openslr/librispeech_asr
Viewer • Updated • 585k • 107k • 225
How to use animaslabs/parakeet-ctc-0.6b-mlx with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir parakeet-ctc-0.6b-mlx animaslabs/parakeet-ctc-0.6b-mlx
How to use animaslabs/parakeet-ctc-0.6b-mlx with Transformers:
# Use a pipeline as a high-level helper
from transformers import pipeline
pipe = pipeline("automatic-speech-recognition", model="animaslabs/parakeet-ctc-0.6b-mlx") # Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("animaslabs/parakeet-ctc-0.6b-mlx", dtype="auto")This model was converted to MLX format from nvidia/parakeet-ctc-0.6b using the scripts in this github repo. Please refer to original model card for more details on the model.
NOTE: Some newer or quantized models may not be supported in parakeet-mlx or mlx-audio
pip install -U parakeet-mlx
parakeet-mlx audio.wav --model animaslabs/parakeet-ctc-0.6b-mlx
pip install -U mlx-audio
python -m mlx_audio.stt.generate --model animaslabs/parakeet-ctc-0.6b-mlx --audio audio.wav
Quantized
Base model
nvidia/parakeet-ctc-0.6b