Datasets:

AudioLLMs
/

seame_dev_sge

dataset_info:
  config_name: seame_dev_sge
  features:
    - name: context
      dtype:
        audio:
          sampling_rate: 16000
    - name: instruction
      dtype: string
    - name: answer
      dtype: string
  splits:
    - name: test
      num_bytes: 283225959.278
      num_examples: 3222
  download_size: 278600000
  dataset_size: 283225959.278
configs:
  - config_name: seame_dev_sge
    data_files:
      - split: test
        path: seame_dev_sge/train-*

This is the dev dataset.