that's tiny

big datasets have more way more than 1000

just start over

and do way less epoch this time

do i go below 50 epoch?

Ok, thank you.

Nope. still the same.

I made sure that the epoch was 50.

then you should improve the dataset

if you want you can also send it here so I can take a look

Could i send in DMs?

yes

Alrighty. gimme a min

Traceback (most recent call last): File "app.py", line 8, in from text2speech import T2S File "C:\Users\SS\Desktop\flask-tacotron2-tts-web-app-master\text2speech.py", line 13, in from waveglow.mel2samp import MAX_WAV_VALUE ModuleNotFoundError: No module named 'waveglow'

read the readme

i did

where should i put the waveglow model from the folder?

then this time with your eyes open

Also does anyone know why google keeps giving me a k80?

they want you to pay?

i'm trying to run the synthesizing notebook

Should I just run it locally?

that should be sufficient for synthesis

the k80 I mean

oh ok

I don't see why it shouldn't work

me to

maybe force using fp16 and then it should work

not sure