update readme

This commit is contained in:
Mike J Innes 2017-09-07 21:15:52 -04:00
parent f55b8cd20e
commit 45141059eb
1 changed files with 1 additions and 36 deletions

View File

@ -4,39 +4,4 @@
Flux is a refreshing approach to machine learning. It provides lightweight abstractions on top of Julia's native GPU and AD support, while remaining fully hackable (right down to the [GPU kernels](https://github.com/FluxML/CuArrays.jl)).
Define a simple model using any Julia code:
```julia
using Flux.Tracker
x, y = rand(10), rand(5) # Dummy input / output
# `track` defines parameters that we can train
W, b = track(randn(5,10)), track(randn(5))
# Transform `x` and calculate the mean squared error
loss = Flux.mse(W*x .+ b, y)
# Calculate and store gradients of `track`ed parameters
back!(loss)
Tracker.grad(W) # Get the gradient of `W` wrt the loss
```
Define a larger model using high-level abstractions:
```julia
using Flux
m = Chain(
Dense(10, 32, relu),
Dense(32, 10), softmax)
m(rand(10))
```
Mix and match the two:
```julia
using Flux.Tracker
x, y = rand(10), rand(5)
d = Dense(10, 5)
loss = Flux.mse(d(x), y)
```
See the [documentation](http://fluxml.github.io/Flux.jl/stable/) or the [model zoo](https://github.com/FluxML/model-zoo/) for more examples.
See the [documentation](http://fluxml.github.io/Flux.jl/stable/) or the [model zoo](https://github.com/FluxML/model-zoo/) for examples.