 0,
"total": 18035
}
},
{
"messageId": "e243fa9e-ddc5-4c01-9fe7-250b82877e35",
"timestamp": "2026-09-29T05:37:11.911Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 18107,
"output": 76,
"cached": 15604,
"thoughts": 195,
"tool": 0,
"total": 18378
}
},
{
"messageId": "d83f00aa-d5fc-4b45-8bac-228094fd1e67",
"timestamp": "2026-09-29T05:43:23.430Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 20644,
"output": 45,
"cached": 0,
"thoughts": 19,
"tool": 0,
"total": 20708
}
},
{
"messageId": "5e4f4dbb-2a7a-4dd8-9774-220f2dd4a4a9",
"timestamp": "2026-09-29T05:43:37.216Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 20794,
"output": 47,
"cached": 15663,
"thoughts": 138,
"tool": 0,
"total": 20979
}
},
{
"messageId": "abf6fa8d-765d-4bbc-88f3-2f8a7aa3916a",
"timestamp": "2026-09-29T05:43:38.999Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 22832,
"output": 47,
"cached": 19634,
"thoughts": 46,
"tool": 0,
"total": 22925
}
},
{
"messageId": "810ea726-d769-402f-8b63-69f75682f202",
"timestamp": "2026-09-29T05:43:41.316Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 26367,
"output": 49,
"cached": 19725,
"thoughts": 90,
"tool": 0,
"total": 26506
}
},
{
"messageId": "5fc7bd0a-fb87-4951-8b77-fad932c2429e",
"timestamp": "2026-09-29T05:43:45.055Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 29513,
"output": 29,
"cached": 23743,
"thoughts": 310,
"tool": 0,
"total": 29852
}
},
{
"messageId": "429c9cad-d91c-4b0d-be8d-59286a9c8861",
"timestamp": "2026-09-29T05:43:52.912Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 31828,
"output": 47,
"cached": 27747,
"thoughts": 63,
"tool": 0,
"total": 31938
}
},
{
"messageId": "64ffd0ef-7c3a-4cfb-ad65-0ba910cade0c",
"timestamp": "2026-09-29T05:43:57.248Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 33560,
"output": 47,
"cached": 27772,
"thoughts": 36,
"tool": 0,
"total": 33643
}
},
{
"messageId": "bb1263ac-0bdf-4365-b668-8340ad362c13",
"timestamp": "2026-09-29T05:43:59.114Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 35334,
"output": 48,
"cached": 31767,
"thoughts": 25,
"tool": 0,
"total": 35407
}
},
{
"messageId": "df186b27-aae1-4f78-bd25-a6d8148832a8",
"timestamp": "2026-09-29T05:44:00.897Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 38686,
"output": 49,
"cached": 31830,
"thoughts": 26,
"tool": 0,
"total": 38761
}
},
{
"messageId": "5576d1b0-e626-43dc-be1b-467109c6d0c9",
"timestamp": "2026-09-29T05:44:02.677Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 40395,
"output": 49,
"cached": 35830,
"thoughts": 26,
"tool": 0,
"total": 40470
}
},
{
"messageId": "a3216f52-ab31-4597-b95f-61fab519345c",
"timestamp": "2026-09-29T05:44:05.153Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 42380,
"output": 49,
"cached": 39840,
"thoughts": 39,
"tool": 0,
"total": 42468
}
},
{
"messageId": "79e75b46-9500-40ae-9c86-13daa92b51de",
"timestamp": "2026-09-29T05:44:09.815Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 44312,
"output": 47,
"cached": 39865,
"thoughts": 54,
"tool": 0,
"total": 44413
}
},
{
"messageId": "f4fd04d7-3853-44be-8287-f8776b7554e0",
"timestamp": "2026-09-29T05:44:11.496Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 46934,
"output": 47,
"cached": 43893,
"thoughts": 27,
"tool": 0,
"total": 47008
}
},
{
"messageId": "8146ef15-c8e3-4bc9-b530-f3482315db58",
"timestamp": "2026-09-29T05:44:13.667Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 47976,
"output": 41,
"cached": 43894,
"thoughts": 61,
"tool": 0,
"total": 48078
}
},
{
"messageId": "7bb28677-8442-4c4f-983c-6237ed75bda2",
