| { | |
| "dataset": "owt", | |
| "evaluation_continuation_length": 64, | |
| "model_length": 1024, | |
| "prefix_length": 64, | |
| "prompt_count": 1024, | |
| "prompt_file": "data/conditional/owt-c64/prompts.jsonl", | |
| "prompt_file_sha256": "68907c7d4eb5104dbaea6d136a951635a11eb46c0f3e8f417d5f406e1ca834fc", | |
| "protocol": "c64_zs_v1", | |
| "schema_version": 1, | |
| "selection_algorithm": "sha256_seed_index_digest_ascending_v1", | |
| "selection_seed": 42, | |
| "source_manifest_path": "data/manifests/owt.json", | |
| "source_manifest_sha256": "6971545df48d6fac81297fee0bdfaa9f1eb937501288b8f9a3a9582beb3b2916", | |
| "source_processed_path": "data/processed/owt-gpt2-1024/validation", | |
| "source_row_count": 110520, | |
| "source_split": "validation", | |
| "tokenizer_id": "gpt2", | |
| "tokenizer_revision": "607a30d783dfa663caf39e06633721c8d4cfcd7e", | |
| "vocabulary_size": 50257 | |
| } | |