Samia Great Tensor Version is an alternative LoRA trained from an expanded dataset and optimized for use on TensorArt workflows.
Compared to the previous release, this version focuses on:
stronger identity preservation
more natural body proportions
improved lighting and skin texture
better performance in realistic and cinematic scenes
cleaner generations with fewer visual artifacts
It works especially well for lifestyle, travel, casual, and portrait-style images while keeping Samia’s distinctive look consistent across different poses, outfits, and camera angles.
This version can be used alone or combined with other style LoRAs for more creative results.