trained on less images than the other. Feel like it works better with other loras than the other version. especially character ones. I only use this one now just because i usually use more than just this lora.
Iphone_quality_FLUX_v2 (1).safetensors