FunASR 批量处理音频事后，我每次执行asr的时候，都会执行load sv model params 这个操作

批量处理音频的时候，每个音频都会都会执行load sv model params 这个操作，模型不是已经加载到显存了吗？这个操作会消耗大量的时间，如何通过设置，不再执行这个操作呢？

Dec 29 '23 10:12 aofengdaxia

Fixed another typo signaled by issue #30 in the CI/CD Pipeline image.

Since i committed to my same branch fix-typo i couldn't create a new pull request so I am commenting here

Oct 23 '23 19:10 AmineAouragh

Fixed another typo signaled by issue #37 in the Monorepo vs Microrepo image

Oct 23 '23 19:10 AmineAouragh

Wondering, how do you usually fix typos on images? @ AmineAouragh

Just wondering how to do it so I can start my own contributions here as well.

Thanks!

Oct 26 '23 05:10 salvadorgonmo

Hey @salvadorgonmo yeah sure

i use Canva to remove the typo and replace with a word, making sure to find the closest font possible to the one in the image

Oct 26 '23 11:10 AmineAouragh

Nice work, @AmineAouragh! But, according to the CONTRIBUTING, I believe you should open an issue for fixing diagrams.

If you notice typos or issues in diagrams, please open a separate issue instead of fixing the images directly. We will update the source images and republish fixes.

https://github.com/ByteByteGoHq/system-design-101/blob/main/CONTRIBUTING.md

Oct 26 '23 19:10 matheusaraujo

批量处理音频事后，我每次执行asr的时候 ，都会执行load sv model params 这个操作

批量处理音频事后，我每次执行asr的时候，都会执行load sv model params 这个操作