| modelscope download --dataset DiffSynth-Studio/diffsynth_example_dataset --include "qwen_image/Qwen-Image-Blockwise-ControlNet-Inpaint/*" --local_dir ./data/diffsynth_example_dataset | |
| accelerate launch examples/qwen_image/model_training/train.py \ | |
| --dataset_base_path data/diffsynth_example_dataset/qwen_image/Qwen-Image-Blockwise-ControlNet-Inpaint \ | |
| --dataset_metadata_path data/diffsynth_example_dataset/qwen_image/Qwen-Image-Blockwise-ControlNet-Inpaint/metadata.csv \ | |
| --data_file_keys "image,blockwise_controlnet_image,blockwise_controlnet_inpaint_mask" \ | |
| --max_pixels 1048576 \ | |
| --dataset_repeat 50 \ | |
| --model_id_with_origin_paths "Qwen/Qwen-Image:transformer/diffusion_pytorch_model*.safetensors,Qwen/Qwen-Image:text_encoder/model*.safetensors,Qwen/Qwen-Image:vae/diffusion_pytorch_model.safetensors,DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Inpaint:model.safetensors" \ | |
| --learning_rate 1e-4 \ | |
| --num_epochs 5 \ | |
| --remove_prefix_in_ckpt "pipe.dit." \ | |
| --output_path "./models/train/Qwen-Image-Blockwise-ControlNet-Inpaint_lora" \ | |
| --lora_base_model "dit" \ | |
| --lora_target_modules "to_q,to_k,to_v,add_q_proj,add_k_proj,add_v_proj,to_out.0,to_add_out,img_mlp.net.2,img_mod.1,txt_mlp.net.2,txt_mod.1" \ | |
| --lora_rank 32 \ | |
| --extra_inputs "blockwise_controlnet_image,blockwise_controlnet_inpaint_mask" \ | |
| --use_gradient_checkpointing \ | |
| --find_unused_parameters | |