Alex2025Job comments

Results 5 comments of


                                            Alex2025Job

PoC for reading cuts in background thread in dynamic bucketing

Hi, Is this PR merged? Maybe I have similar problems, that reading the Cuts is not so fast.

SIGSEGV when running the example provided in the Readme

I have same issues when running example code. is there a lack of reserve the memory? can some help to fix the issue?

Training speed is not improved by using a better GPU

> > (Of course, the H100 has more memory than the A100, and we can use a larger number for “maximum duration”. This test is just to compare the performance...

Training speed is not improved by using a better GPU

> > Which torch/CUDA version you used for test? so for the training settings above (wenetspeech L), one step around 0.5s for both. H100 is slightly faster, but 0.36 is...

Training speed is not improved by using a better GPU

> > Which torch/CUDA version you used for test? so for the training settings above (wenetspeech L), one step around 0.5s for both. H100 is slightly faster, but 0.36 is...