---
dataset_info:
  features:
  - name: sample_id
    dtype: string
  - name: dataset_id
    dtype: string
  - name: subset_id
    dtype: string
  - name: reference
    dtype: string
  - name: hypothesis
    dtype: string
  - name: system_id
    dtype: string
  - name: wer
    dtype: float64
  - name: cer
    dtype: float64
  - name: error_type
    dtype: string
  - name: severity
    dtype: int64
  - name: audio_quality_ok
    dtype: bool
  - name: reference_correct
    dtype: bool
  - name: notes
    dtype: string
  - name: annotator_id
    dtype: string
  - name: llm_error_categories
    sequence: string
  - name: llm_severity
    dtype: string
  - name: llm_confidence
    dtype: float64
  - name: llm_model_id
    dtype: string
  - name: split
    dtype: string
  splits:
  - name: train
    num_bytes: 4674
    num_examples: 11
  download_size: 11429
  dataset_size: 4674
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---