andreae293

looks like an indentation problem did you modify the script? you may have messed up the indentation or something if i have to guess your code at line 855 looks...

my bad, the scripts should end at line 853 anything after is just copypasted garbage i did mistakenly please re-download again [main.py](https://github.com/andreae293/Dreambooth-Stable-Diffusion-cpu/blob/58cee8ff8f142604fe8c7696b2af828c72ee5b98/main.py) i checked this time and everything works correctly

im using python 3.8.5 but i had the same error using the original main.py now i deleted everything after line 853 and it works as it should, with no "return...

i'm glad it worked and sorry for the time wasting bug. out of curiosity, do you have a lot of RAM or are you using virtual RAM? how many seconds...

yeah i get it its always like that when working with ML..it takes a lot of time and effort to correctly setup the environment. let me know if you stumble...

i get 118 hours on A4000 with the following command, so i guess its normal `python finetune.py --base_model='decapoda-research/llama-7b-hf' --num_epochs=10 --cutoff_len=512 --group_by_length --output_dir='./lora-alpaca' --lora_r=16 --micro_batch_size=8 `

i too noticed heavy distorsion and artifact the more i trained, is there a way to avoid those artifacts?

but is this overfitting? the face gets totally distorted and does not resemble the original one at all those artifacts for me starts after 2000 steps ![01974-3460212814-a photo of a...

has any progress been made on this issue? i ran a confrontation between this repo and the xavier one same amount of steps, same regularization images, same training images, same...

> Hi. can you share a link to xavier repo ? https://github.com/XavierXiao/Dreambooth-Stable-Diffusion