{ "experiment": "raw_baseline", "method": "raw", "base_dataset_repo": "Bykot/c4_ru_200k_split", "source_train_docs": 198000, "cleaned_docs": 198000, "removed_docs": 0, "removed_percent": 0.0, "eval_docs": 2000, "created_at_utc": "2026-05-05T02:43:19.833028+00:00" }