-
*********************************************
温馨提示:根据社区不完全统计,按照模板提问,可以加快回复和解决问题的速度
*********************************************
## 环境
- 【FastDeploy版本】: 说明具体的版本,如fastdeploy-linux-gpu-0.8.0
- …
-
请提供下述完整信息以便快速定位问题/Please provide the following information to quickly locate the problem
- 系统环境/System Environment:Windows和Linux都有此问题
- 版本号/Version:Paddleocr和paddlepaddle版本都为最新
部分代码:
```python
…
-
模型是ppocr训练的,已转换成pplite
目前用flutter及rust无法部署,出错太多
请求给个方向指导
-
用的官方的demo,使用的识别模型是:ch_rec_mv3_crnn_opt.nb
如何设置才可以让识别的单词结果有空格呢?
![1677947244275](https://user-images.githubusercontent.com/29346132/222917405-188f3e35-e68e-4e5d-99f5-a8a89883a0c3.png)
-
请提供下述完整信息以便快速定位问题/Please provide the following information to quickly locate the problem
- 系统环境/System Environment:
**win11中使用wsl的conda环境,wsl中安装了11.2,cudnn8.2**
- 版本号/Version:Paddle: PaddleOCR: …
-
我想要识别的文本比较长,所以想尝试将max_text_length设置为一个更大的数,但是改变该值后(默认为25,修改成50后),训练报错:
```
Traceback (most recent call last):
File "PaddleOCR/tools/train.py", line 208, in
main(config, device, logger, vd…
-
I have been trying to train PaddleOCR using NIST Special Database 19 for handwritten text. I have followed all the steps shown in the git page except for text detection training. After using "python t…
-
中文识别精度真的很低啊,虽然框框都画出来了,但是能把溯和湖混淆。
- 版本号/Version:"@paddle-js-models/ocr": "4.0.0"
-
请提供下述完整信息以便快速定位问题/Please provide the following information to quickly locate the problem
- 系统环境/System Environment:
- mac osx 14,intel x86 64bit
- 版本号/Version:Paddle: PaddleOCR: 问题相关组件/Re…
-
请提供下述完整信息以便快速定位问题/Please provide the following information to quickly locate the problem
- 系统环境/System Environment:Ubuntu 20.04
- 版本号/Version:Paddle:2.6.0 PaddleNLP 2.7.2 : 问题相关组件/Related compone…