训练模型

[Harryoung]

我们提供了issue bot来帮助你解答问题，你是否想要它来解答(请填写yes/no)?/We provide issue bot to solve issues, do you want it? (Please write yes/no): yes

请注意：issue bot基于大语言模型实现，其回答仅供参考。目前issue bot尚不能解析issue中的图片。/Please note: The issue bot is implemented based on large language models, and its answers are for reference only. Currently the issue bot cannot parse images in issues.

请在这里描述你的问题/Please describe your issue here:
为什么PP-OCRv4超轻量模型的训练模型，下载解压后没有类似PP-OCRv3训练模型的best-accuracy文件，只是一个student文件

[llm-issue-bot]

Issue Response

• ✨ Introduction: Thanks for opening this issue, I am the issue bot that will help you with your questoins!
• 🎯 First round response: The answer to your question is:

The PP-OCRv4 lightweight model does not have a "best-accuracy" file like the PP-OCRv3 training model. It only contains a "student" file.

Related Sources

https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.7/applications/PCB字符识别/PCB字符识别.md, https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.7/applications/轻量级车牌识别.md, https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.7/doc/doc_ch/PP-OCRv3_det_train.md

How to use

To invoke me for more questions, add a comment using following command:
@llm-issue-bot <QUESTION>

[Harryoung]

@llm-issue-bot 你不能使用中文回答我的问题吗

[llm-issue-bot]

Thanks for mentioning me!

我不知道。

Relevant Sources：

How to use

My answer may not be accurate. I suggest you refer to relevant sources to make a comprehensive judgment.
To invoke me for more questions, add a comment using following command:
@llm-issue-bot <QUESTION>