train/mlfoundations-dclm-baseline-1.0-parquet-sampled-v4/split_31885-tokenized-chunked-1024-512-128-backfill-nodups/index.json

{"shards": [{"column_encodings": ["str", "ndarray:uint16"], "column_names": ["id", "input_ids"], "column_sizes": [null, null], "compression": "zstd", "format": "mds", "hashes": [], "raw_data": {"basename": "shard.00000.mds", "bytes": 67107995, "hashes": {}}, "samples": 42902, "size_limit": 67108864, "version": 2, "zip_data": {"basename": "shard.00000.mds.zstd", "bytes": 48012035, "hashes": {}}}, {"column_encodings": ["str", "ndarray:uint16"], "column_names": ["id", "input_ids"], "column_sizes": [null, null], "compression": "zstd", "format": "mds", "hashes": [], "raw_data": {"basename": "shard.00001.mds", "bytes": 19733189, "hashes": {}}, "samples": 12483, "size_limit": 67108864, "version": 2, "zip_data": {"basename": "shard.00001.mds.zstd", "bytes": 14114506, "hashes": {}}}], "version": 2}