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

使用 ct2 轉換後掉辨識率 #52

Closed
ken19980727 opened this issue Jan 30, 2024 · 2 comments
Closed

使用 ct2 轉換後掉辨識率 #52

ken19980727 opened this issue Jan 30, 2024 · 2 comments

Comments

@ken19980727
Copy link

各位大老好

我用 ct2 轉換後 RTF 是有變快
但是WER上升明顯
請問各位有遇到這種情況嗎

我是用 whisper-large-v2 做 benchmark
common voice 的 WER 上升了 3%
指令我都是用版主提供的指令
不過我沒有加 quantization 這個 function
想問有人跟我遇到一樣的問題嗎

@ken19980727
Copy link
Author

package version:
transformers==4.36.2
peft==0.7.1
fast-whisper==0.10.0

@yeyupiaoling
Copy link
Owner

ct2会有损失吗?我也没有测试过。ct2官方也没有说明有损失。

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

No branches or pull requests

2 participants