Closed zhouhuirun2015 closed 4 years ago
Hi, It seems like it was caused by the out-of-memory issue, CPU ram. Reduce the batch size might solve the problem. You can reference the link https://ai-experiments.com/model-training-in-keras-and-python/ .
B&R Jagger
On Mon, Jun 8, 2020 at 2:28 PM zhouhuirun2015 notifications@github.com wrote:
clean audio is silent audio, noise audio is pink noise audio, 175MB/48000hz/1channel/S16le
while exec python rnn_train.py feature.h5, it always stop here [image: image] https://user-images.githubusercontent.com/15274287/83998878-250bb680-a994-11ea-9a34-4de43e7d6e2f.png
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/jagger2048/rnnoise-windows/issues/11, or unsubscribe https://github.com/notifications/unsubscribe-auth/AINFTAP3FE45DX63MAEAN4TRVSAJLANCNFSM4NYBLZGA .
Thanks for your reply. It did not work when i change batch size from 128 to 32. It did work to use a 16GB memory pc to train.
clean audio is silent audio, noise audio is pink noise audio, 175MB/48000hz/1channel/S16le
while exec python rnn_train.py feature.h5, it always stop here