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.

Sign up for GitHub

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

Jump to bottom

4090的显卡,这个速度正常吗? #1746

Open
femis opened this issue Nov 6, 2024 · 9 comments
Open

4090的显卡,这个速度正常吗? #1746

femis opened this issue Nov 6, 2024 · 9 comments

Comments

@femis
Copy link

femis commented Nov 6, 2024 •
edited
Loading

2
1
i9 处理器 4090 的显卡,这个速度正常吗?
训练的时候1左右
语音合成的时候 30 左右
@XXXXRT666
Copy link
Contributor

装的pytorch cpu吧

@femis
Copy link
Author

femis commented Nov 6, 2024

装的pytorch cpu吧

😅 不太懂啊,大佬求告知, 是我配置的不对吗, 反正有点慢.

@XXXXRT666
Copy link
Contributor

用整合包

@teneous
Copy link

teneous commented Nov 7, 2024

4090 的推理速度应该在80～130左右
1～2分钟的语音训练时间默认配置差不多2分钟结束。你需要看你下你tts_infer.yaml文件中的custom配置指定的device是cpu么？如果是cpu改为cuda

@femis
Copy link
Author

femis commented Nov 8, 2024

4090 的推理速度应该在80～130左右 1～2分钟的语音训练时间默认配置差不多2分钟结束。你需要看你下你tts_infer.yaml文件中的custom配置指定的device是cpu么？如果是cpu改为cuda

感谢大佬, 我前几天已经改成 cuda了, 目前速度提升了!, 我这有1-3并发情况下, 目前大概是 70 左右

@XXXXRT666
Copy link
Contributor

API_V2要是没记错的话并发不支持得排队,你可以尝试一下半精度

@femis
Copy link
Author

femis commented Nov 11, 2024

API_V2要是没记错的话并发不支持得排队,你可以尝试一下半精度

是不支持, 改成了启多个服务

@XXXXRT666
Copy link
Contributor

启动多个服务竞争资源可能会更慢

@femis
Copy link
Author

femis commented Nov 12, 2024

启动多个服务竞争资源可能会更慢

恩, 我启动了6个, 自己瞎改了一下, 把 new TTS() 缓存了起来, 目前速度还挺快的

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
3 participants
@femis @teneous @XXXXRT666