Model card auto-generated by SimpleTuner

README.md

@@ -58,9 +58,9 @@ You may reuse the base model text encoder for inference.
 
 ## Training settings
 
-- Training epochs: 35
-- Training steps: 250
-- Learning rate: 0.0003
+- Training epochs: 14
+- Training steps: 100
+- Learning rate: 0.0001
 - Effective batch size: 6
   - Micro-batch size: 6
   - Gradient accumulation steps: 1
@@ -73,7 +73,7 @@ You may reuse the base model text encoder for inference.
 - LoRA Rank: 64
 - LoRA Alpha: None
 - LoRA Dropout: 0.1
-- LoRA initialisation style: loftq
+- LoRA initialisation style: default
 
 
 ## Datasets