Fine Tuning Video Generation Models | Make Your Own AI Videos
Training video models with… Images? Turns out video diffusion and image diffusion is similar enough that we can train the latest open source video generation models with a limited number of labeled images.
Resources:
Model Trained: https://huggingface.co/AdamLucek/Wan2...
Written Blog Tutorial: / fine-tuning-wan-2-1-with-a-curated-dataset...
Video Diffusion Blog: https://lilianweng.github.io/posts/20...
Diffusion Pipe: https://github.com/tdrussell/diffusio...
ComfyUI: https://github.com/comfyanonymous/Com...
Wan Paper: https://arxiv.org/pdf/2503.20314
Wan2.1-T2V-14B: https://huggingface.co/Wan-AI/Wan2.1-...
Chapters:
00:00 - Introduction
00:34 - How Video Generation Works
06:40 - Training: Dataset Preparation
11:42 - Training: Environment Setup
18:28 - Training: Configuration
23:38 - Training: Testing the Models
#ai #datascience #finetuning