cstr/llama3-8b-spaetzle-v33-mlx-4bit
Quantization vs cstr/llama3-8b-spaetzle-v20. cstr/llama3-8b-spaetzle-v33-mlx-4bit is a quantized version of the other (U32/F16).
Quantization vs cstr/llama3-8b-spaetzle-v20. cstr/llama3-8b-spaetzle-v33-mlx-4bit is a quantized version of the other (U32/F16).