LoRA
LoRA is a novel technique in AI image generation, applying minimal tweaks to the standard models for significant enhancements. For example, Pixel-Art-XL is a LoRA adaptation for Stable Diffusion XL, specifically designed to enhance its pixel art generation capabilities. Using the same prompt, SDXL and SDXL+Pixl-Art-LoRA generate completely different images.
Shell

curl -X POST "https://api.hyperbolicai.xyz/v1/image/generation" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $HYPERBOLIC_API_KEY" \
-d '{
"model_name": "SDXL1.0-base",
"prompt": "a cute cat",
"height": 1024,
"width": 1024,
"backend": "auto",
"lora": {"Pixel_Art": 1.0}
}' | jq -r ".images[0].image" | base64 -d > result.jpg
We support various LoRAs for Stable Diffusion v1.5 and Stable Diffusion XL, each introducing unique functionalities to these models.

You can also mix LoRAs to customize your creations. Adjusting the weight of each LoRA enables us to control its influence, like combining Logo and Pixel Art LoRAs for a distinctive logo.

Last updated