ran the below command to create the LoRA
accelerate launch train_dreambooth_lora_sd3.py --pretrained_model_name_or_path=$MODEL_NAME --instance_data_dir=ft_dataset --output_dir=trained_lora2 --mixed_precision="fp16" --instance_prompt="a photo of v9xq man" --resolution=512 --train_batch_size=4 --gradient_accumulation_steps=4 --learning_rate=1e-4 --report_to="wandb" --lr_scheduler="cosine" --lr_warmup_steps=0 --max_train_steps=600 --validation_prompt="A photo of v9xq man wearing chainmail armor" --validation_epochs=40 --seed="0" --rank=128 --push_to_hub