v0.2.3

Reactant v0.2.3

Diff since v0.2.2

Merged pull requests:

• Export GetDefaultTargetTriple symbol (#160) (@mofeing)
• feat: implement a separate TracedRNumber (#161) (@avik-pal)
• Generalize Base._cat to non-Val, typed Base._cat_t and implement typed_hcat, typed_vcat, typed_hvcat, typed_hvncat (#163) (@mofeing)
• Bypass identical construction on TracedRArray (#165) (@mofeing)
• test: separate out test groups (#167) (@avik-pal)

Closed issues:

• Compiling leads to KeyError (#107)
• Error with vcat (#139)
• Inconsistent semantics of 0-dim RArrays compared to Julia semantics (#141)
• Missing hcat/vcat/hvcat dispatches (#148)

v0.2.2

Reactant v0.2.2

Diff since v0.2.1

Merged pull requests:

• feat: compiling vision transformers (#130) (@avik-pal)
• feat: robust handling of wrapped arrays of reactant arrays (#133) (@avik-pal)
• Replace getfield for traced_getfield in set! (#134) (@mofeing)
• perf: benchmarking CI (#136) (@avik-pal)
• Add fallback gpu setindex (#137) (@wsmoses)
• Bump dependencies (#143) (@wsmoses)
• Build (#145) (@wsmoses)
• feat: generalize indexing to all wrappers (#146) (@avik-pal)
• Arm64 (#147) (@wsmoses)
• Format code of branch "main" (#153) (@github-actions[bot])
• fix: broadcasted type casting (#156) (@avik-pal)
• Fix vararg (#158) (@wsmoses)

Closed issues:

• Handling exponents (#27)
• Incosistent semantics of element-wise application of array functions compared to Julia (#51)
• NNlib activations cannot be compiled without manual dispatches (#54)
• Linear algebra on TracedRArray (#131)
• scalar broadcasting not functional (#135)
• libdevice not found (#149)
• Vararg compilation broken (#151)
• Can I compile in-place functions? (#152)
• Optimisers.jl fails to compile (#154)
• Type-casting doesn't compile (#155)
• Unbreak Reactant benchmark CI (#157)

v0.2.1

Reactant v0.2.1

Diff since v0.2.0

Merged pull requests:

• Format code of branch "main" (#101) (@github-actions[bot])
• Compile NNlib.maxpool and NNlib.meanpool (#102) (@Pangoraw)
• feat: handle cat/hcat/vcat (#103) (@avik-pal)
• feat: implementing view/getindex/setindex (#104) (@avik-pal)
• perf: test compile times for neural networks -- VGG16 (#105) (@avik-pal)
• fix: local JLL build script (#108) (@avik-pal)
• NNlib: forward feature group counts to convolution (#109) (@Pangoraw)
• tests: more comprehensive NN testing (#110) (@avik-pal)
• ci(github-actions): update the formatter action (#111) (@avik-pal)
• Format code of branch "main" (#112) (@github-actions[bot])
• Format code of branch "main" (#113) (@github-actions[bot])
• feat: allow building for CPU target on CUDA-enabled machines (#114) (@avik-pal)
• NNlib: handle kernel flip with reverse instead of window_reversal (#115) (@Pangoraw)
• Format code of branch "main" (#116) (@github-actions[bot])
• Configure CompatHelper workflow (#118) (@mofeing)
• chore: testing Enzyme 0.13 (#119) (@avik-pal)
• Fix non-used argument (#123) (@wsmoses)
• Fix typo in mlirIsNull call (#124) (@mofeing)
• Format code of branch "main" (#125) (@github-actions[bot])
• Refactor compile and test buffer donation (#126) (@mofeing)
• Replace IdDict for OrderedIdDict (#128) (@mofeing)
• CompatHelper: add new compat entry for OrderedCollections at version 1, (keep existing compat) (#129) (@github-actions[bot])

Closed issues:

• Set up GPU CI (#41)
• Print "donated" attribute on @code_hlo (#97)
• Grouped Convolution generates incorrect code (#106)
• Compiling leads to KeyError (#107)
• Update to Enzyme v0.13 (#117)
• gelu needs to be fixed (#121)
• Max Pooling doesn't give correct answers for certain padding (#122)

v0.2.0

Reactant v0.2.0

Diff since v0.1.9

Merged pull requests:

