lyc1995452

paddle_ocr VL

3

### 🔎 Search before asking - [x] I have searched the PaddleOCR [Docs](https://paddlepaddle.github.io/PaddleOCR/) and found no similar bug report. - [x] I have searched the PaddleOCR [Issues](https://github.com/PaddlePaddle/PaddleOCR/issues) and found no...

Support for Qwen2-VL 2B/4B? or Qwen2.5-VL

3

### System Info Is this support capable yet for Qwen2-VL 2B/4B or Qwen2.5-VL? I am curious to run TensorRT with these models. Thank you! ❤️ ### How would you like...

你好 我在网上下载了权重 推理是有问题

模型效果不行？

2

1、你好 我用你网上开源的weights权重 2、按照你的方法初始化：# 根据字典生成模型配置 python tool/init_custdata_model.py \ --cust_vocab ./cust-data/vocab.txt \ --pretrain_model ./weights \ --cust_data_init_weights_path ./cust-data/weights # 将生成vocab.json 替换到./weights中 cp ./cust-data/weights/vocab.json ./weights # 再次生成，获得正确的tokenizer.json python tool/init_custdata_model.py \ --cust_vocab ./cust-data/vocab.txt \ --pretrain_model...