search

lucasnewman / best-rq-pytorch

Implementation of BEST-RQ - a model for self-supervised learning of speech signals using a random projection quantizer, in Pytorch.
MIT License
71 stars 7 forks source link

changed three parameters: sample_rate win_length and hop_length #2

Open a897456 opened 9 months ago

a897456 commented 9 months ago

brq = BestRQ( sample_rate = 22050, win_length = 1024， hop_length = 256,

hi, I changed three parameters, as above, then the program may have crashed. so these three parameters can not be changed?

lucasnewman commented 9 months ago

The sample rate parameter should match the sample rate of your source audio — you'll need to resample your audio manually if it doesn't match what you specify when setting up the network. You should be able to modify the window length and hop length (aka stride) to meet your needs, though!

a897456 commented 9 months ago

The sample rate parameter should match the sample rate of your source audio — you'll need to resample your audio manually if it doesn't match what you specify when setting up the network. You should be able to modify the window length and hop length (aka stride) to meet your needs, though!

My training samples are all audio files with a sampling rate of 22050HZ, which are taken from LJSpeech-1.1, which are used on many GAN, where the win_length is 1024 and the hop_length is 256. So i changed the sample_rate the win_length and the hop length. but your program is not running, and I don't know how to modify it.