• Move shape information of RArray to field (#58) (@mofeing)
• Refactor traced_type to allow customization for external types (#59) (@mofeing)
• Despecialize some methods to lower excessive codegen (#61) (@mofeing)
• Add benchmarking infrastructure (#65) (@mofeing)
• Tpu (#66) (@wsmoses)
• Refactor compilation (#68) (@mofeing)
• Format code of branch "main" (#69) (@github-actions[bot])
• Format code of branch "main" (#70) (@github-actions[bot])
• ci(buildkite): fix scripts (#71) (@avik-pal)
• Fix permutedims on TracedRArray (#72) (@mofeing)
• Fix broadcasting on ConcreteRArray (#73) (@mofeing)
• Small fixes on abstract interpreter (#74) (@mofeing)
• Try hermetic cuda (#75) (@wsmoses)
• Implement create_result method for AbstractDict (#76) (@mofeing)
• Implement only for TracedRArray (#77) (@mofeing)
• Reuse code cache in abstract interpretation to minimize compile time (#78) (@mofeing)
• TPU index setup (#81) (@wsmoses)
• Format code of branch "main" (#82) (@github-actions[bot])
• Export LoadPjrtPlugin, InitializePjrtPlugin, GetCApiClient functions (#84) (@mofeing)
• Compile NNlib convolution (#87) (@Pangoraw)
• Format code of branch "main" (#89) (@github-actions[bot])
• Cleanup api (#90) (@wsmoses)
• Imrpove FFI (#91) (@wsmoses)
• Format code of branch "main" (#92) (@github-actions[bot])
• Fix libtpu auto download (#93) (@wsmoses)
• CUDA build local (#94) (@wsmoses)
• Format code of branch "main" (#95) (@github-actions[bot])
• Add back support for buffer donation (#96) (@mofeing)

Closed issues:

• [Eager Mode] Broadcasting ConcreteRArray (#18)
• Roadmap to v0.2 (#19)
• Fix precompilation in macOS (#33)
• stablehlo.compare derivative isn't implemented (#57)
• Avoid mlir context on copy (#83)

v0.1.9

Reactant v0.1.9

Diff since v0.1.8

Merged pull requests:

• Run pass pipeline from Julia (#49) (@Pangoraw)
• Fix multisymbol issues (#53) (@wsmoses)
• feat: more coverage for common NN operations (#55) (@avik-pal)
• Remove PackageExtensionsCompat (#62) (@mofeing)
• Fix aarch build (#64) (@wsmoses)

v0.1.8

Reactant v0.1.8

Diff since v0.1.7

Merged pull requests:

• Replace Cassette for an overlayed MethodTable (#40) (@mofeing)

Closed issues:

• Display releases on GitHub? (#46)

v0.1.7

Reactant v0.1.7

Merged pull requests:

• Support struct types in create_result (#6) (@mofeing)
• Patches for Lux integration (#7) (@avik-pal)
• Fix warning box in README (#10) (@mofeing)
• Refactor create_result for Tuple types to include NamedTuples (#11) (@mofeing)
• Format code of branch "main" (#12) (@github-actions[bot])
• Format code of branch "main" (#13) (@github-actions[bot])
• Format code of branch "main" (#15) (@github-actions[bot])
• Lux (#20) (@wsmoses)
• Format code of branch "main" (#21) (@github-actions[bot])
• Fix worldage issues (#22) (@wsmoses)
• Format code of branch "main" (#23) (@github-actions[bot])
• Add elem apply of convert (#25) (@wsmoses)
• Format code of branch "main" (#26) (@github-actions[bot])
• Fix create_result on Vector{Symbol} (#28) (@mofeing)
• Format code of branch "main" (#31) (@github-actions[bot])
• Missing Meta.quot (#34) (@avik-pal)
• Generalize broadcast (#35) (@wsmoses)
• Support closures (#36) (@mofeing)
• fix elem_apply usage in mapreduce (#38) (@Pangoraw)
• add @code_hlo macro to get mlir code (#39) (@Pangoraw)
• Bump jll and rewrite testing (#42) (@wsmoses)

Closed issues:

• Crash on precompiling ReactantNNlibExt (#1)
• Crash on Reactant.compile (#2)
• Define similar for Reactant.ConcreteRArray (#4)
• Define create_result for NamedTuple (#8)
• Pipeline Error Reactant + Enzyme (#14)
• Macro to pretty print the stable hlo code (#17)
• xla RuntimeError on differentiating simple neural network (#24)
• Track closures' captured variables (#32)