Beta test Llama3 serving and GUI on MAX Nightly
You can beta test the upcoming MAX serving and Llama3 chatbot GUI using the
nightly
release, I'd love feedback if you run into any problems.
Start the GUI by running these commands on arm64 macOS
or x86/arm64 Linux
(Note the smallest llama3 model is 4.5GB):
```bash
rm -rf ~/.modular...mojonet
PyTorch Neural Network wrapper to help transition existing projects into the Mojo programming language. https://github.com/mytechnotalent/mojonet
#🔥mojo...
Simple little JSON parser
https://github.com/ZacHooper/mojo-json
Trying to follow Python's json module API to make it easier to convert Python code into Mojo....
Gojo: Experiments in porting over Golang stdlib into Mojo
https://github.com/thatstoasty/gojo/tree/main
This library is my attempt at porting over some modules and structs from Golang's stdlib. You can find the list below! Most of these topics were new to me when starting the port, so if you find bugs or opportunities for improvement feel free to drop some comments!
It's been a good learning opportunity for both Go and Mojo....
Small typestate for heap using parameters
https://github.com/rd4com/small_typestate
POC, to cultivate ideas about parameters...
Mojo in the Tiobe-index
Arguments can be made about the usefulness of the Tiobe index, but the measure of Mojo's popularity in it grew from #184 in Nov 2023 to somewhere in the range from #51 to #100 in Jun 2024 (the index doesn't specify the position in that range).
This is an amazing growth-rate in only 7-8 months, many interesting languages (for example Pony) never made it that far. Mojo certainly has the potential to raise to the range 20-50 (where we will know the exact position) and beyond, we'll notify you when it does....
KANs in Mojo - second attempt
This week I finally found time to dive into Kolmogorov–Arnold Networks again and give a Mojo implementation another shot.
https://github.com/dorjeduck/kamo
I decided to scrap my original source code and port A from-scratch implementation of Kolmogorov-Arnold Networks (KAN)…and MLP to Mojo....
Specials: Accurate, Hardware Accelerated, Special Functions in Mojo
Specials is a Mojo package designed to provide highly-optimized and hardware-acceleration-friendly special functions implementations for AI computing.
https://github.com/leandrolcampos/specials...
Mist and Weave: ANSI Styling and text formatting libraries
Mist
mist
lets you safely use advanced styling options on the terminal. It offers you convenient methods to colorize and style your output, without you having to deal with all kinds of weird ANSI escape sequences and color conversions.
https://github.com/thatstoasty/mist
```rust...MAX âš¡ with quantization and macOS support thread
Launch blog: https://www.modular.com/blog/max-24-4-introducing-quantization-apis-and-max-on-macos
Post any issues you're having with the new release or macOS support here. You can get started installing here: https://modul.ar/install-max and running the new llama3 quantization example here: https://modul.ar/llama3...
Setup Mojo on Mac M1 and use Python integration without crash
Some developers setup Mojo on Mac M1 but when they try to use Python integration they will get crash so I write blog how to setup Mojo on Mac M1 and use Python integration without crash
https://medium.com/@maicmi/install-mojo-on-mac-m1-and-run-python-integration-with-no-crash-d730e9965d60...
A Complex Math Library for Mojo
Hi folks, I've had way too much free time recently, so I wrote a complex math library for Mojo! It massively expands the existing
ComplexSIMD
type as well as adding all the standard functions like csqrt
, csin
, cexp
, and so on.
Of course, it's still a WIP, but if you want more let me know! I'm trying to align this as close to the stdlib as possible so it feels natural to use.
https://github.com/JakeEBrooks/mojocmath...Quick and Dirty JSON parser
I hacked together a JSON parser.
https://gist.github.com/mzaks/c1947212fd23ba010129d2f6b7c0fdfa
As it is just a hack I will not even put it into a proper repo, but I think it might be useful as a learning exercise, or for some other hacks 🙂...
NuMojo Moved into a GitHub Organization Updated to Include access to different backends
NuMojo now exists under the Mojo Numerics and Algorithms group (MoNA).
For those of you that haven't heard of NuMojo, we are essentially trying to fill the same niche as numpy does for python. So far we have Tensor input versions of many of the standard library math functions. In the future we will have our own Tensor type and cover most, if not all, of the functionality of numpy and much of the functionality of scipy.
As for our new backend functionality, we added a number of function-only structs that conform to a trait called
Backend
. From my testing, Vectorized is the fastest and is therefore the default backend....