Fizgig
A tool for LoRA training of AI image generation models, designed to run in cloud GPU environments and providing standardized configuration of training parameters.
Timeline (last 90 days)
The developer uses RunPod cloud GPUs running Fizgig to complete LoRA training for Krea 2.
The author's publicly shared standard training configuration is: Krea 2 Standard model, Standard LoRA, Network rank 32, Alpha 32, Epochs 64 (saving every 4 epochs), Batch size 1, target resolution 0.25 MP, EMA 0.98.
Fizgig is a LoRA training frontend tool designed specifically for the Krea series of models; it runs on cloud GPU platforms such as RunPod and encapsulates dataset preprocessing, training scheduling, and checkpoint management.
All Facts (3)
The developer uses RunPod cloud GPUs running Fizgig to complete LoRA training for Krea 2.
50%UnverifiedThe author's publicly shared standard training configuration is: Krea 2 Standard model, Standard LoRA, Network rank 32, Alpha 32, Epochs 64 (saving every 4 epochs), Batch size 1, target resolution 0.25 MP, EMA 0.98.
50%UnverifiedFizgig is a LoRA training frontend tool designed specifically for the Krea series of models; it runs on cloud GPU platforms such as RunPod and encapsulates dataset preprocessing, training scheduling, and checkpoint management.
50%