pvcxl-v1-lora
PVC style LoRA trained on cagliostrolab/animagine-xl-3.0-base.
Download model
Weights for this model are available in Safetensors format.
Download them in the Files & versions tab.
Training details
Dataset
Curated high quality 2.5k pvc figure images from p1atdev/pvc
p1atdev/pvc-quality-swinv2-base was used to curate images.
Training Config
Configuration Item |
|
IaaS |
Compute Engine of Google Cloud Platform |
Machine type |
g2-standard-8 (8 vCPU, 32 GB RAM) |
GPU |
1 x NVIDIA L4 |
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 |