Aug 13 17:02 Clarify MTS-Dialog fact coverage 16:22 Document current reproducible training inputs Aug 11 06:41 Upload dpo_sft_r128/train (5312 rows) 06:41 Upload kto_sft_r128/train (10624 rows) 06:41 Upload dpo_sft/train (5312 rows) 06:41 Upload kto_sft/train (10624 rows) 06:40 Remove dpo_sft/test2_dpo_chat_format.jsonl (archived) 06:40 Archive dpo_sft/test2_dpo_chat_format.jsonl 06:40 Remove dpo_sft/test1_dpo_chat_format.jsonl (archived) 06:40 Archive dpo_sft/test1_dpo_chat_format.jsonl 06:40 Remove dpo_sft/validation_dpo_chat_format.jsonl (archived) 06:40 Archive dpo_sft/validation_dpo_chat_format.jsonl 06:40 Remove dpo_sft/train_dpo_chat_format.jsonl (archived) 06:40 Archive dpo_sft/train_dpo_chat_format.jsonl 06:40 Remove kto_sft/test2_kto_chat_format.jsonl (archived) 06:40 Archive kto_sft/test2_kto_chat_format.jsonl 06:40 Remove kto_sft/test1_kto_chat_format.jsonl (archived) 06:40 Archive kto_sft/test1_kto_chat_format.jsonl 06:40 Remove kto_sft/validation_kto_chat_format.jsonl (archived) 06:40 Archive kto_sft/validation_kto_chat_format.jsonl 06:40 Remove kto_sft/train_kto_chat_format.jsonl (archived) 06:40 Archive kto_sft/train_kto_chat_format.jsonl Aug 10 23:14 Approved dialogues list (1700 use, 1 excluded) 22:59 Remove manually extracted facts for dialogue 1115. Will use approved dialogues list instead. 22:42 Add manually extracted facts for dialogue 1115. Dialogue 1115 (a 9000-char multi-doctor consultation) could not be extracted by GLM-5.2 after all retry rounds. Facts were manually extracted from the MTS-Dialog gold section_text (doctor-written medical summary), following the same fact format as GLM-5.2. With this, all 1201 training dialogues have facts. No downstream script needs to handle missing facts. 12:08 Remove old dpo_sft1/test2_dpo_chat_format.jsonl 12:08 Move dpo_sft1/test2_dpo_chat_format.jsonl → archive/dpo_sft1/test2_dpo_chat_format.jsonl 12:08 Remove old dpo_sft1/test1_dpo_chat_format.jsonl 12:08 Move dpo_sft1/test1_dpo_chat_format.jsonl → archive/dpo_sft1/test1_dpo_chat_format.jsonl 12:08 Remove old dpo_sft1/validation_dpo_chat_format.jsonl 12:08 Move dpo_sft1/validation_dpo_chat_format.jsonl → archive/dpo_sft1/validation_dpo_chat_format.jsonl 12:08 Remove old dpo_sft1/train_dpo_chat_format.jsonl 12:08 Move dpo_sft1/train_dpo_chat_format.jsonl → archive/dpo_sft1/train_dpo_chat_format.jsonl 12:08 Remove old kto_sft1/eval/train_negatives_geval_summary.json 12:08 Move kto_sft1/eval/train_negatives_geval_summary.json → archive/kto_sft1/eval/train_negatives_geval_summary.json 12:08 Remove old kto_sft1/eval/train_negatives_geval.jsonl 12:08 Move kto_sft1/eval/train_negatives_geval.jsonl → archive/kto_sft1/eval/train_negatives_geval.jsonl 12:08 Remove old kto_sft1/eval/train_negatives_candidates.jsonl 12:08 Move kto_sft1/eval/train_negatives_candidates.jsonl → archive/kto_sft1/eval/train_negatives_candidates.jsonl 12:08 Remove old kto_sft1/test2_kto_chat_format.jsonl 12:08 Move kto_sft1/test2_kto_chat_format.jsonl → archive/kto_sft1/test2_kto_chat_format.jsonl 12:08 Remove old kto_sft1/test1_kto_chat_format.jsonl 12:08 Move kto_sft1/test1_kto_chat_format.jsonl → archive/kto_sft1/test1_kto_chat_format.jsonl 12:08 Remove old kto_sft1/validation_kto_chat_format.jsonl 12:08 Move kto_sft1/validation_kto_chat_format.jsonl → archive/kto_sft1/validation_kto_chat_format.jsonl 12:08 Remove old kto_sft1/train_kto_chat_format.jsonl 12:08 Move kto_sft1/train_kto_chat_format.jsonl → archive/kto_sft1/train_kto_chat_format.jsonl 12:08 Remove old dpo_base/test2_dpo_chat_format.jsonl 12:08 Move dpo_base/test2_dpo_chat_format.jsonl → archive/dpo_base/test2_dpo_chat_format.jsonl 12:08 Remove old dpo_base/test1_dpo_chat_format.jsonl 
