jan-hq's picture
Upload dataset
fcc5a0c verified
metadata
dataset_info:
  - config_name: VTSNLP-instruct-filtered
    features:
      - name: text_prompt
        dtype: string
      - name: answer
        dtype: string
      - name: compressed_prompt
        dtype: string
      - name: conversations
        list:
          - name: content
            dtype: string
          - name: role
            dtype: string
    splits:
      - name: train
        num_bytes: 932086063.6919165
        num_examples: 253940
    download_size: 252682325
    dataset_size: 932086063.6919165
  - config_name: english-multiturn
    features:
      - name: text_conversations
        list:
          - name: content
            dtype: string
          - name: role
            dtype: string
      - name: index
        dtype: int64
      - name: conversations
        list:
          - name: content
            dtype: string
          - name: role
            dtype: string
    splits:
      - name: train
        num_bytes: 2813901490.5094438
        num_examples: 158261
    download_size: 996062852
    dataset_size: 2813901490.5094438
  - config_name: instruction-speech-v1
    features:
      - name: index
        dtype: int64
      - name: text_prompt
        dtype: string
      - name: answer
        dtype: string
      - name: length
        dtype: int64
      - name: compressed_prompt
        dtype: string
      - name: conversations
        list:
          - name: content
            dtype: string
          - name: role
            dtype: string
    splits:
      - name: train
        num_bytes: 6279403560.552255
        num_examples: 677289
    download_size: 1867497688
    dataset_size: 6279403560.552255
  - config_name: instruction-speech-v1-rephrase
    features:
      - name: index
        dtype: int64
      - name: text_prompt
        dtype: string
      - name: original_answer
        dtype: string
      - name: rephrased_answer
        dtype: string
      - name: difference_level
        dtype: string
      - name: compressed_prompt
        dtype: string
      - name: conversations
        list:
          - name: content
            dtype: string
          - name: role
            dtype: string
    splits:
      - name: train
        num_bytes: 207820285.4091829
        num_examples: 36479
    download_size: 69530680
    dataset_size: 207820285.4091829
  - config_name: instruction-speech-v2
    features:
      - name: index
        dtype: int64
      - name: text_prompt
        dtype: string
      - name: answer
        dtype: string
      - name: length
        dtype: int64
      - name: compressed_prompt
        dtype: string
      - name: conversations
        list:
          - name: content
            dtype: string
          - name: role
            dtype: string
    splits:
      - name: train
        num_bytes: 8046176474.880298
        num_examples: 884100
    download_size: 2611849293
    dataset_size: 8046176474.880298
  - config_name: sailor2-instruct
    features:
      - name: prompt_id
        dtype: string
      - name: messages
        list:
          - name: role
            dtype: string
          - name: content
            dtype: string
      - name: language
        dtype: string
      - name: category
        dtype: string
      - name: mapped_category
        dtype: string
      - name: ppl
        dtype: float64
      - name: reward
        dtype: float64
      - name: sailor_ppl
        dtype: float64
      - name: sailor_user_ppl
        dtype: float64
      - name: sailor_assistant_ppl
        dtype: float64
      - name: ppl_rank
        dtype: float64
      - name: reward_rank
        dtype: float64
      - name: ppl_percentile
        dtype: float64
      - name: reward_percentile
        dtype: float64
      - name: combined_score
        dtype: float64
      - name: user_message_token_lengths
        sequence: int64
      - name: conversations
        list:
          - name: content
            dtype: string
          - name: role
            dtype: string
    splits:
      - name: train
        num_bytes: 1578331109.2335124
        num_examples: 191841
    download_size: 532773553
    dataset_size: 1578331109.2335124
  - config_name: transcription-speech-vi-en-550k
    features:
      - name: text_prompt
        dtype: string
      - name: answer
        dtype: string
      - name: compressed_prompt
        dtype: string
      - name: conversations
        list:
          - name: content
            dtype: string
          - name: role
            dtype: string
    splits:
      - name: train
        num_bytes: 3265755403.3161187
        num_examples: 560643
    download_size: 908279115
    dataset_size: 3265755403.3161187
configs:
  - config_name: VTSNLP-instruct-filtered
    data_files:
      - split: train
        path: VTSNLP-instruct-filtered/train-*
  - config_name: english-multiturn
    data_files:
      - split: train
        path: english-multiturn/train-*
  - config_name: instruction-speech-v1
    data_files:
      - split: train
        path: instruction-speech-v1/train-*
  - config_name: instruction-speech-v1-rephrase
    data_files:
      - split: train
        path: instruction-speech-v1-rephrase/train-*
  - config_name: instruction-speech-v2
    data_files:
      - split: train
        path: instruction-speech-v2/train-*
  - config_name: sailor2-instruct
    data_files:
      - split: train
        path: sailor2-instruct/train-*
  - config_name: transcription-speech-vi-en-550k
    data_files:
      - split: train
        path: transcription-speech-vi-en-550k/train-*