TrevorJS/voxtral-tts-q4-gguf

Quantization vs mistralai/Voxtral-4B-TTS-2603. TrevorJS/voxtral-tts-q4-gguf is a GGUF quantization (Q4) of the same weights: 169 tensors match exactly and 207 within quantization error.

Tensors changed: other 0/6, Attention 0/156, Norms 0/92, MLP 0/119, Embeddings 0/3

On the Hub · Downloads on Model Pulse