Instructions to use vdeturckheim/skywave-segmenter-small with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use vdeturckheim/skywave-segmenter-small with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir skywave-segmenter-small vdeturckheim/skywave-segmenter-small
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
| { | |
| "balancedRecall" : 0.9955027, | |
| "classes" : [ | |
| "music", | |
| "speech" | |
| ], | |
| "heldOut" : [ | |
| "h20", | |
| "h12" | |
| ], | |
| "hop" : 320, | |
| "mode" : "stream", | |
| "nFft" : 400, | |
| "nMels" : 64, | |
| "sampleRate" : 16000, | |
| "seed" : 2, | |
| "spec" : { | |
| "causal" : false, | |
| "channels" : [ | |
| 16, | |
| 32, | |
| 48, | |
| 64 | |
| ], | |
| "classes" : [ | |
| "music", | |
| "speech" | |
| ], | |
| "dilations" : [ | |
| 1, | |
| 4, | |
| 16 | |
| ], | |
| "headWidth" : 64, | |
| "melPools" : [ | |
| 2, | |
| 2, | |
| 2, | |
| 2 | |
| ], | |
| "name" : "small", | |
| "separable" : false, | |
| "timePools" : [ | |
| 2, | |
| 2, | |
| 1, | |
| 1 | |
| ] | |
| }, | |
| "stride" : 4 | |
| } |