AxionML/Qwen3.5-2B-NVFP4

Quantization vs Qwen/Qwen3.5-2B. AxionML/Qwen3.5-2B-NVFP4 is a quantized version of the other (BF16/U8).

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

On the Hub · Downloads on Model Pulse