README.md

---
dataset_info:
  config_name: meta-active-reading-upsample
  features:
  - name: id
    dtype: string
  - name: text
    dtype: string
  - name: thinking
    dtype: string
  - name: template
    dtype: string
  - name: family
    dtype: string
  - name: pool
    dtype: string
  - name: system
    dtype: string
  - name: temperature
    dtype: float32
  - name: top_p
    dtype: float32
  - name: max_tokens
    dtype: int32
  - name: seed
    dtype: int64
  - name: enable_thinking
    dtype: bool
  - name: response
    struct:
    - name: id
      dtype: string
    - name: choices
      list:
      - name: finish_reason
        dtype: string
      - name: index
        dtype: int32
      - name: logprobs
        list:
        - name: content
          struct:
          - name: token
            dtype: string
          - name: bytes
            list: int32
          - name: logprob
            dtype: float32
          - name: top_logprobs
            list:
            - name: token
              dtype: string
            - name: bytes
              list: int32
            - name: logprob
              dtype: float32
        - name: refusal
          struct:
          - name: token
            dtype: string
          - name: bytes
            list: int32
          - name: logprob
            dtype: float32
          - name: top_logprobs
            list:
            - name: token
              dtype: string
            - name: bytes
              list: int32
            - name: logprob
              dtype: float32
      - name: message
        struct:
        - name: content
          dtype: string
        - name: refusal
          dtype: string
        - name: role
          dtype: string
        - name: annotations
          list:
          - name: type
            dtype: string
          - name: url_citation
            struct:
            - name: end_index
              dtype: int32
            - name: start_index
              dtype: int32
            - name: title
              dtype: string
            - name: url
              dtype: string
        - name: audio
          struct:
          - name: id
            dtype: string
          - name: data
            dtype: string
          - name: expires_at
            dtype: int64
          - name: transcript
            dtype: string
        - name: function_call
          struct:
          - name: arguments
            dtype: string
          - name: name
            dtype: string
        - name: tool_calls
          list:
          - name: id
            dtype: string
          - name: function
            struct:
            - name: arguments
              dtype: string
            - name: name
              dtype: string
          - name: type
            dtype: string
    - name: created
      dtype: int64
    - name: model
      dtype: string
    - name: object
      dtype: string
    - name: service_tier
      dtype: string
    - name: system_fingerprint
      dtype: string
    - name: usage
      struct:
      - name: completion_tokens
        dtype: int32
      - name: prompt_tokens
        dtype: int32
      - name: total_tokens
        dtype: int32
      - name: completion_tokens_details
        struct:
        - name: accepted_prediction_tokens
          dtype: int32
        - name: audio_tokens
          dtype: int32
        - name: reasoning_tokens
          dtype: int32
        - name: rejected_prediction_tokens
          dtype: int32
      - name: prompt_tokens_details
        struct:
        - name: audio_tokens
          dtype: int32
        - name: cached_tokens
          dtype: int32
  splits:
  - name: train
    num_bytes: 38636152
    num_examples: 1000
  download_size: 20359407
  dataset_size: 38636152
configs:
- config_name: meta-active-reading-upsample
  data_files:
  - split: train
    path: meta-active-reading-upsample/train-*
---

Samples of synthetic upsamples for MV2