FastVideo is a lightweight framework for accelerating large video diffusion models.

FastMochi-Demo.mp4

🤗 FastMochi | 🤗 FastHunyuan | 🔍 Discrod

FastVideo currently offers: (with more to come)

• FastHunyuan and FastMochi: consistency distilled video diffusion models for 8x inference speedup.
• First open distillation recipes for video DiT, based on PCM.
• Support distilling/finetuning/inferencing state-of-the-art open video DiTs: 1. Mochi 2. Hunyuan.
• Scalable training with FSDP, sequence parallelism, and selective activation checkpointing, with near linear scaling to 64 GPUs.
• Memory efficient finetuning with LoRA, precomputed latent, and precomputed text embeddings.

Dev in progress and highly experimental.

Change Log

• 2024/12/17: FastVideo v0.1 is released.

🔧 Installation

The code is tested on Python 3.10.0, CUDA 12.1 and H100.

./env_setup.sh fastvideo

🚀 Inference

We recommend using a GPU with 80GB of memory. To run the inference, use the following command:

FastHunyuan

# Download the model weight
python scripts/huggingface/download_hf.py --repo_id=FastVideo/FastHunyuan --local_dir=data/FastHunyuan --repo_type=model
# CLI inference
sh scripts/inference/inference_hunyuan.sh

You can also inference FastHunyuan in the official Hunyuan github.

FastMochi

# Download the model weight
python scripts/huggingface/download_hf.py --repo_id=FastVideo/FastMochi-diffusers --local_dir=data/FastMochi-diffusers --repo_type=model
# CLI inference
bash scripts/inference/inference_mochi_sp.sh

FastHunyuan-Demo.mp4

Distillation

Please refer to the distillation guide.

Finetuning

Please refer to the finetuning guide.

Acknowledgement

We learned and reused code from the following projects: PCM, diffusers OpenSoraPlan, and xDiT.

We thank MBZUAI and Anyscale for their support throughout this project.