├── .gitignore ├── Dockerfile ├── LICENSE ├── README.md ├── datasets ├── KADID-10K │ ├── KADID-10K_mos.txt │ ├── KADID-10K_normalized_mos.txt │ ├── KADID-10K_test_set.txt │ ├── KADID-10K_train_set.txt │ ├── KADID-10K_val_set.txt │ └── scoring │ │ └── RL-KADID-10K_train_scoring.jsonl ├── combined │ └── RL-622-KADID-TID2013-KONIQ-LIVEC_train_scoring.jsonl ├── make_data.py └── test_data_label │ ├── AGIQA-3K_test_all.txt │ ├── BID_test_all.txt │ ├── Deblur_test_all.txt │ ├── KONIQ-10K_test_all.txt │ ├── LIVEC_test_all.txt │ ├── SPAQ_test_all.txt │ ├── SRIQA-Bench_test_all.txt │ └── dehaze_test_all.txt ├── images ├── intro.png ├── logo.png └── realistic_scenario.png ├── setup.sh └── src └── open-r1-multimodal ├── .gitignore ├── LICENSE ├── Makefile ├── configs ├── ddp.yaml ├── qwen2vl_sft_config.yaml ├── zero2.yaml └── zero3.yaml ├── data_config ├── rec.yaml └── rec_internvl.yaml ├── data_jsonl └── gui_multi-image.jsonl ├── local_scripts ├── create_vision_cot_data.py ├── lmms_eval_qwen2vl.sh ├── prepare_hf_data.py ├── train_aria_moe.sh ├── train_qwen2_vl.sh ├── zero2.json ├── zero3.json ├── zero3.yaml └── zero3_offload.json ├── run_scripts ├── KADID-10K │ ├── multi_run_kadid.sh │ └── one_node_run_kadid.sh └── multinode_training_args.yaml ├── setup.cfg ├── setup.py └── src └── open_r1 ├── __init__.py ├── configs.py ├── evaluate.py ├── generate.py ├── grpo.py ├── grpo_jsonl.py ├── grpo_rec.py ├── qwen2_5vl_monkey_patch.py ├── sft.py ├── trainer ├── __init__.py ├── grpo_config.py ├── grpo_trainer.py └── vllm_grpo_trainer.py ├── utils ├── __init__.py ├── callbacks.py ├── evaluation.py ├── hub.py ├── math.py └── pycocotools │ ├── coco.py │ └── cocoeval.py └── vlm_modules ├── __init__.py ├── internvl_module.py ├── qwen_module.py └── vlm_module.py /.gitignore: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/.gitignore -------------------------------------------------------------------------------- /Dockerfile: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/Dockerfile -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/LICENSE -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/README.md -------------------------------------------------------------------------------- /datasets/KADID-10K/KADID-10K_mos.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/datasets/KADID-10K/KADID-10K_mos.txt -------------------------------------------------------------------------------- /datasets/KADID-10K/KADID-10K_normalized_mos.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/datasets/KADID-10K/KADID-10K_normalized_mos.txt -------------------------------------------------------------------------------- /datasets/KADID-10K/KADID-10K_test_set.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/datasets/KADID-10K/KADID-10K_test_set.txt -------------------------------------------------------------------------------- /datasets/KADID-10K/KADID-10K_train_set.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/datasets/KADID-10K/KADID-10K_train_set.txt -------------------------------------------------------------------------------- /datasets/KADID-10K/KADID-10K_val_set.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/datasets/KADID-10K/KADID-10K_val_set.txt -------------------------------------------------------------------------------- /datasets/KADID-10K/scoring/RL-KADID-10K_train_scoring.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/datasets/KADID-10K/scoring/RL-KADID-10K_train_scoring.jsonl -------------------------------------------------------------------------------- /datasets/combined/RL-622-KADID-TID2013-KONIQ-LIVEC_train_scoring.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/datasets/combined/RL-622-KADID-TID2013-KONIQ-LIVEC_train_scoring.jsonl -------------------------------------------------------------------------------- /datasets/make_data.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/datasets/make_data.py -------------------------------------------------------------------------------- /datasets/test_data_label/AGIQA-3K_test_all.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/datasets/test_data_label/AGIQA-3K_test_all.txt -------------------------------------------------------------------------------- /datasets/test_data_label/BID_test_all.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/datasets/test_data_label/BID_test_all.txt -------------------------------------------------------------------------------- /datasets/test_data_label/Deblur_test_all.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/datasets/test_data_label/Deblur_test_all.txt -------------------------------------------------------------------------------- /datasets/test_data_label/KONIQ-10K_test_all.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/datasets/test_data_label/KONIQ-10K_test_all.txt -------------------------------------------------------------------------------- /datasets/test_data_label/LIVEC_test_all.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/datasets/test_data_label/LIVEC_test_all.txt -------------------------------------------------------------------------------- /datasets/test_data_label/SPAQ_test_all.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/datasets/test_data_label/SPAQ_test_all.txt -------------------------------------------------------------------------------- /datasets/test_data_label/SRIQA-Bench_test_all.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/datasets/test_data_label/SRIQA-Bench_test_all.txt -------------------------------------------------------------------------------- /datasets/test_data_label/dehaze_test_all.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/datasets/test_data_label/dehaze_test_all.txt -------------------------------------------------------------------------------- /images/intro.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/images/intro.png -------------------------------------------------------------------------------- /images/logo.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/images/logo.png -------------------------------------------------------------------------------- /images/realistic_scenario.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/images/realistic_scenario.png -------------------------------------------------------------------------------- /setup.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/setup.sh -------------------------------------------------------------------------------- /src/open-r1-multimodal/.gitignore: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/.gitignore -------------------------------------------------------------------------------- /src/open-r1-multimodal/LICENSE: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/LICENSE -------------------------------------------------------------------------------- /src/open-r1-multimodal/Makefile: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/Makefile -------------------------------------------------------------------------------- /src/open-r1-multimodal/configs/ddp.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/configs/ddp.yaml -------------------------------------------------------------------------------- /src/open-r1-multimodal/configs/qwen2vl_sft_config.