search

Aria-K-Alethia / BigCodec

Official implementation of the paper "BigCodec: Pushing the Limits of Low-Bitrate Neural Speech Codec"
https://aria-k-alethia.github.io/bigcodec-demo/
MIT License
86 stars 4 forks source link

a quetion about scaling up the model size #9

Open VJJJJJJ1 opened 1 week ago

VJJJJJJ1 commented 1 week ago

hi, thank you for your great work. Do you increase the model parameters simply by increasing the number of conv layers while keeping the sample rate constant, thereby improving performance? Have you ever tried to increase N to 6, 7 or more?

Aria-K-Alethia commented 1 week ago

Hi,

Yes, we increase parameters while keeping the sample rate constant to improve performance. I haven't tried larger N, but you are welcomed to try it.