FluxML/Flux.jl

Pull Requests Commits

Define `expm1` for `Nil`, so `outputsize` works with `elu`, `celu`, `selu`

CarloLucibello committed 15 days ago

dc5d5657

Add `Attention` layer on NNlib's fused attention, deprecate `MultiHeadAttention` (#2739)

CarloLucibello committed 16 days ago

Verified c2dcfc01

Update GPUArrays requirement from 11.2 to 11.2, 12.1 (#2741)

dependabot[bot] committed 16 days ago

Verified 2eb39211

Reactant on CUDA: separate Buildkite job, backend check, wider tests (#2738)

CarloLucibello committed 20 days ago

Verified 394703b1

Fix `gpu` falling back to the CPU on Julia 1.13, and the Reactant GPU tests (#2737)

CarloLucibello committed 21 days ago

Verified efcab379

Bump actions/checkout from 4 to 7 (#2731)

dependabot[bot] committed 22 days ago

Verified da1f0d16

Bump julia-actions/setup-julia from 2 to 3 (#2732)

dependabot[bot] committed 22 days ago

Verified 81374fb2

Bump JuliaRegistries/TagBot from 1.25.11 to 1.25.13 (#2735)

dependabot[bot] committed 22 days ago

Verified 3a9f84f8

Fix three DDP correctness issues: PMI guardrails, data padding, unuse… (#2694)

edopica committed 44 days ago

Verified b617b334

[tests] add enzyme gradient tests (#2721)

hardikxk committed 44 days ago

Verified de0fb135

Update SparseArrays requirement to 1.10 (#2727)

dependabot[bot] committed 44 days ago

Verified c0d08771

Update LinearAlgebra requirement to 1.10 (#2728)

dependabot[bot] committed 44 days ago

Verified e4b50ab1

Update Random requirement to 1.10 (#2729)

dependabot[bot] committed 44 days ago

Verified 967d47a4

Update Project.toml

CarloLucibello committed 48 days ago

Verified 64542243

return send/receive buffers in __reduce! and __reduceall! for non-GPU aware MPI implementations (#2725)

Alexander-Barth committed 48 days ago

Verified 4dbfac8c

add missing pad_mode argument (#2724)

Alexander-Barth committed 48 days ago

Verified 574a3bf7

Add `:named` show option to `@layer` macro (#2719)

hardikxk committed 77 days ago

Verified 404ff37d

Bump JuliaRegistries/TagBot from 1.25.10 to 1.25.11 (#2720)

dependabot[bot] committed 79 days ago

Verified 95860ee1

Non-zero padding modes, `:same`, and even-filter same-padding fix (#2717)

CarloLucibello committed 82 days ago

Verified 9121ba9d

Fix Dense with a zero-sized batch dimension (#2407) (#2718)

CarloLucibello committed 82 days ago

Verified bc7324e7

fix stack overflow when passing empty struct (#2670)

hardikxk committed 83 days ago

Verified 18f475ef

BatchNorm: error on a single value per channel in training (fixes #1992) (#2716)

CarloLucibello committed 85 days ago

Verified 3f2f828b

Bump version from 0.16.10 to 0.16.11

CarloLucibello committed 87 days ago

Verified 443ed1bc

Resolve AD backend once for `train!`/`trainstep!`; honour explicit `adtype` on Reactant (#2714)

CarloLucibello committed 87 days ago

Verified 6af22a1e

Reactant docs guide + `trainstep!` in the ResNet example & benchmark (#2713)

CarloLucibello committed 87 days ago

Verified c7b0217c

Reactant trainstep!: single-forward aux, host-side outputs, cache warning (#2712)

CarloLucibello committed 87 days ago

Verified cea47b23

train_step! -> trainstep!

CarloLucibello committed 88 days ago

067e83d5

CarloLucibello committed 88 days ago

2dda675e

Support auxiliary loss outputs in train_step! and train! (#2711)

CarloLucibello committed 88 days ago

Verified c234ff91

add padding_idx to Embedding (#2708)

CarloLucibello committed 88 days ago

Verified 0170e9e2

Older