Attempt to recreate my Dark Cinematic Lora for videos.
Reran the training with the same data
dark_video_h3_1_0.safetensors