FP8 inference optimization

Qwen Releases Qwen3-VL-30B-A3B Instruct/Thinking: 3B Activation Parameters, New Multimodal Combination Supporting FP8

On October 4, 2025, Qwen officially announced the launch of two new multimodal models, Qwen3-VL-30B-A3B-Instruct and -Thinking, in its codebase, and s...