Miiche/visualrl-qwen3vl4b-vppo-k01-clean
Upload logs/eval-visualrl-qwen3vl4b-vppo-k01-clean-s270-vllm.log with huggingface_hub
eval visualrl-qwen3vl4b-vppo-k01-clean-s270-vllm: mathvision (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s270-vllm: mmmu_pro (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s270-vllm: mathverse (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s270-vllm: mmmu (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s270-vllm: hallusionbench (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s270-vllm: mathvista (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s270-vllm: mme (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s270-vllm: perceptionbench (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s270-vllm: scienceqa (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s270-vllm: mmstar (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s270-vllm: blink (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s270-vllm: logicvista (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s270-vllm: mmbench (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s270-vllm: realworldqa (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s270-vllm: mmvet (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s270-vllm: chartqa (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s270-vllm: visnumbench (scores+output+timing+log, single push)
Upload logs/eval-visualrl-qwen3vl4b-vppo-k01-clean-s180-vllm.log with huggingface_hub
eval visualrl-qwen3vl4b-vppo-k01-clean-s180-vllm: mathvision (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s180-vllm: mmmu_pro (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s180-vllm: mathverse (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s180-vllm: mmmu (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s180-vllm: hallusionbench (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s180-vllm: mathvista (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s180-vllm: mme (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s180-vllm: perceptionbench (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s180-vllm: scienceqa (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s180-vllm: mmstar (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s180-vllm: blink (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s180-vllm: logicvista (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s180-vllm: mmbench (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s180-vllm: realworldqa (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s180-vllm: mmvet (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s180-vllm: chartqa (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s180-vllm: visnumbench (scores+output+timing+log, single push)
Upload logs/eval-visualrl-qwen3vl4b-vppo-k01-clean-s90-vllm.log with huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
merged eval-form step_270
merged eval-form step_180
Upload logs/eval-visualrl-qwen3vl4b-vppo-k01-clean-s90-vllm.log with huggingface_hub
eval visualrl-qwen3vl4b-vppo-k01-clean-s90-vllm: mathvision (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s90-vllm: mmmu_pro (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s90-vllm: mathverse (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s90-vllm: mmmu (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s90-vllm: hallusionbench (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s90-vllm: mathvista (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s90-vllm: mme (scores+output+timing+log, single push)
eval visualrl-qwen3vl4b-vppo-k01-clean-s90-vllm: perceptionbench (scores+output+timing+log, single push)
