INCModel3/Qwen3.5-9B-MXFP8-FP8KV-FP8Attn-CT-AutoRound

Quantization vs Qwen/Qwen3.5-9B. INCModel3/Qwen3.5-9B-MXFP8-FP8KV-FP8Attn-CT-AutoRound is a quantized version of the other (F8_E4M3/BF16).

On the Hub · Downloads on Model Pulse