entity_tokenized / README.md
darshanmakwana's picture
Upload dataset
00312dc verified
|
raw
history blame
529 Bytes
metadata
dataset_info:
  features:
    - name: audio_tokens
      sequence: int64
    - name: raw_text
      dtype: string
    - name: transcript
      dtype: string
    - name: entities
      sequence: string
    - name: prompt
      dtype: string
    - name: input_ids
      sequence: int64
    - name: attention_mask
      sequence: int64
  splits:
    - name: train
      num_bytes: 2895627778
      num_examples: 70397
  download_size: 184378881
  dataset_size: 2895627778
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*