Automatic Speech Recognition
MLX
meralion3
rotorquant
kv-cache-quantization
meralion
speech-to-text
multimodal
audio
quantized
4bit
apple-silicon
custom_code
Instructions to use majentik/MERaLiON-3-10B-RotorQuant-MLX-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use majentik/MERaLiON-3-10B-RotorQuant-MLX-4bit with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir MERaLiON-3-10B-RotorQuant-MLX-4bit majentik/MERaLiON-3-10B-RotorQuant-MLX-4bit
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
| { | |
| "auto_map": { | |
| "AutoProcessor": "processing_meralion3.MERaLiON3Processor" | |
| }, | |
| "do_normalize": true, | |
| "fixed_speech_embeds_length": 300, | |
| "processor_class": "MERaLiON3Processor", | |
| "speech_token_index": 255999, | |
| "time_duration_limit": 300, | |
| "whisper_chunk_size": 30 | |
| } | |