Model card auto-generated by SimpleTuner

README.md

@@ -134,9 +134,9 @@ You may reuse the base model text encoder for inference.
 
 ## Training settings
 
-- Training epochs: 499
-- Training steps: 3000
-- Learning rate: 0.00012
+- Training epochs: 7
+- Training steps: 500
+- Learning rate: 0.00013
 - Effective batch size: 2
   - Micro-batch size: 2
   - Gradient accumulation steps: 1
@@ -147,7 +147,7 @@ You may reuse the base model text encoder for inference.
 - Precision: Pure BF16
 - Quantised: No
 - Xformers: Not used
-- LoRA Rank: 4
+- LoRA Rank: 16
 - LoRA Alpha: None
 - LoRA Dropout: 0.1
 - LoRA initialisation style: default
@@ -156,7 +156,7 @@ You may reuse the base model text encoder for inference.
 ## Datasets
 
 ### default_dataset
-- Repeats: 0
+- Repeats: 10
 - Total number of images: 12
 - Total number of aspect buckets: 1
 - Resolution: 0.147456 megapixels