Instructions to use aufklarer/Cohere-Transcribe-2B-MLX-5bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use aufklarer/Cohere-Transcribe-2B-MLX-5bit with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir Cohere-Transcribe-2B-MLX-5bit aufklarer/Cohere-Transcribe-2B-MLX-5bit
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
| { | |
| "auto_map": { | |
| "AutoFeatureExtractor": "processing_cohere_asr.CohereAsrFeatureExtractor" | |
| }, | |
| "dither": 1e-05, | |
| "feature_extractor_type": "CohereAsrFeatureExtractor", | |
| "feature_size": 128, | |
| "frame_splicing": 1, | |
| "log": true, | |
| "n_fft": 512, | |
| "n_window_size": 400, | |
| "n_window_stride": 160, | |
| "normalize": "per_feature", | |
| "pad_to": 0, | |
| "padding_value": 0.0, | |
| "sampling_rate": 16000, | |
| "window": "hann" | |
| } | |