Comments (4)
Thank you, will do a pull request.
I would think the other way , the more people start using, the more requests you get ( know the demand for using Flux on Mac with GPU ), so more incentive to make it faster, quicker.
from flux.jl.
There's some documentation here: http://fluxml.ai/Flux.jl/stable/gpu/#Selecting-GPU-backend Could always be better though.
Support for anything but CUDA is a bit experimental. For me e.g. this model http://fluxml.ai/Flux.jl/stable/models/quickstart/ does not work on Metal, but I didn't try hard & maybe have wrong versions or something. Do some models work for you?
I believe that many functions in NNlib need to call the Metal equivalents, parallel to how things here https://github.com/FluxML/NNlib.jl/tree/master/ext/NNlibCUDAExt call CUDA functions.
from flux.jl.
Thank you .
For example the quick start example works perfectly with Apple silicon but needs to be documented.
The only thing I had to do is switch to "using Metal" instead of "CUDA"
But also exit and create new session . This allowed re-compilation of Flux with Metal.
Can I recommend you put comments in the code ?
From this :
using Flux, CUDA, Statistics, ProgressMeter
Like this :
using CUDA # switch to 'using Metal' for Apple silicon
using Flux, Statistics, ProgressMeter
from flux.jl.
Can I recommend you put comments in the code ?
Feel free to submit a pull request! However, it takes 95 seconds to train on my Apple Silicon GPU and 7 seconds to train on my Pi5's CPU so it might be worth waiting for better support before recommending it in the documentation.
from flux.jl.
Related Issues (20)
- Why is Flux.destructure type unstable? HOT 3
- bad formatting for PairwiseFusion docstring HOT 1
- Zero-sized arrays cannot be applied to Dense layers. HOT 4
- Adding Simple Recurrent Unit as a recurrent layer
- Collecting PyTorch -> Flux migration notes
- tests are failing due to ComponentArrays HOT 2
- deprecate Flux.params HOT 7
- Significant time spent moving medium-size arrays to GPU, type instability HOT 10
- ConvTranspose errors with symmetric non-constant pad
- SamePad() for even sized filters.
- Dense layers with shared parameters HOT 5
- Implementation of `AdamW` differs from PyTorch HOT 10
- `gpu` should warn if cuDNN is not installed HOT 2
- Cannot take `gradient` of L2 regularization loss HOT 1
- Create a flag to use Enzyme as the AD in training/etc. HOT 13
- test Enzyme gradient for loss functions
- test Enzyme gpu support
- Enzyme fails with MultiHeadAttention layer HOT 13
- Enable github Discussions
- Stacked RNN in Flux.jl?
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from flux.jl.