v0.1.3

Torch v0.1.3

Diff since v0.1.2

Merged pull requests:

• CompatHelper: bump compat for "FillArrays" to "0.9" (#34) (@github-actions[bot])
• Added nnlib layers and test cases, added upsampling layers (#38) (@bjosv79)
• CompatHelper: bump compat for "FillArrays" to "0.10" (#41) (@github-actions[bot])
• CompatHelper: bump compat for "FillArrays" to "0.11" (#44) (@github-actions[bot])
• CompatHelper: bump compat for "FillArrays" to "0.12" (#48) (@github-actions[bot])
• Add buildkite pipeline (#51) (@DhairyaLGandhi)
• CompatHelper: bump compat for FillArrays to 0.13, (keep existing compat) (#53) (@github-actions[bot])
• Limit compat of Torch_jll to v1.4 (#55) (@stemann)
• Added C wrapper generator (#56) (@stemann)
• CompatHelper: bump compat for FillArrays to 1, (keep existing compat) (#57) (@github-actions[bot])
• Updated Julia wrapper generator (#59) (@stemann)
• Fixed compat to enable tests on CI (#60) (@stemann)
• Updated C wrapper wrt. Torch v1.10 (#61) (@stemann)
• CompatHelper: add new compat entry for Statistics at version 1, (keep existing compat) (#63) (@github-actions[bot])
• Updated for Torch v1.10.2, and NNlib >= v0.8 (#64) (@stemann)
• Updated authors, and copyright in LICENSE (#65) (@stemann)

Closed issues:

• Add tests and relevant integrations (#12)
• Required version of glibc? (#17)
• could not load library "libdoeye_caml" on windows (#32)
• Using Torch.jl with Flux 0.11 (#33)
• Clarify if Python is needed in the README (#39)
• Error upon installing (#42)
• Failed installing Torch.jl (#50)
• Update wrapper (#54)

v0.1.2

Torch v0.1.2

Diff since v0.1.1

Closed issues:

• Could not load library libdoeye_caml.so (#27)

Merged pull requests:

• Add instructions of CUDA version (#30) (@findmyway)

v0.1.1

Torch v0.1.1

Diff since v0.1.0

Closed issues:

• Should this be in the FluxML org? (#15)

Merged pull requests:

• Make MaxPool grads work (#13) (@dhairyagandhi96)
• Add Julia Computing to license. (#14) (@ViralBShah)
• Expand compat for YAML, bump version number, install CompatHelper, and update manifests (#18) (@DilumAluthge)
• Report correct rank for matmul and aggregrations (#19) (@ToucheSir)

v0.1.0

Torch v0.1.0

Closed issues:

• Add a decent build setup (#3)
• Different behavior for Tensor{Float32,4} and Array{Float32,4} (#5)
• CUDA OOM error when Benchmarking (#6)
• getindex not defined for Tensor{Float32,4} (#7)

Merged pull requests:

• Move to using status error codes (#4) (@dhairyagandhi96)
• add LICENSE (#8) (@dhairyagandhi96)
• added chunk (#9) (@gartangh)
• Link CUDNN directly while building (#10) (@dhairyagandhi96)
• add Torch artifact (#11) (@dhairyagandhi96)