KreakxX

Encountered it, even with the prefix and the new code, but after resuming a checkpoint it works. Pls help

accelerate launch --mixed_precision="fp16" train_dreambooth_lora.py --pretrained_model_name_or_path="stabilityai/stable-diffusion-2-1" --instance_data_dir="C:/Users/Henri/logo-generator/Dataset/images" --output_dir="C:/Users/Henri/logo-lora-output" --resolution=512 --train_batch_size=1 --mixed_precision="fp16" --gradient_accumulation_steps=1 --learning_rate=3e-5 --lr_scheduler="constant" --lr_warmup_steps=0 --max_train_steps=2500 --checkpointing_steps=750 --validation_prompt="a henrikstyle logo for a tech company, minimal flat design, white background" --instance_prompt="a logo...