mlx-community/DeepSeek-R1-Distill-Llama-70B-3bit

Quantization vs deepseek-ai/DeepSeek-R1-Distill-Llama-70B. mlx-community/DeepSeek-R1-Distill-Llama-70B-3bit is a quantized version of the other (U32/F16).

On the Hub · Downloads on Model Pulse