INCModel2/Qwen3-30B-A3B-Instruct-2507-MXFP8-FP8ATTN-CT-AutoRound

Quantization vs Qwen/Qwen3-30B-A3B-Instruct-2507. INCModel2/Qwen3-30B-A3B-Instruct-2507-MXFP8-FP8ATTN-CT-AutoRound is a quantized version of the other (F8_E4M3/BF16).

On the Hub · Downloads on Model Pulse