stukenov/sozkz-corpus-tokenized-kk-llama50k-v4

--- dataset_info: features: - name: input_ids sequence: int32 - name: token_type_ids sequence: int8 - name: labels sequence: int64 splits: - name: train num_bytes: 6175594056.0 num_examples: 463494 - name: validation num_bytes: 325038980.0 num_examples: 24395 download_size: 2158067242 dataset_size: 6500633036.0 configs: - config_name: default data_files: - split: train path: data/train-* - split: validation path: data/validation-* ---