Zero training needed, Image-to-LoRA (i2L) V2
I2L Research has released Image-to-LoRA (i2L) V2, an updated version of their tool that generates style LoRAs from reference images without requiring explicit training. This new version can process one or more images in a single forward pass, making it applicable to various base models such as Z-Image, Klein-4B, and Hidream-O1. AI
IMPACT Simplifies the creation of custom image styles for AI art generation, potentially lowering the barrier to entry for users.