Update README.md

README.md

@@ -89,6 +89,7 @@ license_link: https://freedevproject.org/faipl-1.0-sd/
 datasets:
 - p1atdev/pvc
 ---
+
 # pvcxl-v1-lora
 
 PVC style LoRA trained on [cagliostrolab/animagine-xl-3.0-base](https://huggingface.co/cagliostrolab/animagine-xl-3.0-base).
@@ -101,4 +102,26 @@ PVC style LoRA trained on [cagliostrolab/animagine-xl-3.0-base](https://huggingf
 
 Weights for this model are available in Safetensors format.
 
-[Download](/p1atdev/pvcxl-v1-lora/tree/main) them in the Files & versions tab.
+[Download](/p1atdev/pvcxl-v1-lora/tree/main) them in the Files & versions tab.
+
+## Training details
+
+### Dataset
+
+Curated high quality 2.5k pvc figure images from [p1atdev/pvc](https://huggingface.co/datasets/p1atdev/pvc)
+
+[p1atdev/pvc-quality-swinv2-base](https://huggingface.co/p1atdev/pvc-quality-swinv2-base) was used to curate images.
+
+### Training Config 
+
+|Configuration Item||
+|:-|-|
+|IaaS|Compute Engine of Google Cloud Platform|
+|GPU|1 x NVIDIA L4 (g2-standard-8)|
+|Dataset size|2576 images|
+|Batch size|4|
+|Training steps|1000+5000 (crashed and resumed)|
+|Train text encoder|False|
+|Image resolution|1024|
+|Optimizer|AdaFactor|
+|Learning rate|constant 1e-5 with 100 steps of warmup|