Download metadata.json from SuperGoatScriptGuy/PreprocessedMIXED: direct link, hf CLI and curl.
- Browser
- Download file 377 Bytes
-
https://huggingface.co/datasets/SuperGoatScriptGuy/PreprocessedMIXED/resolve/main/metadata.json
- Command line
-
hf download hf://datasets/SuperGoatScriptGuy/PreprocessedMIXED/metadata.json
-
curl -L -o metadata.json https://huggingface.co/datasets/SuperGoatScriptGuy/PreprocessedMIXED/resolve/main/metadata.json
377 Bytes
| { | |
| "num_sequences": 26840062, | |
| "num_shards": 2685, | |
| "mix_ratios": { | |
| "fineweb": 0.7, | |
| "wikipedia": 0.15, | |
| "code": 0.15 | |
| }, | |
| "sequences_per_dataset": { | |
| "fineweb": 18788044, | |
| "wikipedia": 4026009, | |
| "code": 4026009 | |
| }, | |
| "source_datasets": { | |
| "fineweb": 2392, | |
| "wikipedia": 230, | |
| "code": 63 | |
| }, | |
| "total_tokens": 54968446976 | |
| } |