Instructions to use Peacockery/MOSS-Transcribe-preview-2B-MLX with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use Peacockery/MOSS-Transcribe-preview-2B-MLX with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir MOSS-Transcribe-preview-2B-MLX Peacockery/MOSS-Transcribe-preview-2B-MLX
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
MOSS-Transcribe-preview-2B-MLX / MOSS-Transcribe-preview-2B-text-decoder-4bit-g64 /quantization-report.json
| { | |
| "bits": 4, | |
| "elapsed_sec": 12.934918417129666, | |
| "group_size": 64, | |
| "mode": "affine", | |
| "output_dir": "/Users/simonpeacocks/GitHub/moss-mlx-conversion/artifacts/mlx/MOSS-Transcribe-preview-2B-text-decoder-4bit-g64", | |
| "quantize_elapsed_sec": 0.030685750301927328, | |
| "save_elapsed_sec": 7.085229250136763, | |
| "saved_tensor_count": 1230, | |
| "scope": "text-decoder", | |
| "source_dir": "/Users/simonpeacocks/GitHub/moss-mlx-conversion/artifacts/mlx/MOSS-Transcribe-preview-2B-bf16", | |
| "weight_bytes": 2811958960, | |
| "weight_path": "/Users/simonpeacocks/GitHub/moss-mlx-conversion/artifacts/mlx/MOSS-Transcribe-preview-2B-text-decoder-4bit-g64/weights.safetensors" | |
| } | |