Commits · 392ddc1867a09359549ce1cc87e803037f3768b8 · Jon Lachmann / torch · GitLab

Feb 10, 2022
- Merge pull request #781 from mlverse/refactor/LanternFunction · 392ddc18
  Daniel Falbel authored Feb 10, 2022
  392ddc18
- We now correctly delete the Rcpp function called by lantern when the LanternNode is deleted. · 5784d485
  Daniel Falbel authored Feb 10, 2022
  
  5784d485
- Don't need to take fun into the closure. · beee8ffa
  Daniel Falbel authored Feb 10, 2022
  
  beee8ffa
- Move a few calls to inside apply. · c9efdd4a
  Daniel Falbel authored Feb 10, 2022
  
  c9efdd4a
- Comment out windows GPU runners as they are too unstable. · 40306a51
  Daniel Falbel authored Feb 10, 2022
  
  40306a51
Feb 09, 2022
- Merge pull request #774 from mlverse/cudnn-version · 93b250ae
  Daniel Falbel authored Feb 09, 2022
  93b250ae
- Merge branch 'main' into cudnn-version · 951bae40
  Daniel Falbel authored Feb 09, 2022
  
  951bae40
- Add NEWS bullets. · aeb51900
  Daniel Falbel authored Feb 09, 2022
  
  aeb51900
- add URL so the search index is correctly built. · 38edab2e
  Daniel Falbel authored Feb 09, 2022
  
  38edab2e
- Merge pull request #778 from mlverse/bs5 · 8886b690
  Daniel Falbel authored Feb 09, 2022
  8886b690
- No need for `cat` · 69294734
  Daniel Falbel authored Feb 09, 2022
  
  69294734
- fix the print function · 6358e00a
  Daniel Falbel authored Feb 09, 2022
  
  6358e00a
- Update pkgdown to bootstrap 5 · b572b0be
  Daniel Falbel authored Feb 09, 2022
  
  b572b0be
- Styling and formatting R and C/C++ code · aaff492e
  GitHub Actions authored Feb 09, 2022
  
  aaff492e
- Merge branch 'cudnn-version' of https://github.com/rstudio/torch into cudnn-version · c7c879fa
  Daniel Falbel authored Feb 09, 2022
  
  c7c879fa
- Only include CUDA Allocator if on NVCC. · 152ac5be
  Daniel Falbel authored Feb 09, 2022
  
  152ac5be
- Add unit test · 9342933d
  Daniel Falbel authored Feb 09, 2022
  
  9342933d
- remove name and force init of device. · e00879df
  Daniel Falbel authored Feb 09, 2022
  
  e00879df
Feb 04, 2022
- Re-document. · 5777b2d5
  Daniel Falbel authored Feb 04, 2022
  
  5777b2d5
- Merge branch 'cudnn-version' of https://github.com/rstudio/torch into cudnn-version · f83b092b
  Daniel Falbel authored Feb 04, 2022
  
  f83b092b
- Add memory diagnostics helper. · 846b7f6f
  Daniel Falbel authored Feb 04, 2022
  
  846b7f6f
- Re-document. · f085e28b
  Daniel Falbel authored Feb 04, 2022
  
  f085e28b
- Add support for querying cudnn version. · 0610350b
  Daniel Falbel authored Feb 04, 2022
  
  0610350b
- Update NEWS.md · d1fa3e73
  Daniel Falbel authored Feb 04, 2022
  
  d1fa3e73
- Merge pull request #773 from mlverse/flatten · 2d3a7f3d
  Daniel Falbel authored Feb 04, 2022
  2d3a7f3d
- Update NEWS.md · 30ffd7ab
  Daniel Falbel authored Feb 04, 2022
  
  30ffd7ab
- Merge pull request #772 from mlverse/max-min-other · 11db6951
  Daniel Falbel authored Feb 04, 2022
  11db6951
- Update NEWS.md · 7fca15d2
  Daniel Falbel authored Feb 04, 2022
  
  7fca15d2
- Merge pull request #771 from mlverse/autograd-grad-deadlock · c033e60b
  Daniel Falbel authored Feb 04, 2022
  c033e60b
- Add `nn_flatten` and `nn_unflatten`. · 0768b3f0
  Daniel Falbel authored Feb 04, 2022
  
  0768b3f0
- Correctly compute min when other has lenght 2. · 1beafdfe
  Daniel Falbel authored Feb 04, 2022
  
  1beafdfe
- Merge pull request #770 from mlverse/truncated-defaults · 0e8139da
  Daniel Falbel authored Feb 04, 2022
  0e8139da
- Add unit test to make sure we don't segfault. · cbbfa61c
  Daniel Falbel authored Feb 04, 2022
  
  cbbfa61c
- Launch `autograd_grad` in a separate thread so the main thread is free to run R callbacks. · 6faa3d3d
  Daniel Falbel authored Feb 04, 2022
  
  6faa3d3d
- Add test case to expose the problem. · e380d17a
  Daniel Falbel authored Feb 04, 2022
  
  e380d17a
- Re-document. · c70b0240
  Daniel Falbel authored Feb 04, 2022
  
  c70b0240
- Regenerate with untruncated defaults. · 88c69bf4
  Daniel Falbel authored Feb 04, 2022
  
  88c69bf4
- Merge pull request #769 from mlverse/nnf_cosine_similarity-default · 67b7dfaf
  Daniel Falbel authored Feb 04, 2022
  67b7dfaf
- Changed default `dim` from `1` to `2` in `nnf_cosine_similarity`. Re-document. · 0d3b0ff1
  Daniel Falbel authored Feb 03, 2022
  
  0d3b0ff1
- Merge pull request #767 from mlverse/purl · c5373ce8
  Daniel Falbel authored Feb 03, 2022
  c5373ce8