Run Damo's text-to-video-synthesis locally

Implementing the amazing work found here and adapted from the HuggingFace space implementation here - this repo serves to allow users to quickly test text-to-video-synthesis.

You will need a RTX 3090 or 4090 to run this model on your local machine as the model is fairly large.

Instructions:

1. Clone this repo using:

git clone https://github.com/chris-alexiuk/modelscope-damo-text-to-video-synthesis-dockerfile.git

2. Create the image with

docker build -t damo-txt2vid .

3. Run the container with

docker run --gpus all -v ${HOME}/.cache:/root/.cache -p 7860:7860 --shm-size 30g -it --rm txt2vid

🎉 That's it, enjoy! 🎉