Hunyuan Video lora train (Currently only supports image dataset training, video dataset training is coming soon!)
datasetTips: hunyuan video lora train dataset zip 1. Need to include 20 to 40 images 2. The aspect ratio of each image should not exceed 2, and the optimal resolution is 512 x 512, 512 x 768, 768 x 512 . 3. caption file and image file name is the same, eg: 1.png corresponding to the caption 1.txt
trigger_word
Tips: hunyuan video lora train trigger_word, Trigger words are automatically added to the caption file. If the caption file already has the trigger word as a prefix, it will not be added repeatedly
Your Results
Feedback is welcome