torchtyping
dex-lang
torchtyping | dex-lang | |
---|---|---|
7 | 25 | |
1,337 | 1,538 | |
- | 0.5% | |
3.2 | 8.8 | |
11 months ago | 3 days ago | |
Python | Haskell | |
Apache License 2.0 | BSD 3-clause "New" or "Revised" License |
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.
torchtyping
-
[D] Have their been any attempts to create a programming language specifically for machine learning?
Not really an answer to your question, but there are Python packages that try to solve the problem of tensor shapes that you mentioned, e.g. https://github.com/patrick-kidger/torchtyping or https://github.com/deepmind/tensor_annotations
-
What's New in Python 3.11?
I disagree. I've had a serious attempt at array typing using variadic generics and I'm not impressed. Python's type system has numerous issues... and now they just apply to any "ArrayWithNDimensions" type as well as any "ArrayWith2Dimenensions" type.
Variadic protocols don't exist; many operations like stacking are inexpressible; the synatx is awful and verbose; etc. etc.
I've written more about this here as part of my TorchTyping project: [0]
[0] https://github.com/patrick-kidger/torchtyping/issues/37#issu...
-
Can anyone point out the mistakes in my input layer or dimension?
also https://github.com/patrick-kidger/torchtyping
-
[D] Anyone using named tensors or a tensor annotation lib productively?
FWIW I'm the author of torchtyping so happy to answer any questions about that. :) I think people are using it!
-
[D] Ideal deep learning library
The one thing I really *really* wish got more attention was named tensors and the tensor type system. Tensor misalignment errors are a constant source of silently-failing bugs. While 3rd party libraries have attempted to fill this gap, it really needs better native support. In particular it seems like bad form to me for programmers to have to remember the specific alignment and broadcasting rules, and then have to apply them to an often poorly documented order of tensor indices. I'd really like to see something like tsalib's warp operator made part of the main library and generalized to arbitrary function application, like a named-tensor version of fold. But preferably using notation closer to that of torchtyping.
-
[P] torchtyping -- documentation + runtime type checking of tensor shapes (and dtypes, ...)
Yes it does work with numerical literals! It support using integers to specify an absolute size, strings to specify names for dimensions that should all be consistently sized (and optionally also checks named tensors), "..." to indicate batch dimensions, and so on. See the full list here.
dex-lang
-
Thinking in an Array Language
A really nice approach to this I've seen recently is Google's research on [Dex](https://github.com/google-research/dex-lang).
- Function Composition in Programming Languages – Conor Hoekstra – CppNorth 2023 [video]
- Dex Lang: Research language for array processing in the Haskell/ML family
-
[D] Have their been any attempts to create a programming language specifically for machine learning?
Dex
-
[D] PyTorch 2.0 Announcement
Have you tried Dex? https://github.com/google-research/dex-lang It is in a relatively early stage, but it is exploring some interesting parts of the design space.
- Mangle, a programming language for deductive database programming
-
Looking for languages that combine algebraic effects with parallel execution
I think [Dex](https://github.com/google-research/dex-lang) might be along the lines of what you're looking for, although its focus is on SIMD GPU-style parallelism rather than thread-level parallelism.
-
“Why I still recommend Julia”
Dex proves indexing correctness without a full dependent type system, including loops.
See: https://github.com/google-research/dex-lang/pull/969
-
Haskell for Artificial Intelligence?
In case you want to see one research direction that's combining practical machine learning and functional programming, one of the authors of JAX (and the main author of its predecessor, Autograd) is writing Dex (https://github.com/google-research/dex-lang), a functional language for array processing. The compiler itself is written in Haskell. JAX is one of the most popular libraries for doing a lot of machine learning these days, along with Tensorflow and PyTorch. You might also want to see the bug in the JAX repo about adding Haskell support, for some context: https://github.com/google/jax/issues/185
What are some alternatives?
jaxtyping - Type annotations and runtime checking for shape and dtype of JAX/NumPy/PyTorch/etc. arrays. https://docs.kidger.site/jaxtyping/
jax - Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more
equinox - Elegant easy-to-use neural networks + scientific computing in JAX. https://docs.kidger.site/equinox/
futhark - :boom::computer::boom: A data-parallel functional programming language
tsalib - Tensor Shape Annotation Library (numpy, tensorflow, pytorch, ...)
julia - The Julia Programming Language
mypy - Optional static typing for Python
Pytorch - Tensors and Dynamic neural networks in Python with strong GPU acceleration
functorch - functorch is JAX-like composable function transforms for PyTorch.
hasktorch - Tensors and neural networks in Haskell
tensor_annotations - Annotating tensor shapes using Python types
CIPs