"timestamp": "2026-09-29T05:44:15.576Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 48514,
"output": 41,
"cached": 43884,
"thoughts": 21,
"tool": 0,
"total": 48576
}
},
{
"messageId": "36086108-3cf3-457b-b415-97046481eef0",
"timestamp": "2026-09-29T05:44:22.391Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 49657,
"output": 79,
"cached": 43888,
"thoughts": 863,
"tool": 0,
"total": 50599
}
},
{
"messageId": "69540163-badd-4f38-91e6-2e4eaa67d17f",
"timestamp": "2026-09-29T05:45:49.209Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 51631,
"output": 45,
"cached": 47902,
"thoughts": 46,
"tool": 0,
"total": 51722
}
},
{
"messageId": "5fa53942-315c-47f1-b58b-f5d9a6edf28b",
"timestamp": "2026-09-29T05:46:04.342Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 51785,
"output": 74,
"cached": 47882,
"thoughts": 57,
"tool": 0,
"total": 51916
}
},
{
"messageId": "581c3484-ab55-43f5-8b96-7c4441ba283f",
"timestamp": "2026-09-29T05:46:10.404Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 51945,
"output": 79,
"cached": 47862,
"thoughts": 44,
"tool": 0,
"total": 52068
}
},
{
"messageId": "faf020d1-77e9-413f-8426-2f95ef2b4209",
"timestamp": "2026-09-29T05:47:29.748Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 53101,
"output": 30,
"cached": 47867,
"thoughts": 3223,
"tool": 0,
"total": 56354
}
},
{
"messageId": "cc603fad-270f-480f-8a18-62f81179bdb1",
"timestamp": "2026-09-29T05:48:06.683Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 57206,
"output": 234,
"cached": 51930,
"thoughts": 5431,
"tool": 0,
"total": 62871
}
},
{
"messageId": "d2a7c79e-2090-4372-aa60-926f68284bd0",
"timestamp": "2026-09-29T05:48:11.954Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 63137,
"output": 254,
"cached": 56033,
"thoughts": 99,
"tool": 0,
"total": 63490
}
},
{
"messageId": "275ec778-c5a6-496b-bc49-769635a6061e",
"timestamp": "2026-09-29T05:48:18.210Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 64491,
"output": 312,
"cached": 60040,
"thoughts": 186,
"tool": 0,
"total": 64989
}
},
{
"messageId": "f873825e-d21a-4f04-a789-bdb86f77fe53",
"timestamp": "2026-09-29T05:48:33.370Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 65055,
"output": 324,
"cached": 60029,
"thoughts": 14,
"tool": 0,
"total": 65393
}
},
{
"messageId": "d3e60eca-e003-42ff-a511-bbcd4712cbd4",
"timestamp": "2026-09-29T05:49:55.026Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 66178,
"output": 105,
"cached": 64031,
"thoughts": 10751,
"tool": 0,
"total": 77034
}
},
{
"messageId": "97db45f0-5b7e-4b72-af8d-9c0d9290dcad",
"timestamp": "2026-09-29T05:55:25.338Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 78975,
"output": 49,
"cached": 8031,
"thoughts": 15,
"tool": 0,
"total": 79039
}
},
{
"messageId": "a5f4f044-4a6e-49e9-89ca-c7283ef55d26",
"timestamp": "2026-09-29T05:55:32.368Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 79128,
"output": 141,
"cached": 76274,
"thoughts": 64,
"tool": 0,
"total": 79333
}
},
{
"messageId": "db6e65a9-ccba-49ff-9a15-42658546c0fb",
"timestamp": "2026-09-29T05:56:12.119Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 79453,
"output": 375,
"cached": 76252,
"thoughts": 5283,
"tool": 0,
"total": 85111
}
},
{
"messageId": "ab60a495-88b5-457d-a1bd-e9c3512952d3",
"timestamp": "2026-09-29T06:01:51.560Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 85191,
"output": 149,
"cached": 80351,
"thoughts": 3997,
"tool": 0,
"total": 89337
}
},
{
"messageId": "22d7db28-3b09-4463-be02-46ee6815c7da",
"timestamp": "2026-09-29T06:05:05.286Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 89428,
