stuck kraken
Last night I tried to set up my Dataset for training only for the entire thing to backfire on me, thanks to me not knowing that I needed to save the Training around the 100th epoch mark
When I tried to start over again from scratch it says i've reached GPU limit and now have to wait a whole 24 hours to try again
Does anyone know an alternative?
sharp reef
use a local fork if you have decent Nvidia GPU
idk about the paperspace alternative but give it a try
stuck kraken
about that my specs are
AMD Ryzen 3 2200G with Radeon Vega Graphics 3.50 GHz
and RX 570 8gb graphics card
idk if that translates well to Nvidia
sharp reef
about that my specs are
AMD Ryzen 3 2200G with Radeon Vega Graphics 3.50 G...
not applies for local training
stuck kraken
so it wont work on my pc?
thorny hull
Last night I tried to set up my Dataset for training only for the entire thing t...
you can use an alt google account and continue
Be sure to not put high epochs for the save frequency
stuck kraken
just keep it around 500
but if i saved it around say the 100th epoch and left for a while to do stuff
and if i try to resume by loading it, and it wont work what next?
thorny hull
just keep it around 500
but if i saved it around say the 100th epoch and left f...
no no i mean, put save frquency to 50, you can put as 500 the epochs to train, btw in google drive you should have a rvcdisconnected folder which inside you have the model name, if you have that, you can follow this https://docs.aihub.wtf/rvc/cloud/rvc-disconnected/#resuming-, also remember you can resume only on the same google acc
stuck kraken
alright i'll give it a try... After 24 hours
thorny hull
you sure you runned everything as said in that guide?
it seems like u didnt run the training variables cell
stuck kraken
i did it first as the instruction said
thorny hull
Last night I tried to set up my Dataset for training only for the entire thing t...
if you want there are other google colabs for training inside the cloud folder inside the rvc folder in our docs https://docs.aihub.wtf like applio
thorny hull
i did it first as the instruction said
kinda weird honestly, you sure your using colab gpu btw
stuck kraken
kinda weird honestly, you sure your using colab gpu btw
yes and the day before i discovered it had limited uses before reaching max limit
thorny hull
you could just try to disconnect and delete runtime and re run everything from the start
sharp reef
free T4 gpu is only available for a session daily, so if you get disconnected or do "disconnect runtime", the next runtime won't use T4 until tomorrow
though doing restart runtime still keeps T4
thorny hull
free T4 gpu is only available for a session daily, so if you get disconnected or...
no, it isnt only a daily session
it depends from how much you abuse it
technically on their site they say 12 hours daily, which depends on the usage so in reality its 4 hours for things like this
but depends cus if u use ot every single day every single month they give less time, so yea, but its not like if u connect 1 min and accidentally disconnect they dont have the free daily gpu till tmr
stuck kraken
Anyways I'll try it again tomorrow since I'm vacationing with my family
I just hope someone will be here and responds before I begin
sharp reef
technically on their site they say 12 hours daily, which depends on the usage so...
not sure if it has changed to 3 hours daily, and might become less when done more usages monthly
thorny hull
not sure if it has changed to 3 hours daily, and might become less when done mor...
depends from user to user
stuck kraken
update it finally finished
but it says i need to download the file thats closest to before the over training point?
so far i downloaded 4 files
I set training as
save_frequency 20
total epoch 500
batch size 8
which one of the 4 i downloaded should i use?
