pytorch/docs/source
Michael Carilli 0f0271e255 [RELAND2] Eager autocasting, out-of-place ops only (with MSVC 2017 fix) (#35102)
Summary:
This is the second reland attempt for https://github.com/pytorch/pytorch/pull/32140.

The first reland attempt https://github.com/pytorch/pytorch/pull/35011 failed due a [small incompatible change](https://github.com/pytorch/pytorch/pull/35011#issuecomment-601754216) in recent master (`skipIfRocm` was removed from `test_data_parallel.py`).

The present PR restores skipIfRocm.

Description from first reland attempt https://github.com/pytorch/pytorch/pull/35011:

> https://github.com/pytorch/pytorch/pull/32140 was approved and merged, but [reverted](d0577e19f0) because it broke builds with versions of Visual Studio older than 15.8 that were not represented in public CI.  The build failures were caused by a [known VS bug](https://developercommunity.visualstudio.com/content/problem/27729/allow-function-with-internal-linkage-as-template-n.html), fixed in versions 15.8 and newer.
>
> The present PR reverts the revert (restoring https://github.com/pytorch/pytorch/pull/32140 's diffs) and adds a workaround to enable compilation with VS < 15.8.  The workaround isn't pretty, but it's guarded by macros such that it's only used when compiling with VS < 15.8.  All other builds compile with the same code/control flow as was merged in https://github.com/pytorch/pytorch/pull/32140.
>
> Original description of https://github.com/pytorch/pytorch/pull/32140:
> > Initial integration of eager autocasting, supporting out-of-place ops only for easier review.
> Relevant issue/RFC: https://github.com/pytorch/pytorch/issues/25081
>
> > In-place ops and ops with user-supplied out=... can certainly be supported as well (my initial WIP https://github.com/pytorch/pytorch/issues/29552 handled many) but require substantially more complex special casing in the autocasting backend and tests. Support for these ops (much of which has already been written) will be broken into later PRs.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/35102

Differential Revision: D20596918

Pulled By: ezyang

fbshipit-source-id: 60caa279bb0ce4a9bb0b28c1d585d42cf1cc7e50
2020-03-24 09:08:04 -07:00
..
_static Improve documentation around builtin functions (#30347) 2019-12-04 13:50:40 -08:00
_templates Generate sphinx docs with secure content. (#18508) 2019-03-27 11:01:48 -07:00
_templates-stable Add layout.html to template for stable docs (#33770) 2020-03-04 03:14:52 -08:00
community Fix broken links in governance.rst 2020-02-04 14:26:09 -08:00
notes [RELAND2] Eager autocasting, out-of-place ops only (with MSVC 2017 fix) (#35102) 2020-03-24 09:08:04 -07:00
org/pytorch [jit] kill script namespace (#34515) 2020-03-11 23:32:48 -07:00
scripts Add torch.nn.GELU for GELU activation (#28944) 2019-11-03 21:55:05 -08:00
__config__.rst Allow a non-OpenMP based build (#19749) 2019-05-06 19:34:48 -07:00
amp.rst [RELAND2] Eager autocasting, out-of-place ops only (with MSVC 2017 fix) (#35102) 2020-03-24 09:08:04 -07:00
autograd.rst functional autograd api (#34066) 2020-03-19 08:24:07 -07:00
bottleneck.rst
checkpoint.rst
conf.py Revert D19320493: Javadoc changes 2020-01-09 14:23:30 -08:00
cpp_extension.rst
cuda_deterministic_backward.rst Typo correction in cuda_deterministic_backward.rst (#25011) 2019-08-22 21:19:39 -07:00
cuda_deterministic.rst
cuda.rst Fix most documentation warnings (#27782) 2019-10-13 10:34:01 -07:00
cudnn_deterministic.rst
cudnn_persistent_rnn.rst
data.rst Fix typo in data.rst (#34624) 2020-03-11 19:40:18 -07:00
distributed.rst Update docs about DP and DDP for CUDA (#35063) 2020-03-20 20:06:37 -07:00
distributions.rst Continuous bernoulli distribution (take 2) (#34619) 2020-03-12 11:53:18 -07:00
dlpack.rst
hub.rst Fix typos, via a Levenshtein-type corrector (#31523) 2020-01-17 16:03:19 -08:00
index.rst Add list of view ops to public doc. (#32560) 2020-02-28 15:05:55 -08:00
jit_builtin_functions.rst Fix builtin function reference (#24056) 2019-08-09 15:58:15 -07:00
jit_language_reference.rst [docs] add experimental warning to TorchScript classes in language reference (#33697) 2020-02-24 14:01:19 -08:00
jit_python_reference.rst Add Python language reference docs (#30686) 2019-12-26 13:21:36 -08:00
jit_unsupported.rst [JIT] add support for torch.norm (#33783) 2020-03-05 14:46:24 -08:00
jit.rst [jit] kill script namespace (#34515) 2020-03-11 23:32:48 -07:00
math-quantizer-equation.png adding quantization.rst file for quantization feature (#27559) 2019-10-09 16:45:09 -07:00
model_zoo.rst add/move a few apis in torch.hub (#18758) 2019-04-10 23:10:39 -07:00
multiprocessing.rst Bag of documentation fixes; fix more sphinx warnings (#27850) 2019-10-15 07:31:14 -07:00
name_inference.rst Fix typos (#30606) 2019-12-02 20:17:42 -08:00
named_tensor.rst Bag of documentation fixes; fix more sphinx warnings (#27850) 2019-10-15 07:31:14 -07:00
nn.functional.rst Breaks up NN module in docs so it loads faster. 2019-06-11 09:38:41 -07:00
nn.init.rst Bag of documentation fixes; fix more sphinx warnings (#27850) 2019-10-15 07:31:14 -07:00
nn.rst Pruning Functionality (#24076) 2019-11-08 19:38:00 -08:00
onnx.rst Adds true_divide function, analogous to Python 's, JAX's, NumPy's (true) division (#34236) 2020-03-09 21:06:33 -07:00
optim.rst Fix capitalization inconsistency in optim.rst 2019-12-04 08:17:03 -08:00
packages.rst Revert D19320493: Javadoc changes 2020-01-09 14:23:30 -08:00
quantization.rst Fixed typos in quantization docs / docstrings (#34182) 2020-03-06 21:53:52 -08:00
random.rst Fix most documentation warnings (#27782) 2019-10-13 10:34:01 -07:00
rpc.rst Update descriptions for transmitting CUDA tensors (#34888) 2020-03-17 17:43:48 -07:00
sparse.rst Bag of documentation fixes; fix more sphinx warnings (#27850) 2019-10-15 07:31:14 -07:00
storage.rst
tensor_attributes.rst Add docs about memory format (#34818) 2020-03-23 15:06:33 -07:00
tensor_view.rst Add list of view ops to public doc. (#32560) 2020-02-28 15:05:55 -08:00
tensorboard.rst Add method add_hparams to API doc (#27344) 2019-10-03 17:07:45 -07:00
tensors.rst Extends true_divide to be a method (#34794) 2020-03-23 23:12:23 -07:00
torch.rst [quant] Add dequantize.tensors (#34348) 2020-03-20 22:51:51 -07:00
type_info.rst