---
dataset_info:
  config_name: sft
  features:
  - name: prompt_id
    dtype: string
  - name: bucket
    dtype: string
  - name: messages
    list:
    - name: role
      dtype: string
    - name: content
      dtype: string
  - name: teacher
    dtype: string
  - name: temperature
    dtype: float32
  - name: top_p
    dtype: float32
  - name: seed
    dtype: int32
  - name: gen_hash
    dtype: string
  - name: schema_v
    dtype: int32
  - name: finish_reason
    dtype: string
  splits:
  - name: train
    num_bytes: 123871
    num_examples: 96
  download_size: 125085
  dataset_size: 123871
configs:
- config_name: sft
  data_files:
  - split: train
    path: sft/train-*
---