GRPO CARE
This repository contains the GRPO-CARE model, presented in the paper GRPO-CARE: Consistency-Aware Reinforcement Learning for Multimodal Reasoning.
Source du modèle
- Type
- Modèle
- Taille du fichier
- 1,58–4,65 GB
- Version
- vRevision 02db101
- Mis à jour
- 30 oct. 2026
- Téléchargements
- 30
- Source
- Hugging Face
Description de la source
This repository contains the GRPO-CARE model, presented in the paper GRPO-CARE: Consistency-Aware Reinforcement Learning for Multimodal Reasoning.
Code released at GRPO-CARE.
Citation
@misc{chen2025grpocareconsistencyawarereinforcementlearning,
title={GRPO-CARE: Consistency-Aware Reinforcement Learning for Multimodal Reasoning},
author={Yi Chen and Yuying Ge and Rui Wang and Yixiao Ge and Junhao Cheng and Ying Shan and Xihui Liu},
year={2025},
eprint={2506.16141},
archivePrefix={arXiv},
primaryClass={cs.CV},
url={https://arxiv.org/abs/2506.16141},
}
Sources
1 source
Sources
1 sourceVérifié 30 oct.
Download files
Qwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500 · 4 weight files
All files listed in the provider index are present.
- Qwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/model.safetensors.index.json · provider index
- Qwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/config.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/generation_config.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/model-00004-of-00004.safetensors · individual shard
- Qwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/model-00001-of-00004.safetensors · individual shard
- Qwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/model-00002-of-00004.safetensors · individual shard
- Qwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/model-00003-of-00004.safetensors · individual shard
Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500 · 4 weight files
All files listed in the provider index are present.
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/model.safetensors.index.json · provider index
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/config.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/generation_config.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/model-00004-of-00004.safetensors · individual shard
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/model-00001-of-00004.safetensors · individual shard
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/model-00002-of-00004.safetensors · individual shard
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/model-00003-of-00004.safetensors · individual shard
Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000 · 4 weight files
All files listed in the provider index are present.
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/model.safetensors.index.json · provider index
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/config.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/generation_config.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/model-00004-of-00004.safetensors · individual shard
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/model-00001-of-00004.safetensors · individual shard
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/model-00002-of-00004.safetensors · individual shard
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/model-00003-of-00004.safetensors · individual shard
Repository files
File inventory only; completeness is unverified.
- Qwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/added_tokens.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/chat_template.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/merges.txt
- Qwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/preprocessor_config.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/special_tokens_map.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/tokenizer.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/tokenizer_config.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/trainer_state.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/training_args.bin
- Qwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/vocab.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/added_tokens.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/chat_template.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/merges.txt
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/preprocessor_config.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/special_tokens_map.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/tokenizer.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/tokenizer_config.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/trainer_state.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/training_args.bin
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/vocab.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/added_tokens.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/chat_template.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/merges.txt
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/preprocessor_config.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/special_tokens_map.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/tokenizer.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/tokenizer_config.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/trainer_state.json
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/training_args.bin
- Qwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/vocab.json
- config.json
Artefacts du modèle
12 artefacts
Artefacts du modèle
12 artefactsQwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/model-00001-of-00004.safetensors
safetensors · 4,63 GB · SHA-256 0fb900d20de9…f99a · Hugging Face
Download shardQwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/model-00002-of-00004.safetensors
safetensors · 4,65 GB · SHA-256 56719e70fd41…889f · Hugging Face
Download shardQwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/model-00003-of-00004.safetensors
safetensors · 4,59 GB · SHA-256 b574ad77cd94…8008 · Hugging Face
Download shardQwen2.5-VL-7B-GRPO-CARE-Margin0-SEED-Bench-R1/checkpoint-1500/model-00004-of-00004.safetensors
safetensors · 1,58 GB · SHA-256 a934cb8e5c16…2ca3 · Hugging Face
Download shardQwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/model-00001-of-00004.safetensors
safetensors · 4,63 GB · SHA-256 962dfb0ba4de…bfc2 · Hugging Face
Download shardQwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/model-00002-of-00004.safetensors
safetensors · 4,65 GB · SHA-256 d3b5e9d980ae…1462 · Hugging Face
Download shardQwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/model-00003-of-00004.safetensors
safetensors · 4,59 GB · SHA-256 b3d7cf00b6f6…0653 · Hugging Face
Download shardQwen2.5-VL-7B-GRPO-CARE-Margin0.01-SEED-Bench-R1/checkpoint-1500/model-00004-of-00004.safetensors
safetensors · 1,58 GB · SHA-256 673687ccb602…994a · Hugging Face
Download shardQwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/model-00001-of-00004.safetensors
safetensors · 4,63 GB · SHA-256 68d9b3336c5b…7a2a · Hugging Face
Download shardQwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/model-00002-of-00004.safetensors
safetensors · 4,65 GB · SHA-256 c2bda7d9a61f…3164 · Hugging Face
Download shardQwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/model-00003-of-00004.safetensors
safetensors · 4,59 GB · SHA-256 09db5dbad952…c7eb · Hugging Face
Download shardQwen2.5-VL-7B-GRPO-CARE-Margin0.01-Video-R1/checkpoint-1000/model-00004-of-00004.safetensors
safetensors · 1,58 GB · SHA-256 f7eae40c4895…2712 · Hugging Face
Download shardExtraits de sources
3 extraits
Extraits de sources
3 extraits--- license: apache-2.0 library_name: transformers pipeline_tag: video-text-to-text --- This repository contains the GRPO-CARE model, presented in the paper [GRPO-CARE: Consistency-Aware Reinforcement Learning for Multimodal Reasoning](https://huggingface.co/papers/2506.16141). Code released at [GRPO-CARE](https://github.com/TencentARC/GRPO-CARE). ## Citation ```bibtex @misc{chen2025grpocareconsistencyawarereinforcementlearning, title={GRPO-CARE: Consistency-Aware Reinforcement Learning for Multimodal Reasoning}, author={Yi Chen and Yuying Ge and Rui Wang and Yixiao Ge and Junhao Cheng and Ying Shan and Xihui Liu}, year={2025}, eprint={2506.16141}, archivePrefix={arXiv}, primaryClass={cs.CV}, url={https://arxiv.org/abs/2506.16141}, } ```
Source context: 30 downloads · 2 likes · Pipeline video-text-to-text · Library transformers · Repo TencentARC/GRPO-CARE
Source context: 18 downloads · 2 likes · Pipeline video-text-to-text · Library transformers · Repo TencentARC/GRPO-CARE
