[Docs]: Valle_v2 pretrained model info

open-mmlab / Amphion

Amphion (/æmˈfaɪən/) is a toolkit for Audio, Music, and Speech Generation. Its purpose is to support reproducible research and help junior researchers and engineers get started in the field of audio, music, and speech generation research and development.

https://openhlt.github.io/amphion/

MIT License

4.45k stars 379 forks source link

[Docs]: Valle_v2 pretrained model info #259

Closed acul3 closed 1 month ago

acul3 commented 1 month ago

hey @jiaqili3 or anyone involve

do you have information about how many epoch your valle_v2 train ?

valle_ar_mls_196000.bin

valle_nar_mls_164000.bin

( i assume 196000, 164000 is step) ?

a tensorboard loss could also be helpfull, thank you

jiaqili3 commented 1 month ago

Hi @acul3, yes the 190k and 160k are steps, and each model has been trained on about 1 epoch. For the loss curve, it's a bit hard to find, but we'll post loss curves for future model releases. Thanks!

acul3 commented 1 month ago

thank for the reply and confirmation