theluyuan/DiffSynth-Studio
1
0
Fork 0
mirror of https://github.com/modelscope/DiffSynth-Studio.git synced 2026-03-25 18:58:11 +00:00
Code Issues Packages Projects Releases Wiki Activity

support qwen-image blockwise controlnet training

Browse Source
This commit is contained in:
Artiprocher
2025-08-15 18:41:01 +08:00
parent 024fdad76d
commit 01a1f48f70
17 changed files with 269 additions and 15 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
examples/qwen_image/model_training/lora/Qwen-Image-Distill-Full.sh
View File

@@ -11,4 +11,5 @@ accelerate launch examples/qwen_image/model_training/train.py \
--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 \
--use_gradient_checkpointing
--use_gradient_checkpointing \
--find_unused_parameters