okay sir :4827spidermanok:
okay sir :4827spidermanok:
Fthanks for bringing into my attention here
FExplore and run machine learning code with Kaggle Notebooks | Using data from T5 prompts DeepFloyd IF
Ayeah its working perfect now
A
M@Furkan Gözükara SECourses can you help me solve this, ive been at it for 2 days i still havent been able to use kohya
M
MRuntimeError: CUDA out of memory. Tried to allocate 20.00 MiB (GPU 0; 8.00 GiB total capacity; 7.20 GiB already allocated; 0 bytes free; 7.29 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation.
Fyou need to enable memory optimizations
Flooks like you already did
Fwhat is your batch size?
Fit is also 1
Fi see
Fwhich torch version did you chose @mayen
HVery cool deepfloyd video 
HAlso I believe I can try your Lora video with colab @Furkan Gözükara SECourses
HThat notebook is a bit tricky to me
Fagreed
Factually i prefer gradio over these but they dont release and i dont get why
Fi even written them from twitter multiple times
Mi chose the one that was NOT 2.0, my batchzie was small wither 2 or 1 eveytime i ran it, 512 x 512 images
Fcould you have anything else that is consuming your GPU vram?
Fwhat is your GPU model
Mrtx 3070ti 8gb ram, no programs running
M
Fcan you try with clip skip 1? i dont know if that would make difference.
Mwhats a clip skip exactly?
Fit is about training weight vectors
Fyou made it 2
Fcould you show screenshots of the all settings if possible
M
Fmake clip skip 1 and check these 2
F
Falso are you using standard lora? what network size?
Msame problem
M
Tstill having same issue where regularization images come out looking like sketches and training ends up looking artifacted no matter how long/short the training session
Fcould it be trying to use your second gpu? amd?
Fwith all models?
For certain model
Tfollowed your videos to the tee, any model
Trealistic vision v20 and sd 1.5 are my main ive been trying
Fso do a fresh install
Fyour installation is problem
Twhich part?
Ti am getting no errors
Fdont know
Ti thought they installed to venvs
Ti followed your videos exactly, even installation
Ti can do it again