Commits · 0.3.2 · pycodegen / pystencils_autodiff

17 Dec, 2019 9 commits
- Fixup · 8abd3aab
  Stephan Seitz authored Dec 17, 2019
  
  8abd3aab
- Remove unnecessary xfails · 19f36b03
  Stephan Seitz authored Dec 17, 2019
  
  19f36b03
- Accelerate gradient check test by using smaller arrays · 64fbd5eb
  Stephan Seitz authored Dec 17, 2019
  
  64fbd5eb
- Fixup · c6e5fda1
  Stephan Seitz authored Dec 17, 2019
  
  c6e5fda1
- Use pystencils compiler as torch host compiler · aa5f89f0
  Stephan Seitz authored Dec 17, 2019
  
  aa5f89f0
- Do not link against cudart as default · 22007af0
  Stephan Seitz authored Dec 17, 2019
  
  22007af0
- Fix (?) autodiff.get/setstate · 7a292ebb
  Stephan Seitz authored Dec 17, 2019
  
  7a292ebb
- Remove AutodiffOp.backward_fields_map · a91a2f9f
  Stephan Seitz authored Dec 17, 2019
  
  a91a2f9f
- Add additional assertions · ddba65e5
  Stephan Seitz authored Dec 17, 2019
  
  ddba65e5
16 Dec, 2019 19 commits
- Copy contents of index.rst to README.rst · 906de4b0
  Stephan Seitz authored Dec 16, 2019
  
  906de4b0
- 💄 Lint · 253c5a9c
  Stephan Seitz authored Dec 16, 2019
  
  253c5a9c
- Also hash input fields and backward_input_fields in Tf hash name · 52564579
  Stephan Seitz authored Dec 16, 2019
  
  52564579
- Don't use CUDA for doctest · cd438f95
  Stephan Seitz authored Dec 16, 2019
  
  cd438f95
- Fix README.rst/index.rst · 75f1a1c0
  Stephan Seitz authored Dec 16, 2019
  
  75f1a1c0
- Try skipping torch GPU test on CI · 1666f0b9
  Stephan Seitz authored Dec 16, 2019
  
  1666f0b9
- Use tensorflow CPU · a10c2d24
  Stephan Seitz authored Dec 16, 2019
  
  a10c2d24
- Allow use_cuda to be truthy (instead of only True/False) · 25ebe582
  Stephan Seitz authored Dec 16, 2019
  
  25ebe582
- Fixup: .gitlab-ci · f3772f74
  Stephan Seitz authored Dec 16, 2019
  
  f3772f74
- Mark test_datahandling xfail · 59808045
  Stephan Seitz authored Dec 16, 2019
  
  59808045
- Try install tensorflow 2.0 · bd7777c8
  Stephan Seitz authored Dec 16, 2019
  
  bd7777c8
- Add additional assertion: with_cuda needs to be bool · 8b80a78e
  Stephan Seitz authored Dec 16, 2019
  
  8b80a78e
- Mark pytorch CUDA xfail for CI · c2cd47b7
  Stephan Seitz authored Dec 16, 2019
  
  c2cd47b7
- Make test_tfmad_gradient_check_torch_native also pass with CUDA · a9797544
  Stephan Seitz authored Dec 16, 2019
```
with_cuda needed to be True not 'with_cuda'
```
  a9797544
- Add TODO · 0e367089
  Stephan Seitz authored Dec 16, 2019
  
  0e367089
- Also rename KernelFunctionCall -> FunctionCall in test · ae1ea301
  Stephan Seitz authored Dec 16, 2019
  
  ae1ea301
- Change requirements for pystencils_autodiff · 2dfe28b0
  Stephan Seitz authored Dec 16, 2019
  
  2dfe28b0
- Add justfile · c6a3bc29
  Stephan Seitz authored Dec 16, 2019
  
  c6a3bc29
- Sort fields by name to avoid confusion when returning gradients by order · f0d770c2
  Stephan Seitz authored Dec 16, 2019
  
  f0d770c2
02 Dec, 2019 2 commits
- Fixup: allow time_constant_fields == None · 20f2ee40
  Stephan Seitz authored Dec 02, 2019
  
  20f2ee40
- Make all fields == not time_constant_fields default · 8b7bdb96
  Stephan Seitz authored Dec 02, 2019
  
  8b7bdb96
29 Nov, 2019 9 commits
- Change torch native for new interface · 1e141dfb
  Stephan Seitz authored Nov 29, 2019
  
  1e141dfb
- Remove openmp for now · 75dd38f2
  Stephan Seitz authored Nov 29, 2019
  
  75dd38f2
- Put openmp in spot where its allowed · 60c9f471
  Stephan Seitz authored Nov 29, 2019
  
  60c9f471
- Oops -fopenmp should go into hash · 7e7f127f
  Stephan Seitz authored Nov 29, 2019
  
  7e7f127f
- Fix unused parameter · 797bf4bd
  Stephan Seitz authored Nov 29, 2019
  
  797bf4bd
- Fix gradient calculation for Tensorflow · 3b21e27a
  Stephan Seitz authored Nov 29, 2019
  
  3b21e27a
- Make openmp for tf cpu optional · 9ed3556c
  Stephan Seitz authored Nov 29, 2019
  
  9ed3556c
- Add openmp flag to tensorflow jit · 8f91bec3
  Stephan Seitz authored Nov 29, 2019
  
  8f91bec3
- Weirdly cache result · 603ad7ef
  Stephan Seitz authored Nov 29, 2019
  
  603ad7ef
28 Nov, 2019 1 commit
- Fix loading of tensorflow modules · 0ef3d7a9
  Stephan Seitz authored Nov 28, 2019
  
  0ef3d7a9