1166: Fix crossentropy when some probabilities are zero r=dhairyagandhi96 a=cossio
Use a function `xlogy(x,y) = x * log(y)` that has the correct limit at `x=0`.
Before this PR:
```julia
julia> Flux.crossentropy([0.1,0.0,0.9], [0.1,0.0,0.9])
NaN
```
After this PR:
```julia
julia> Flux.crossentropy([0.1,0.0,0.9], [0.1,0.0,0.9])
0.3250829733914482
```
Co-authored-by: cossio <j.cossio.diaz@gmail.com>
1160: Build docs on Julia 1.3 r=dhairyagandhi96 a=dhairyagandhi96
This causes red CI otherwise
Co-authored-by: Dhairya Gandhi <dhairya@juliacopmuting.com>
1156: Add correct overload for apply! in docs r=dhairyagandhi96 a=dhairyagandhi96
Maybe we should considering adding a `const` name that is better than `apply!` (or rename `apply!`) and export it, so folks can just overload `descriptive_apply_my_optimiser_rule!` rather than have to go to the sub-project `Flux.Optimise`?
Co-authored-by: Dhairya Gandhi <dhairya@juliacopmuting.com>
901: Add option for "Same" padding to conv and pooling layers r=dhairyagandhi96 a=DrChainsaw
Fixes#813
This adds the possibility to set "pad=SamePad()" to automatically calculate the amount of padding to apply so that outputsize==inputsize (assuming stide == 1).
Comments on API more than welcome. I considered the following options:
* Call the type just Same and export it, but I was afraid to cause name collisions due to a too generic name
* Call the type Same and not export it
* Dispatch on type instead of instance (so that one can type pad=Same instead of pad=Same())
* Supply a method instead of a type, giving a similar API as above.
Happy to change to any of the above or to anything else.
I don't think that same padding is common for pooling layers, but I added it just for the sake of consistency. It is a separate commit so it can easily be removed if not wanted.
Co-authored-by: DrChainsaw <Christian.kyril.skarby@gmail.com>
1131: Update glorot_normal doc r=dhairyagandhi96 a=AdarshKumar712
Just a minute correction in glorot_normal function doc.
Co-authored-by: Adarsh Kumar <45385384+AdarshKumar712@users.noreply.github.com>
853: Improve docs r=CarloLucibello a=janEbert
If you disagree with any of the changes, please tell me what to reverse or fix.
I am unsure about the docstrings I added to `src/utils.jl` for `unsqueeze` and
the `[un]stack` functions so please give those a more detailed look.
Update Documenter.jl version for new features, fix deprecation warnings in
`docs/make.jl` and import Flux for all doctests.
Add missing docstrings to `src/utils.jl`, `src/layers/stateless.jl` and `src/data/`; add
these and other missing functions to Markdown docs.
Improve docstrings by...
- fixing typos,
- removing trailing or double whitespaces,
- using `jldoctest` blocks where applicable,
- fixing, updating or correctly setting up existing doctests,
- improving consistency (for example, always use "# Examples" instead
of other variants),
- removing empty lines between docstrings and functions,
- instead of mentioning keywords, put them into the docstring,
- adding some missing but useful keywords,
- adding references (`@ref`),
- using LaTeX math where applicable, and
- linking papers.
Debatable stuff that is untouched:
- BE/AE s/z irregularities (e.g. "normalise" versus "normalize") since
most papers use the AE version while the Flux source code was
written with BE spelling.
- Names of normalization functions are capitalized
("Batch Normalization" instead of "batch normalization").
- Default values in argument lists have spaces around the equals sign (`arg = x` instead of `arg=x`).
Co-authored-by: janEbert <janpublicebert@posteo.net>
1110: fix tests and new version r=CarloLucibello a=CarloLucibello
Add to set the Boston Housing dataset tests as broken due to as SSL certificate expiration problem wich is not our fault
Co-authored-by: Carlo Lucibello <carlo.lucibello@gmail.com>
