Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[TTS]使用gpu合成后显存未释放 #2908

Closed
shiyv132 opened this issue Feb 12, 2023 · 3 comments
Closed

[TTS]使用gpu合成后显存未释放 #2908

shiyv132 opened this issue Feb 12, 2023 · 3 comments
Assignees

Comments

@shiyv132
Copy link

For support and discussions, please use our Discourse forums.

If you've found a bug then please create an issue with the following information:

Describe the bug
我使用https://zhuanlan.zhihu.com/p/587765776中的代码块在anaconda的命令行窗口进行合成,windows11,gpu是rtx3050移动端,4g显存,发现每次运行完保存完文件之后显存一直维持在近满的状态,直到关闭窗口重进或退出python环境后重置。

To Reproduce
Steps to reproduce the behavior:
1.打开anaconda,进入虚拟环境,python版本3.8.16
2. 运行更改了待合成文本的代码块
3. 第一次合成正常,运行完后显存维持在3.7g左右,等待一两分钟无果
4. 再次合成即报错,显示(翻译)未能分配足够(大概500m)显存,剩余0,退出重新进入合成能正常运行,但又出现以上情况

Expected behavior
A clear and concise description of what you expected to happen.

Screenshots
wrong

If applicable, add screenshots to help explain your problem.

Environment (please complete the following information):

  • OS: [Windows11家庭版美版22623.1255]

  • GCC/G++ Version [gcc (MinGW.org GCC-6.3.0-1) 6.3.0]

  • Python Version [3.8.16]

  • PaddlePaddle Version [paddlepaddle-gpu 2.4.1.post117]
    image

  • Model Version [?]声码器是pwgan_aishell3_static_1.1.0,模型是我在飞桨ai studio上代码块微调后导出的

  • GPU/DRIVER Informationo [e.g. Tesla V100-SXM2-32GB/440.64.00]

  • CUDA/CUDNN Version [cuda-11.7/cudnn8.8?or8.4]
    image
    image
    image

  • MKL Version

  • TensorRT Version

Additional context
我paddlespeech本来是使用源码编译安装的,会显示paddlespeech0.0.0,于是我用pip install paddlespeech==1.3重新安装的

@iftaken
Copy link
Collaborator

iftaken commented Feb 13, 2023

在命令行里面变量占用的显存在关闭前是不释放的,你把这部分的代码写在py文件里面,看看多次运行是否可以

@shiyv132
Copy link
Author

在命令行里面变量占用的显存在关闭前是不释放的,你把这部分的代码写在py文件里面,看看多次运行是否可以

感谢回复,是正常的就好,我是小白不懂写文件而且要经常合成新文本所以就重开吧,抱歉
还有就是paddlespeech源码编译安装的时候会自动安装版本0.0.0
我是按安装说明的命令git clone https://github.com/PaddlePaddle/PaddleSpeech.git
cd PaddleSpeech
pip install pytest-runner
pip install .
安装的,不知道什么问题,希望优化一下,没问题就没事了

@iftaken
Copy link
Collaborator

iftaken commented Feb 14, 2023

在命令行里面变量占用的显存在关闭前是不释放的,你把这部分的代码写在py文件里面,看看多次运行是否可以

感谢回复,是正常的就好,我是小白不懂写文件而且要经常合成新文本所以就重开吧,抱歉 还有就是paddlespeech源码编译安装的时候会自动安装版本0.0.0, 我是按安装说明的命令git clone https://github.com/PaddlePaddle/PaddleSpeech.git cd PaddleSpeech pip install pytest-runner pip install . 安装的,不知道什么问题,希望优化一下,没问题就没事了

这是正常的,使用源码编译的方式就是最新develop版本的0.0.0,如果是pip install paddlespeech 则会安装发行版本的,带有版本号,比如当前最新发行版本是1.3,后续会持续升级

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

3 participants