We are continuing with the most open-source Stable Diffusion fine-tuning series. This time we decided to address the issues of the Vodka v3 model using custom LoRAs, and the results look interesting!
Check out the full post here https://followfoxai.substack.com/p/getting-started-with-loras-vodka

LoRA trained on realistic portrait shots on top of Vodka V3.
Adds photorealism to all kinds of generations when used in combination with the Vodka model.

Overall observations so far:

• The model can generate very high-quality photorealistic images

• It also introduced many issues; in some generations, fingers are now extra deformed, and we observe some forgetting of subjects they knew before.

• The model's diversity decreased, generating more similar images within the same prompt.