Hugging Face's logo

Datasets:
Evo-LMM
/
EMILIA

Dataset card Files Community
Dataset Viewer
Full Screen
The dataset viewer is not available for this subset.
Job manager crashed while running this job (missing heartbeats).

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

YAML Metadata Warning: empty or missing yaml metadata in repo card (https://huggingface.co/docs/hub/datasets-cards)

File Structure 

EMILIA/
├── worker_X_Y_batch1/           # Folder containing audio files for batch 1
│   ├── audio_0.zip             # Contains samples 0-999
│   ├── audio_1.zip             # Contains samples 1000-1999
│   └── ...                     # Total 200 zip files per batch folder
├── emilia_worker_X_Y_batch1.jsonl  # Corresponding JSONL file for batch 1
├── worker_X_Y_batch2/           # Next batch folder
└── emilia_worker_X_Y_batch2.jsonl  # Next batch JSONL file

Batch Information

  • Each batch folder contains 200,000 samples total
  • Samples are split into 200 zip files
  • Each zip file contains 1,000 audio samples
  • The JSONL file named emilia_worker_X_Y_batchN.jsonl corresponds to the folder worker_X_Y_batchN and contains metadata for all 200,000 samples in that batch
Downloads last month
5