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/configs/qwen2vl_sft_config.yaml -------------------------------------------------------------------------------- /src/open-r1-multimodal/configs/zero2.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/configs/zero2.yaml -------------------------------------------------------------------------------- /src/open-r1-multimodal/configs/zero3.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/configs/zero3.yaml -------------------------------------------------------------------------------- /src/open-r1-multimodal/data_config/rec.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/data_config/rec.yaml -------------------------------------------------------------------------------- /src/open-r1-multimodal/data_config/rec_internvl.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/data_config/rec_internvl.yaml -------------------------------------------------------------------------------- /src/open-r1-multimodal/data_jsonl/gui_multi-image.jsonl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/data_jsonl/gui_multi-image.jsonl -------------------------------------------------------------------------------- /src/open-r1-multimodal/local_scripts/create_vision_cot_data.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/local_scripts/create_vision_cot_data.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/local_scripts/lmms_eval_qwen2vl.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/local_scripts/lmms_eval_qwen2vl.sh -------------------------------------------------------------------------------- /src/open-r1-multimodal/local_scripts/prepare_hf_data.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/local_scripts/prepare_hf_data.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/local_scripts/train_aria_moe.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/local_scripts/train_aria_moe.sh -------------------------------------------------------------------------------- /src/open-r1-multimodal/local_scripts/train_qwen2_vl.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/local_scripts/train_qwen2_vl.sh -------------------------------------------------------------------------------- /src/open-r1-multimodal/local_scripts/zero2.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/local_scripts/zero2.json -------------------------------------------------------------------------------- /src/open-r1-multimodal/local_scripts/zero3.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/local_scripts/zero3.json -------------------------------------------------------------------------------- /src/open-r1-multimodal/local_scripts/zero3.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/local_scripts/zero3.yaml -------------------------------------------------------------------------------- /src/open-r1-multimodal/local_scripts/zero3_offload.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/local_scripts/zero3_offload.json -------------------------------------------------------------------------------- /src/open-r1-multimodal/run_scripts/KADID-10K/multi_run_kadid.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/run_scripts/KADID-10K/multi_run_kadid.sh -------------------------------------------------------------------------------- /src/open-r1-multimodal/run_scripts/KADID-10K/one_node_run_kadid.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/run_scripts/KADID-10K/one_node_run_kadid.sh -------------------------------------------------------------------------------- /src/open-r1-multimodal/run_scripts/multinode_training_args.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/run_scripts/multinode_training_args.yaml -------------------------------------------------------------------------------- /src/open-r1-multimodal/setup.cfg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/setup.cfg -------------------------------------------------------------------------------- /src/open-r1-multimodal/setup.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/setup.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/configs.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/configs.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/evaluate.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/evaluate.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/generate.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/generate.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/grpo.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/grpo.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/grpo_jsonl.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/grpo_jsonl.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/grpo_rec.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/grpo_rec.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/qwen2_5vl_monkey_patch.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/qwen2_5vl_monkey_patch.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/sft.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/sft.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/trainer/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/trainer/__init__.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/trainer/grpo_config.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/trainer/grpo_config.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/trainer/grpo_trainer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/trainer/grpo_trainer.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/trainer/vllm_grpo_trainer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/trainer/vllm_grpo_trainer.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/utils/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/utils/callbacks.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/utils/callbacks.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/utils/evaluation.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/utils/evaluation.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/utils/hub.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/utils/hub.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/utils/math.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/utils/math.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/utils/pycocotools/coco.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/utils/pycocotools/coco.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/utils/pycocotools/cocoeval.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/utils/pycocotools/cocoeval.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/vlm_modules/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/vlm_modules/__init__.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/vlm_modules/internvl_module.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/vlm_modules/internvl_module.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/vlm_modules/qwen_module.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/vlm_modules/qwen_module.py -------------------------------------------------------------------------------- /src/open-r1-multimodal/src/open_r1/vlm_modules/vlm_module.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TianheWu/VisualQuality-R1/HEAD/src/open-r1-multimodal/src/open_r1/vlm_modules/vlm_module.py --------------------------------------------------------------------------------