Actions: EvolvingLMMs-Lab/lmms-eval
Actions
Showing runs from all workflows
2,500+ workflow runs
2,500+ workflow runs
vllm model drops video timestamps (to_openai_messages), degrading Qwen3-VL/Qwen3.5 on motion benchmarks
Claude Code
#1346:
Issue #1404
opened
by
njb-nvidia