metaseq
jax
metaseq | jax | |
---|---|---|
53 | 82 | |
6,389 | 28,082 | |
0.4% | 1.8% | |
6.2 | 10.0 | |
11 days ago | about 4 hours ago | |
Python | Python | |
MIT License | Apache License 2.0 |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
metaseq
-
Training great LLMs from ground zero in the wilderness as a startup
This is a super important issue that affects the pace and breadth of iteration of AI almost as much as the raw hardware improvements do. The blog is fun but somewhat shallow and not technical or very surprising if you’ve worked with clusters of GPUs in any capacity over the years. (I liked the perspective of a former googler, but I’m not sure why past colleagues would recommend Jax over pytorch for LLMs outside of Google.) I hope this newco eventually releases a more technical report about their training adventures, like the PDF file here: https://github.com/facebookresearch/metaseq/tree/main/projec...
- Chronicles of Opt Development
-
See the pitch memo that raised €105M for four-week-old startup Mistral
The number of people who can actually pre-train a true LLM is very small.
It remains a major feat with many tweaks and tricks. Case in point: the 114 pages of OPT175B logbook [1]
[1] https://github.com/facebookresearch/metaseq/blob/main/projec...
- Technologie: „Austro-ChatGPT“ – aber kein Geld zum Testen
- OPT (Open Pre-trained Transformers) is a family of NLP models trained on billions of tokens of text obtained from the internet
- Current state-of-the-art open source LLM
-
Elon Musk Buys Ten Thousand GPUs for Secretive AI Project
Reliability at scale: take a look at the OPT training log book for their 175B model run. It needed a lot of babysitting. In my experience, that scale of TPU training run requires a restart about once every 1-2 weeks—and they provide the middleware to monitor the health of the cluster and pick up on hardware failures.
-
Is AI Development more fun than Software Development?
I really appreciated this log of Facebook training a large language model of how troublesome AI development can be: https://github.com/facebookresearch/metaseq/tree/main/projects/OPT/chronicles
-
Visual ChatGPT
Stable Diffusion will run on any decent gaming GPU or a modern MacBook, meanwhile LLMs comparable to GPT-3/ChatGPT have had pretty insane memory requirements - e.g., <https://github.com/facebookresearch/metaseq/issues/146>
-
Ask HN: Is There On-Call in ML?
It seems so, check this log book from Meta: https://github.com/facebookresearch/metaseq/blob/main/projec...
jax
-
The Elements of Differentiable Programming
The dual numbers exist just as surely as the real numbers and have been used well over 100 years
https://en.m.wikipedia.org/wiki/Dual_number
Pytorch has had them for many years.
https://pytorch.org/docs/stable/generated/torch.autograd.for...
JAX implements them and uses them exactly as stated in this thread.
https://github.com/google/jax/discussions/10157#discussionco...
As you so eloquently stated, "you shouldn't be proclaiming things you don't actually know on a public forum," and doubly so when your claimed "corrections" are so demonstrably and totally incorrect.
-
Julia GPU-based ODE solver 20x-100x faster than those in Jax and PyTorch
On your last point, as long as you jit the topmost level, it doesn't matter whether or not you have inner jitted functions. The end result should be the same.
Source: https://github.com/google/jax/discussions/5199#discussioncom...
-
Apple releases MLX for Apple Silicon
The design of MLX is inspired by frameworks like NumPy, PyTorch, Jax, and ArrayFire.
-
MLPerf training tests put Nvidia ahead, Intel close, and Google well behind
I'm still not totally sure what the issue is. Jax uses program transformations to compile programs to run on a variety of hardware, for example, using XLA for TPUs. It can also run cuda ops for Nvidia gpus without issue: https://jax.readthedocs.io/en/latest/installation.html
There is also support for custom cpp and cuda ops if that's what is needed: https://jax.readthedocs.io/en/latest/Custom_Operation_for_GP...
