ADPO
[CVPR 2026 Highlight] Unified Generation and Self-Verification for Vision-Language Models via Advantage Decoupled Preference Optimization
File Explorer
- carousel1.jpg
- pipeline.png
- ciou.py
- majority_vote.py
- merge_json.py
- test_od_r1.py
- test_rec_baseline.py
- test_rec_r1.py
- test_rec_r1_internvl.py
- vllm_inference.py
- ddp.yaml
- qwen2vl_sft_config.yaml
- zero2.yaml
- zero3.yaml
- rec.yaml
- rec_internvl.yaml
- gui_multi-image.jsonl
- 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
- zero_stage2_config.json
- multinode_training_args.yaml
- run_adpo_rec.sh
- run_grpo_rec.sh
- __init__.py
- adpo_trainer.py
- grpo_config.py
- grpo_trainer.py
- vllm_grpo_trainer.py
- coco.py
- cocoeval.py
- __init__.py
- callbacks.py
- evaluation.py
- hub.py
- math.py
- __init__.py
- glm_module.py
- internvl_module.py
- qwen_module.py
- vlm_module.py
- __init__.py
- adpo_jsonl.py
- configs.py
- evaluate.py
- generate.py
- grpo.py
- grpo_jsonl.py
- grpo_rec.py
- qwen2_5vl_monkey_patch.py
- sft.py
- .gitignore
- LICENSE
- Makefile
- setup.cfg
- setup.py
- .gitignore
- CODE_OF_CONDUCT.md
- eval.sh
- LICENSE
- README.md
- setup.sh
- train.sh
# Use via CDN
jsDelivrjsDelivr serves any public GitHub repository as a CDN with zero setup. Pick a version and a file to get a ready-to-paste link and snippet.
Command Glossary
Commands referenced in this DOCs, explained below.
conda activate
View Details ▼
conda activate
Activate a conda environment.
See also: `conda deactivate`.
conda activate myenv
Activate an existing environment named `myenv`:
conda activate {{path/to/myenv}}
Activate an existing environment located at custom path:
conda activate --stack myenv
Stack `myenv` environment on top of a previous environment making libraries/commands/variables from both accessible:
conda create
View Details ▼
conda create
Create new conda environments.
conda create {{[-y|--yes]}} {{[-n|--name]}} py39 python=3.9 "numpy>=1.11" scipy
Create a new environment named `py39`, install Python 3.9, NumPy v1.11 or above in it, and the latest stable version of SciPy. Say yes to all confirmations:
conda create {{[-n|--name]}} myenv --file {{file1.yml}} --file {{file2.yml}}
Create a new environment named `myenv` and install packages listed in files:
conda create {{[-p|--prefix]}} {{path/to/myenv}}
Create a new environment at a custom path (i.e. prefix):
huggingface-cli
View Details ▼
huggingface-cli
This command is an alias of `hf`.
tldr hf
View documentation for the original command:
# Project Badges
-
arXiv Visit
