【Hackathon 8th No.36】Add gme-Qwen2-VL for PaddleMix

[ZhijunLStudio]

Add gme-Qwen2-VL for PaddleMix

[New Feature] Implement GmeQwen2VL Model and Multimodal Inference Pipeline

Changes Proposed

1. Model Architecture & Inference Pipeline Implementation

   • Added the text, image, and fused embedding computation modules of the GmeQwen2VL model to paddlemix/models.
   • Integrated a multimodal retrieval task example in paddlemix/examples, including text-to-image similarity calculation and information retrieval workflows.

2. Functional Enhancements

   • Supports custom instructions for text embeddings (e.g., "Find an image that matches the given text.").
   • Provides fused embedding interfaces (get_fused_embeddings()) and encoding functions for queries/databases (encode_queries() / encode_corpus()).

3. Performance Validation

   • Achieved performance comparable to the original repository through internal testing, with validated accuracy in similarity calculations (see documentation for test results).

[paddle-bot]

Thanks for your contribution!

[lyuwenyu]

把这些图片去掉吧；用到的地方使用链接的方式（把图片拉到issue里，会生成相关的链接）

[ZhijunLStudio]

已修改

[lyuwenyu]

py文件都加上copyright吧；可以从已有的文件里copy

[ZhijunLStudio]

已修改