animaslabs/parakeet-tdt-0.6b-v3-mlx-4bit

Quantization vs nvidia/parakeet-tdt-0.6b-v3. animaslabs/parakeet-tdt-0.6b-v3-mlx-4bit is a quantized version of the other (F32/U32).

1 other repos hold byte-for-byte identical weight files.

On the Hub · Downloads on Model Pulse