{ "description": "Per-length CP data roots for DataLoaderLite. Each bucket directory is a self-contained, flat CP_DATA_ROOT with its own train and val shards.", "encoding": "gpt2 (tiktoken)", "dtype": "uint16", "eot_token": 50256, "eot_appended_per_document": true, "dataloader_max_tokens_per_shard": 100000000, "hf_repo": "unparallelled/BluScriptCP-data", "hf_prefix": "cp_data_by_length", "buckets": { "b0_lt_4096": { "token_range": [ 0, 4096 ], "train_shards": 736, "val_shards": 1, "tokens": 73663368206, "max_shard_tokens": 99953973, "within_dataloader_cap": true, "bytes": 147326736412, "published_to_hf": false }, "b1_4096_8192": { "token_range": [ 4096, 8192 ], "train_shards": 139, "val_shards": 1, "tokens": 13980346665, "max_shard_tokens": 99866141, "within_dataloader_cap": true, "bytes": 27960693330, "published_to_hf": true }, "b2_8192_16384": { "token_range": [ 8192, 16384 ], "train_shards": 69, "val_shards": 1, "tokens": 6991372853, "max_shard_tokens": 99889424, "within_dataloader_cap": true, "bytes": 13982745706, "published_to_hf": true }, "b3_16384_32768": { "token_range": [ 16384, 32768 ], "train_shards": 30, "val_shards": 1, "tokens": 3028006031, "max_shard_tokens": 97701502, "within_dataloader_cap": true, "bytes": 6056012062, "published_to_hf": true }, "b4_32768_65536": { "token_range": [ 32768, 65536 ], "train_shards": 12, "val_shards": 1, "tokens": 1269502884, "max_shard_tokens": 97692527, "within_dataloader_cap": true, "bytes": 2539005768, "published_to_hf": true }, "b5_65536_131072": { "token_range": [ 65536, 131072 ], "train_shards": 9, "val_shards": 1, "tokens": 962072457, "max_shard_tokens": 96334986, "within_dataloader_cap": true, "bytes": 1924144914, "published_to_hf": true }, "b6_gte_131072": { "token_range": [ 131072, null ], "train_shards": 1, "val_shards": 1, "tokens": 114467449, "max_shard_tokens": 57334448, "within_dataloader_cap": true, "bytes": 228934898, "published_to_hf": true } }, "totals": { "buckets": 7, "train_shards": 996, "val_shards": 7, "tokens": 100009136545 } }