Instructions to use faraday/ema-lightning-mlx with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use faraday/ema-lightning-mlx with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] hf download faraday/ema-lightning-mlx --local-dir ema-lightning-mlx
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
Download conversion-manifest.json from faraday/ema-lightning-mlx: direct link, hf CLI and curl.
- Browser
- Download file 1.27 kB
-
https://huggingface.co/faraday/ema-lightning-mlx/resolve/main/conversion-manifest.json
- Command line
-
hf download hf://faraday/ema-lightning-mlx/conversion-manifest.json
-
curl -L -o conversion-manifest.json https://huggingface.co/faraday/ema-lightning-mlx/resolve/main/conversion-manifest.json
1.27 kB
| { | |
| "created_date": "2026-10-18", | |
| "format": "ema-mlx-voice", | |
| "format_version": 2, | |
| "source_revision": "12797c5f4dbcced8e4b3392b12bba2fe8b9f5ec6", | |
| "checkpoints": { | |
| "ema.pt": "95aec03dafbe0e1d69bca774ab597c779464729a14bc99bfcb52480090c7dfe6", | |
| "decoder.pt": "9595819b173f411340f63d11332695121a97f8bf1f6d8b6fef0b21cf99c7ad67" | |
| }, | |
| "source_export_directory": "Engines/mlx/artifacts/OriginalVoice", | |
| "exporter_sha256": "ab447f7eab12a0a536668a97a73c19cf932c6bfd0a0b5776042c378627151ba8", | |
| "asset_checks": { | |
| "acoustic.safetensors": { | |
| "sha256": "86e00e2c197c34b53f4eef2e330706e6a5aa7ba442fa3d8024db23a7821d2cd8", | |
| "bytes": 22240080, | |
| "tensor_count": 105, | |
| "dtype": "float32", | |
| "shapes_verified": true | |
| }, | |
| "decoder.safetensors": { | |
| "sha256": "834ba4d7671663c790348c55d39fb95f56b78ff29c39c52726518ad73c28c4a2", | |
| "bytes": 11993644, | |
| "tensor_count": 232, | |
| "dtype": "float32", | |
| "shapes_verified": true | |
| }, | |
| "constants.safetensors": { | |
| "sha256": "b63d3764d951f3e21e7800947e8e54a18a303bda4e992660ad341c6513848737", | |
| "bytes": 758872, | |
| "tensor_count": 7, | |
| "dtype": "float32", | |
| "shapes_verified": true | |
| } | |
| }, | |
| "stored_precision": "float32", | |
| "training_performed": false | |
| } | |