I haven't worked with float4, but can imagine that new numerical types would require some special handling. But I assume that's the case for any ml environment.
But really you probably mean fixed point 4bit integer types? Looks like that has had at least some work done in Jax: https://github.com/google/jax/issues/8566
-
MatX: Efficient C++17 GPU numerical computing library with Python-like syntax
>
Are they even comparing apples to apples to claim that they see these improvements over NumPy?
> While the code complexity and length are roughly the same, the MatX version shows a 2100x over the Numpy version, and over 4x faster than the CuPy version on the same GPU.
NumPy doesn't use GPU by default unless you use something like Jax [1] to compile NumPy code to run on GPUs. I think more honest comparison will mainly compare MatX running on same CPU like NumPy as focus the GPU comparison against CuPy.
[1] https://github.com/google/jax
-
JAX – NumPy on the CPU, GPU, and TPU, with great automatic differentiation
Actually that never changed. The README has always had an example of differentiating through native Python control flow:
https://github.com/google/jax/commit/948a8db0adf233f333f3e5f...
The constraints on control flow expressions come from jax.jit (because Python control flow can't be staged out) and jax.vmap (because we can't take multiple branches of Python control flow, which we might need to do for different batch elements). But autodiff of Python-native control flow works fine!
-
Julia and Mojo (Modular) Mandelbrot Benchmark
For a similar "benchmark" (also Mandelbrot) but took place in Jax repo discussion: https://github.com/google/jax/discussions/11078#discussionco...
-
Functional Programming 1
2. https://github.com/fantasyland/fantasy-land (A bit heavy on jargon)
Note there is a python version of Ramda available on pypi and there’s a lot of FP tidbits inside JAX:
3. https://pypi.org/project/ramda/ (Worth making your own version if you want to learn, though)
4. For nested data, JAX tree_util is epic: https://jax.readthedocs.io/en/latest/jax.tree_util.html and also their curry implementation is funny: https://github.com/google/jax/blob/4ac2bdc2b1d71ec0010412a32...
Anyway don’t put FP on a pedestal, main thing is to focus on the core principles of avoiding external mutation and making helper functions. Doesn’t always work because some languages like Rust don’t have legit support for currying (afaik in 2023 August), but in those cases you can hack it with builder methods to an extent.
Finally, if you want to understand the middle of the midwit meme, check out this wiki article and connect the free monoid to the Kleene star (0 or more copies of your pattern) and Kleene plus (1 or more copies of your pattern). Those are also in regex so it can help you remember the regex symbols. https://en.wikipedia.org/wiki/Free_monoid?wprov=sfti1
The simplest example might be {0}^* in which case
0: “” // because we use *
-
Best Way to Learn JAX
Hello! I'm trying to learn JAX over the next couple of weeks. Ideally, I want to be comfortable with using it for projects after about 3 weeks to a month, although I understand that may not be realistic. I currently have experience with PyTorch and TensorFlow. How should I go about learning JAX? Is there a specific YouTube tutorial or online course I should use, or should I just use the tutorial on https://jax.readthedocs.io/? Any information, advice, or experience you can share would be much appreciated!
- Codon: Python Compiler
What are some alternatives?
stable-diffusion - A latent text-to-image diffusion model
Numba - NumPy aware dynamic Python compiler using LLVM
nlp-resume-parser - NLP-powered, GPT-3 enabled Resume Parser from PDF to JSON.
functorch - functorch is JAX-like composable function transforms for PyTorch.
GLM-130B - GLM-130B: An Open Bilingual Pre-Trained Model (ICLR 2023)
julia - The Julia Programming Language
gpt-2 - Code for the paper "Language Models are Unsupervised Multitask Learners"
Pytorch - Tensors and Dynamic neural networks in Python with strong GPU acceleration
manim - Animation engine for explanatory math videos
Cython - The most widely used Python to C compiler
cupscale - Image Upscaling GUI based on ESRGAN
jax-windows-builder - A community supported Windows build for jax.