"output": 134,
"cached": 8040,
"thoughts": 247,
"tool": 0,
"total": 89809
}
},
{
"messageId": "58f3a799-b347-4ed2-9e1c-f6c79867e2b0",
"timestamp": "2026-09-29T06:05:24.884Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 89893,
"output": 47,
"cached": 84405,
"thoughts": 1654,
"tool": 0,
"total": 91594
}
},
{
"messageId": "25c73690-1445-40ef-a7de-087af3b01753",
"timestamp": "2026-09-29T06:05:28.416Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 92654,
"output": 43,
"cached": 88450,
"thoughts": 35,
"tool": 0,
"total": 92732
}
},
{
"messageId": "b65f88d1-21c1-4785-84d6-30fd9e3f9358",
"timestamp": "2026-09-29T06:05:38.857Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 93287,
"output": 47,
"cached": 88436,
"thoughts": 881,
"tool": 0,
"total": 94215
}
},
{
"messageId": "e68fb99d-a519-4820-a645-86e74dbdb82b",
"timestamp": "2026-09-29T06:05:45.641Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 94924,
"output": 159,
"cached": 92461,
"thoughts": 604,
"tool": 0,
"total": 95687
}
},
{
"messageId": "ee028d17-83e2-45e5-8708-2fa00535de33",
"timestamp": "2026-09-29T06:06:03.546Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 95776,
"output": 139,
"cached": 92452,
"thoughts": 1561,
"tool": 0,
"total": 97476
}
},
{
"messageId": "cfd59a3d-06cd-4419-a808-511a9c532c76",
"timestamp": "2026-09-29T06:06:39.409Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 97482,
"output": 37,
"cached": 92448,
"thoughts": 41,
"tool": 0,
"total": 97560
}
},
{
"messageId": "3d21d4c0-0834-4faa-895f-ea753fbd9cdb",
"timestamp": "2026-09-29T06:06:43.315Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 97813,
"output": 42,
"cached": 92430,
"thoughts": 38,
"tool": 0,
"total": 97893
}
},
{
"messageId": "e1bb5a7c-c9ff-46f9-a696-1a70b6a3e342",
"timestamp": "2026-09-29T06:06:46.031Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 97922,
"output": 139,
"cached": 92408,
"thoughts": 11,
"tool": 0,
"total": 98072
}
},
{
"messageId": "a3e0d64f-46af-4b46-9eb9-bad27d1b6568",
"timestamp": "2026-09-29T06:06:50.731Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 98255,
"output": 47,
"cached": 92381,
"thoughts": 66,
"tool": 0,
"total": 98368
}
},
{
"messageId": "6943b2c5-805b-427d-8f8b-68ff104b4749",
"timestamp": "2026-09-29T06:06:57.247Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 98816,
"output": 523,
"cached": 96383,
"thoughts": 17,
"tool": 0,
"total": 99356
}
},
{
"messageId": "6ac5f531-afe3-4a12-a3ad-bd30bf83fd26",
"timestamp": "2026-09-29T06:07:01.210Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 99722,
"output": 47,
"cached": 96366,
"thoughts": 46,
"tool": 0,
"total": 99815
}
},
{
"messageId": "5d35ca80-babf-4b68-90d9-3378836e3aa9",
"timestamp": "2026-09-29T06:07:04.043Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 101212,
"output": 47,
"cached": 96370,
"thoughts": 29,
"tool": 0,
"total": 101288
}
},
{
"messageId": "d2790dae-cf7c-4818-aa72-26a0eeff893c",
"timestamp": "2026-09-29T06:07:10.132Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 101735,
"output": 312,
"cached": 96355,
"thoughts": 20,
"tool": 0,
"total": 102067
}
},
{
"messageId": "54ecb56a-0237-458d-9700-b605f704ef53",
"timestamp": "2026-09-29T06:07:13.096Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 102242,
"output": 47,
"cached": 96331,
"thoughts": 36,
"tool": 0,
"total": 102325
}
},
{
"messageId": "7c28f774-1f4e-4132-add3-33d6ce1ff294",
"timestamp": "2026-09-29T06:07:17.238Z",
"model": "gemini-3.8-flash",
"tokens": {
"input": 102882,
"output": 133,
"cached": 100333,
"thoughts": 22,
"tool": 0,
"total": 103037
}
}
]
}
}