TensorChord: 2022 in Review
In just six months, TensorChord has accomplished amazing feats with the launch of the first open source project
envd! To celebrate this success and look ahead to 2023, we wanted to share our journey so far and provide insight into what's in store for us over the next year!
On June 18th, 2022
envd officially opened its doors to the public. We were overwhelmed with feedback from our community and we could not be more delighted about it! Since then, we've had 100+ Pull Requests in just over 4 months - a remarkable achievement by any standard. Even better news: With 1024 GitHub stars in under 5 months since open source release.
It's been an amazing year for
envd! We've increased our incredible contributor base by nearly 60 people from all around the globe, and have gained a whopping 1,400 stars on GitHub. To top it off, 85 releases with three major updates including various new features and improvements have been released. We’ve also had in excess of 20,000 Docker Hub image downloads so far - giving us over 20k successful development environment builds this past 6 months alone.
We couldn't have done it without our amazing community! Our roadmap and contributor guide are constantly being updated to ensure that everyone is on the same page when it comes to building
envd together. Thank you all for your support -- we truly appreciate it.
At TensorChord, we strive to be open and transparent in all aspects of our work. To extend this ethos even further, we've made available a set of internal guidelines (https://internals.tensorchord.ai/) that outline how we help our new members onboard - so people can benefit from understanding more about us! We welcome everyone interested in learning or contributing.
envd takes the hassle out of setting up a development environment for AI/ML. Don't get bogged down in tedious configuration - let
envd make it easy to focus on your work!
Development environments are full of python and system dependencies, CUDA, BASH scripts, Dockerfiles, SSH configurations, Kubernetes YAMLs, and many other clunky things that are always breaking.
envd is to solve the problem.
envd, your development process can be more efficient. It enables you to create tight-knit and easily reusable workflows with just a few lines of code. In no time at all, running 'envd up' will have you training models in an isolated environment - tapping into the power of modern development tools without any extra hassle.
def build(): base(os="ubuntu20.04", language="python3") # Configure the pip index if needed. # config.pip_index(url = "https://pypi.tuna.tsinghua.edu.cn/simple") install.python_packages(name = [ "numpy", ]) shell("zsh")
As we enter 2023, there is still plenty of work for us to be done in order to fully optimize the system. We are dedicating our focus to build speed, as well as fine-tuning and serving large models - all essential tasks that will help drive further progress this year!
Not only that, but get ready to experience the power of
envd with our brand-new online demo server running on Kubernetes! If you're already using it, tell us about the experience at Discord.
It's no secret that the machine learning process still needs a few tweaks here and there. But don't worry - we won't stop until it runs like an absolute dream! With any luck, 2023 holds something special in store for us all – so stay tuned to see what exciting new developments we can bring your way.