Welcome to Issue 5!
This is our weekly high level summary of everything happening in the Modverse across the AI Engine, Mojo, and our Community. Sign Up for the Newsletter Here!
- Mojo is coming to native macOS Apple Silicon tomorrow! 🔥 we'll host a livestream on Friday here
- Chris, Abul, and Jeff presented the keynote on Mojo at the LLVM Developer Conference 2023. Check out the slides here and stay tuned for the full YouTube presentation.
- Modular ❤️ PyTorch, we attended their conference this week and we're now a gold sponsor!
- We featured our first community blog post from Aydyn: How I Built llama2
We already have very significant demand for our EAP, but you can add yourself to our waiting list and we will try to reach out when new spots open.
An overview of language changes, relevant community projects, and the most common issues.
- Stijn Woestenborghs and Cedric Van Heck benchmarked Mojo against Jax, C++, and Python on gradient descent and wrote a blog post about it
- Aydyn created a blog post on benchmarking llama2.🔥 on macOS across different languages.
- Gabriel created a repo to fill in gaps between Mojo and the Python stdlib named mojo-stdlib-extensions with an example for running tests in GitHub Actions
- Traincheck released a blog post: Mojomics: Supercharging bioinformaticians with 🔥
- VMois created mojo-gym for short exercises related to Mojo
- polyglotengineer ran a workshop on Mojo and uploaded it to YouTube
- rd4com has been working on tutorials in their repo: mojo-learning
Most Common Issues
Please upvote tickets with a thumbs up that you want to see supported and fixed most!
Support for different operating systems still remain the most highly upvoted open tickets:
- #510: Support for macOS: most requested feature, support coming this Thursday!
- #620: Native Windows Support: coming after macOS, you can use WLS2 or docker containers for now
- #572: Can't install on arch linux: please see GitHub repo arch-mojo, for a community alternative until we have official support
Please upvote with a thumbs up if you're receiving the same error:
- #1002: Illegal instruction running minimal mojo program please report the output of lscpu to the ticket if you're receiving the same error
- #551: Unable to locate a suitable libpython when running Mojo executable you can use conda to install libpython and link Mojo to that learning resource here
- #670: Failed to reach URL, couldn't resolve host name VPN can resolve this issue in some cases
- #598: Failed to reach URL timeout was reached requires connection speed of 600kbs+ to download before timeout. Working on compressing artifacts and adding a download bar