I see that this code example only has the training part, and the post-training part is not used. After this execution, accelerate launch main.py \
--concept_image_dir="./examples/concept_image" \
--content_image_dir="./examples/content_image" \
--output_image_path="./outputs" \
--initializer_token="girl" \
--max_train_steps=500 \
--concept_embedding_num=3 \
--cross_attention_injection_ratio=0.2 \
--self_attention_injection_ratio=0.9 \
--use_l1 How to use the trained model
I see that this code example only has the training part, and the post-training part is not used. After this execution, accelerate launch main.py \ --concept_image_dir="./examples/concept_image" \ --content_image_dir="./examples/content_image" \ --output_image_path="./outputs" \ --initializer_token="girl" \ --max_train_steps=500 \ --concept_embedding_num=3 \ --cross_attention_injection_ratio=0.2 \ --self_attention_injection_ratio=0.9 \ --use_l1 How to use the trained model