qa-formatting-data / reports /build_report.json
nzhao721's picture
Add the OpenStax QA-formatting data
6d1b0e7 verified
Raw History Blame Contribute Delete
1.91 kB
{
"artifact_dir": "data/stage2_educational/artifacts/openstax_stage2_v4_llm_augmented_k4_no_pack_300m",
"claim_count": 5886,
"conditions": {
"q0_normal_fact": {
"documents": 23544,
"target_tokens": 530692
},
"q1_diverse_prose": {
"documents": 23544,
"target_tokens": 585401
},
"q3_diverse_qa": {
"documents": 23544,
"target_tokens": 623510
}
},
"decision_boundary": "This release tests Stage 2 LLM-generated diverse-prose versus LLM-generated diverse-QA augmentation. Held-out questions remain sealed and are not used as training variants.",
"exact_probe_config_path": "data/stage2_educational/config/stage2_openstax_stage2_v4_llm_augmented_k4_no_pack_300m_exact_probe_jobs.jsonl",
"exact_training_evaluation_count": 23544,
"heldout_evaluation_count": 11774,
"manifest_dir": "data/stage2_educational/manifests/openstax_stage2_v4_llm_augmented_k4_no_pack_300m/fixed_token",
"max_claims": null,
"output_stem": "openstax_stage2_v4_llm_augmented_k4_no_pack_300m",
"rendered_document_count": 70632,
"renderer_version": "stage2-v4-llm-rewrite-k4-v1",
"rewrite_path": "data/stage2_educational/llm_rewrites/openstax_stage2_v4_llm_augmented_k4_gpt54mini_full_20260725/accepted_rewrites.jsonl",
"run_date": "20260725",
"schema_version": "stage2-v4-llm-rewrite-build-v1",
"source_count": 1962,
"tokenizer_backend": {
"name": "transformers_auto_tokenizer",
"tokenizers_version": "0.21.4",
"version": "4.48.3"
},
"tokenizer_identifier": "allenai/OLMo-Ladder-760M-0.5xC@c8f6875e640a985de0709e8be1f38eef0cb4bc48",
"training_conditions": [
"q0_normal_fact",
"q1_diverse_prose",
"q3_diverse_qa"
],
"training_config_path": "data/stage2_educational/config/stage2_openstax_stage2_v4_llm_augmented_k4_no_pack_300m_jobs.jsonl",
"variants_per_claim": 4
}