Skip to content

Experiments with tokenizing video, audio, and image

Notifications You must be signed in to change notification settings

MaxGubin/video_encoders

Repository files navigation

Introduction

"From scratch" implementations of image processing algorithms using Jax

Goals

  1. Learn about the algorithms (my background is NLP, learning the new area)
  2. Refresh my knowledge of the latest changes in JAX
  3. Have fun
  4. This is not production quality implementations, the focus is on simplicity and algorithm understanding

Reproducing

Everything was tried using Ubuntu 24.04 CUDA 12.7

Setting up and running

python3 -m venv .venv:wa
source .venv/bin/activate
pip install -r requirements.txt
jupyter lab

Using Docker

docker run --runtime=nvidia -p 8888:8888 -it jax-cuda

About

Experiments with tokenizing video, audio, and image

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages