[BugFix] Qwen-image performance regressed - Avoid mapping diffusion_batch_size onto scheduler max_num_seqs - #6525

[BugFix] Qwen-image performance regressed - Avoid mapping diffusion_batch_size onto scheduler max_num_seqs#6525

Gaohan123 merged 2 commits into

vllm-project:mainvllm-project/vllm-omni:mainfrom

NumberWan:fix/diffusion-seq-capacity-request-mode-onlyNumberWan/vllm-omni:fix/diffusion-seq-capacity-request-mode-onlyCopy head branch name to clipboard