Skip to content

Commit 92f3b10

Browse files
committed
shape inference sounds better
1 parent bf49b66 commit 92f3b10

File tree

2 files changed

+7
-7
lines changed

2 files changed

+7
-7
lines changed

docs/make.jl

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -35,7 +35,7 @@ makedocs(
3535
"GPU Support" => "gpu.md",
3636
"Model Tools" => [
3737
"Saving & Loading" => "saving.md",
38-
"Size Propagation" => "outputsize.md",
38+
"Shape Inference" => "outputsize.md",
3939
"Weight Initialisation" => "utilities.md",
4040
],
4141
"Performance Tips" => "performance.md",

docs/src/outputsize.md

Lines changed: 6 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -1,10 +1,10 @@
1-
# Size Propagation
1+
# Shape Inference
22

3-
Flux provides some utility functions to help you generate models in an automated fashion.
3+
To help you generate models in an automated fashion, [`Flux.outputsize`](@ref) lets you
4+
calculate the size returned produced by layers for a given size input.
5+
This is especially useful for layers like [`Conv`](@ref).
46

5-
[`Flux.outputsize`](@ref) enables you to calculate the output sizes of layers like [`Conv`](@ref)
6-
when applied to input samples of a given size. This is achieved by passing a "dummy" array into
7-
the model that preserves size information without running any computation.
7+
It works by passing a "dummy" array into the model that preserves size information without running any computation.
88
`outputsize(f, inputsize)` works for all layers (including custom layers) out of the box.
99
By default, `inputsize` expects the batch dimension,
1010
but you can exclude the batch size with `outputsize(f, inputsize; padbatch=true)` (assuming it to be one).
@@ -44,4 +44,4 @@ end
4444

4545
```@docs
4646
Flux.outputsize
47-
```
47+
```

0 commit comments

Comments
 (0)