which config you use and where do you train?
which config you use and where do you train?
DI train locally on a 3090, I use your last 24gb slow preset, I was changing some parameters like crossatention for a speed up with no gradient checkpointing, I dont remember changing anything else
Dother questions did you tried this new learning rates on the base model for a run at least, may be works perfect on realvis4 but is too high for the base model
Fif you are using latest it shouldnt get cooked
Fi trained 4500 steps with it
Ffor base model yes could be too high
Fyou can reduce LR of UNET to 6e-06
Fbut i havent compared yet
Dthanks for your great work, may be, be good if you make a run with your dataset, same settings and samples on the base model to be sure that is consistent on all scenarios.
Ftutorial recorded for download upload cloud
PI've tried it and it's still installing, it's been installing for 3 hours on runpod rtx4090. Surely it shouldn't take so long?
Fso many bad pods they have
Fwe have kohya on massed compute
Fit is latest version
POh I see. How long should it take?
Finstall should take like maximum 10 - 15 min
PI can't believe i've waited 5 hours
thank you
thank youFdmn just for install?
Fhorrific
PYes, the comfy ui tutorial you linked for me earlier ha. I think I'll wait for your video tutorial on how to upload lora safetensor from huggingface to runpod
Fi am editing it
Fmade a perfect tutorial
Rhello Can you make one click for CodeFormer I am a photographer and I just want a sample gradio like the one huggingface I will be a Gold Supporter if you make it
Fsure let me make right now :d
Fyou can get membership working
![Dazzastrous [4090]](https://cdn.discordapp.com/avatars/871389071686135838/c4a5f4d81d9eeb18d856295d692e2c07.webp?size=40)
Doh it works well
Finstaller almost ready testing
Fwow this code is old :d
Fneed to fix some dependencies
Fi think finally made it
Fwow taking longer than i expected
Fapp doesnt give any error but nothing happens trying to figure out
Fyes finally made it
F
Fman the guy had so many errors 
Fi need to add auto save and open outputs folder
Fhe didnt even made those logic
Fhere published
FF
F
F
Fall generations auto saved
Fworks with latest gradio man it took huge time
Fi will hopefully add batch processing as well later, input output folder
AI think doctor is the wrong title furkan
Ayou may be a mad scientist
A
F