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

whisper在rk3588中推理的RTF为1.3左右,和介绍的RTF0.392不太一样 #180

Open
Gooddz1 opened this issue Sep 11, 2024 · 2 comments

Comments

@Gooddz1
Copy link

Gooddz1 commented Sep 11, 2024

-- init_whisper_decoder_model use: 221.897003 ms
-- read_mel_filters & read_vocab & read_audio use: 11.515000 ms
-- inference_whisper_model use: 7662.529785 ms

Whisper result: Mr. Quilter is the apostle of the middle classes and we are glad to welcome his gospel.

Real Time Factor (RTF): 7.663 / 5.855 = 1.309

@Gooddz1
Copy link
Author

Gooddz1 commented Sep 11, 2024

音频文件是model/test.wav

@jxx-xi
Copy link

jxx-xi commented Sep 12, 2024

需要定频后再测性能,请使用定频脚本(https://github.com/airockchip/rknn_model_zoo/blob/main/scaling_frequency.sh),用法:./scaling_frequency.sh -c rk3588

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