Model card auto-generated by SimpleTuner

README.md

@@ -61,6 +61,8 @@ This is a LoRA derived from [black-forest-labs/FLUX.1-dev](https://huggingface.c
 
 The main validation prompt used during training was:
 
+
+
 ```
 julie, in photograph style
 ```
@@ -86,17 +88,17 @@ You may reuse the base model text encoder for inference.
 
 ## Training settings
 
-- Training epochs: 63
-- Training steps: 700
-- Learning rate: 1.0
-- Effective batch size: 1
+- Training epochs: 147
+- Training steps: 2500
+- Learning rate: 1e-05
+- Effective batch size: 2
   - Micro-batch size: 1
-  - Gradient accumulation steps: 1
+  - Gradient accumulation steps: 2
   - Number of GPUs: 1
 - Prediction type: flow-matching
 - Rescaled betas zero SNR: False
-- Optimizer: Prodigy
-- Precision: no
+- Optimizer: AdamW, stochastic bf16
+- Precision: Pure BF16
 - Xformers: Not used
 - LoRA Rank: 16
 - LoRA Alpha: 16.0
@@ -108,9 +110,9 @@ You may reuse the base model text encoder for inference.
 
 ### julia
 - Repeats: 0
-- Total number of images: 11
+- Total number of images: 34
 - Total number of aspect buckets: 1
-- Resolution: 1.0 megapixels
+- Resolution: 512 px
 - Cropped: True
 - Crop style: random
 - Crop aspect: square