Skip to content

The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface. The fastest local inference engine in the world.

Support SimpleTuner lycoris lora for Qwen-Image (#9280)

PsychoLogicAu2208aa616d3a
diff --git a/comfy/lora.py b/comfy/lora.py
index 6686b7229..00358884b 100644
--- a/comfy/lora.py
+++ b/comfy/lora.py
@@ -301,6 +301,7 @@ def model_lora_keys_unet(model, key_map={}):
key_map["{}".format(key_lora)] = k
# Support transformer prefix format
key_map["transformer.{}".format(key_lora)] = k
+ key_map["lycoris_{}".format(key_lora.replace(".", "_"))] = k #SimpleTuner lycoris format
return key_map