| { | |
| "dataset": "lm1b", | |
| "evaluation_continuation_length": 64, | |
| "model_length": 128, | |
| "prefix_length": 64, | |
| "prompt_count": 1024, | |
| "prompt_file": "data/conditional/lm1b-c64/prompts.jsonl", | |
| "prompt_file_sha256": "b83d6fc574efda562fce68a23bffe106b238d1739a656dade72fda4ab18bc422", | |
| "protocol": "c64_zs_v1", | |
| "schema_version": 1, | |
| "selection_algorithm": "sha256_seed_index_digest_ascending_v1", | |
| "selection_seed": 42, | |
| "source_manifest_path": "data/manifests/lm1b.json", | |
| "source_manifest_sha256": "a8465f3fba8c5dc529fc0751039c8ffa32a7710f480208779b55545cb141017c", | |
| "source_processed_path": "data/processed/lm1b-bert-128/validation", | |
| "source_row_count": 71979, | |
| "source_split": "validation", | |
| "tokenizer_id": "bert-base-uncased", | |
| "tokenizer_revision": "86b5e0934494bd15c9632b12f734a8a67f723594", | |
| "vocabulary_size": 30522 | |
| } | |