tuwonga / fast_dreambooth_4_kaggle Goto Github PK

EDIT: NVM

bro every code is getting errors literally. I hope you will fix it

I get this no matter what my input is

Only replaced the "insert your" parts but get the following

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
/tmp/ipykernel_23/305713317.py in <module>
    135 
    136   txtenc_train(Caption, stpsv, stp, MODELT_NAME, INSTANCE_DIR, CLASS_DIR, OUTPUT_DIR, PT, Seed, precision, Training_Steps=stptxt)
--> 137   unet_train(Caption, stpsv, stp, MODELT_NAME, INSTANCE_DIR, OUTPUT_DIR, PT, Seed, precision, Training_Steps)
    138 
    139 else:

TypeError: unet_train() missing 1 required positional argument: 'Training_Steps'

I've had great success with training on 1.5. Just tried to train on the v2 model (512) and I keep getting the following error on the training cell.

"No model found, use the "Model Download" cell to download a model."

I've tried using a link from the model on my google drive, i've tried directly linking to the huggingface CKPT (https://huggingface.co/stabilityai/stable-diffusion-2-base/resolve/main/512-base-ema.ckpt)

But i'm obviously doing something wrong.

Any help would be appreciated, thank you.

at first it with stuck at download model with an error(run with Path_to_HuggingFace=)

if i edit code to skip this error, it with have someting wrong when i train

and this is my setting on downlaod model

No matter what I do I still get this message when it comes time to load the training images. I tried everything I coould think of or google.

Hi there,

Was using your fantastic dreambooth port on here for a few weeks perfectly, but it seems like i'm having issues since SD2.0 and various other updates.

I'm trying to train on 1.5 and every time I get an error when it gets to the training cell.

I feel like i'm inputting all my session name, instance name etc in all the right places, as I was before, but I must be doing something wrong.

Any advice would be appreciated.

Thanks!