Datasets:
Download dataset_info.json from mlfoundations/dcvlm-balanced-200b: direct link, hf CLI and curl.
- Browser
- Download file 641 Bytes
-
https://huggingface.co/datasets/mlfoundations/dcvlm-balanced-200b/resolve/main/dataset_info.json
- Command line
-
hf download hf://datasets/mlfoundations/dcvlm-balanced-200b/dataset_info.json
-
curl -L -o dataset_info.json https://huggingface.co/datasets/mlfoundations/dcvlm-balanced-200b/resolve/main/dataset_info.json
641 Bytes
| { | |
| "name": "dcvlm_balanced_flat_shuffled", | |
| "seed": 40, | |
| "world_size": 512, | |
| "num_shards": 11318, | |
| "num_samples": 112358849, | |
| "num_samples_skipped": 0, | |
| "total_length_tokens": 200000681349.0, | |
| "per_data_type_counts": { | |
| "captioning": 44967501, | |
| "text": 16856093, | |
| "instruction": 44961975, | |
| "mmdoc_mt": 5573280 | |
| }, | |
| "per_data_type_fraction": { | |
| "captioning": 0.4002132577915603, | |
| "text": 0.15002016441090457, | |
| "instruction": 0.4001640760844747, | |
| "mmdoc_mt": 0.049602501713060444 | |
| }, | |
| "globally_shuffled": true, | |
| "reshuffle_num_buckets": 512, | |
| "reshuffle_seed": 40, | |
| "reshuffle_shuffle_bufsize": 20000 | |
| } | |