a897456
a897456
他们其实想问Trainer.test()或者Trainer.validate()怎么用，尤其test.step()和validate_step()，怎么设计？
> 他们其实想问Trainer.test()或者Trainer.validate()怎么用，尤其test.step()和validate_step()，怎么设计？ 我猜他们想问这一套测试和验证的方法，因为你的main(),只涉及到fit（）训练
好像是相对过程，不能用了。
> 好像是相对过程，不能用了。  Part two, just quit the program
> I find a solution: try 'relative.XXX' instead of 'XXX' alone. Do you know what the corresponding parameter should be changed to when you change XXX？ I mean that such...
> In the usage there is a explanation: > > Just call the object returned by load with your estimated file (and your reference in case of relative metrics.) scores...
https://github.com/bastibe/python-soundfile/issues/388 same problem, but `librosa 0.10.1 requires soundfile>=0.12.1, but you have soundfile 0.12.0 which is incompatible.`
same problem+1
PS：4080 16G; dataset is LibriSpeech(29.3G); sr=16kHz; so how can I set this parameters: `batch_size` `num_workers` `num_iters_per_epoch` `dataset_type`
> The model is trained about 1.5 day on LibriTTS clean subset with an A800 GPU, and the batch size is 10240 tokens. Hi @ZhihaoDU 10240 tokens？How do you calculate...