i am getting that error after restarting the run time is still the same

change your batch size to a lower one

example? i skipped that and move to the next page. and no error what so ever . would it effect the training ?

depends where you got the error

if in duration or pitch prediction then it might not be an issue

will uberduck will have the same option system much like 15.ai

DeepMoji emotional contextualizers

pretty sure deepmoji is now implemented, but on certain models only...

models trained under the pipeline support torchmoji i think

no clue about actually being able to control them in the app though

pikachu tested torchmoji on his multispeaker model and it barely did anything

imo the best way currently to do emotions is to separate them into speakers

personally i'm waiting for when/if audio-based gst gets implemented

like 1 speaker for each emotion question mark

yes

and also a speaker for question mark

When training TalkNet getting "no such file or directory, all files.json"

On Step 6.

Seem to be an issue with FFMPeg.

did you run the cells before that

Yeah. My friend was trying to train a model, and I was helping him out by giving him instructions. We think it might actually be the data set, but we're not 100% sure yet. I'm currently trying to train one and it seems to be working, so I'm assuming it's his dataset that's the problem.

this should be pinned

Quite pinteresting lol

OK, figured out training, but now I'm out of GPULMAO. Am I still able to synthesize on CPU?

Not DeepMoji, Torchmoji

Hey everyone does anyone know how to run tacotorn 2 model in python directly

?

Hey guys. How long do you think Highfi GAN should be trained for?

I don't think there's a limit. Go as long as you want. It will eventually get to a point where it won't get better, but that's up to you