bmaltais

What happen if you run the command manually in the activated venv: ``` accelerate launch --num_cpu_threads_per_process=2 "train_network.py" --pretrained_model_name_or_path="runwayml/stable-diffusion-v1-5" --train_data_dir="E:/pics/Eva/birme-512x512/processed/Rev lora/image" --resolution=512,512 --output_dir="E:/pics/Eva/birme-512x512/processed/EvaG lora/model" --logging_dir="E:/pics/Eva/birme-512x512/processed/Rev lora/log" --network_alpha="128" --save_model_as=safetensors --network_module=networks.lora --text_encoder_lr=5e-5 --unet_lr=0.0001...

Something appear to be wrong with your setup. Did you folow all the steps for the installation? It can't find train_db.py... but the path it is listing is wrong... it...

I have no idea. Best I can recommend at this point is to try to run `upgrade.ps1` to see if there a no modules missing or requiring updates.

Did this start with last night update? Was it working before? This is a bug in the kohya trainer code base. You need to open an issue on his repo...

did you complete the `accelerate config` step? Try running that again.

Change the bf16 to fp16 in the training parameters.

@Deejay85 Don't despair. It look like your card dows not support bitsandbytes for AdamW8bit. Change the optimizer to AdamW instead. That should get you passed this. If NVidia supported the...

I wish I could help more but it is hard to troubleshoot when you don't have the issue yourself ;-( Have you tried deleting the kohya_ss folder and then re-install...

It is complaining about socket access. Might be some sort of 3rd party antivirus blocking the access? Do you run a non Microsoft AV on your PC?

There is something weird on your system. The socket error is for a domain that does not even exist: www.007guard.com It is hard to tell what is the root cause......