# Qwen-Image-2.0-RL 技术报告

- 来源：HuggingFace Daily Papers（社区热门论文）
- 发布时间：2026-06-25 08:00
- AIHOT 分数：48
- AIHOT 链接：https://aihot.virxact.com/items/cmqylfgci02xusliv4qvht0r6
- 原文链接：https://arxiv.org/abs/2606.27608

## AI 摘要

Qwen-Image-2.0-RL 后训练流程应用 RLHF 和 on-policy distillation（OPD）提升 Qwen-Image-2.0 扩散模型的视觉质量与指令跟随。通过微调视觉语言模型构建任务特定复合奖励模型，覆盖文生图的对齐、美学、肖像保真度，以及图像编辑的指令准确性与人脸身份保持。基于 GRPO 的 RL 框架引入混合 CFG 策略、组内奖励范围过滤和分类权重校准。最终通过轨迹级速度匹配的 OPD 合并多个教师策略。评测显示，Qwen-Image-2.0-RL 在 Qwen-Image-Bench 总分 57.84（+2.61），文生图 Elo 1193（+78），图像编辑 Elo 1349（+93）。

## 正文

We present Qwen-Image-2.0-RL, a post-training pipeline that applies reinforcement learning from human feedback (RLHF) and on-policy distillation (OPD) to improve both the visual quality and instruction-following capability of the Qwen-Image-2.0 diffusion model. To provide reliable reward signals, we construct task-specific composite reward models by fine-tuning vision-language models with a pointwise scoring paradigm and chain-of-thought reasoning. For text-to-image generation, the reward models cover alignment, aesthetics, and portrait fidelity dimensions. For image editing tasks, the reward system addresses instruction-following accuracy and face identity preservation. Building on this reward system, we develop a scalable GRPO-based RL training framework, incorporating a hybrid classifier-free guidance (CFG) strategy to preserve pre-trained knowledge, prompt curation via intra-group reward range filtering, and per-category reward weight calibration. To merge the task-specialized RL policies for T2I and editing, we propose on-policy distillation as the final training stage, which consolidates multiple teachers into a single student model through trajectory-level velocity matching. Extensive evaluation shows that Qwen-Image-2.0-RL achieves 57.84 overall score on Qwen-Image-Bench (+2.61 over the base model), Elo ratings of 1193 in text-to-image arena (+78) and 1349 in image edit arena (+93), demonstrating consistent gains in aesthetic quality, prompt adherence, and editing accuracy.