OSSForks/pytorch
1
0
Fork 0
mirror of https://github.com/zebrajr/pytorch.git synced 2025-12-07 12:21:27 +01:00
Code Issues Actions 38 Packages Projects Releases Wiki Activity
24,043 Commits 8,304 Branches 150 Tags 1.5 GiB
4bef344210
Commit Graph

750 Commits

Author SHA1 Message Date
Nicki Skafte
4bef344210 Implementation of mixture distributions (#22742) 
Summary:
Addressing issue https://github.com/pytorch/pytorch/issues/18125
This implements a mixture distributions, where all components are from the same distribution family. Right now the implementation supports the ```mean, variance, sample, log_prob``` methods.

cc: fritzo and neerajprad

- [x] add import and `__all__` string in `torch/distributions/__init__.py`
- [x] register docs in docs/source/distributions.rst

### Tests
(all tests live in tests/distributions.py)
- [x] add an `Example(MixtureSameFamily, [...])` to the `EXAMPLES` list,
     populating `[...]` with three examples:
     one with `Normal`, one with `Categorical`, and one with `MultivariateNormal`
     (to exercise, `FloatTensor`, `LongTensor`, and nontrivial `event_dim`)
- [x] add a `test_mixture_same_family_shape()` to `TestDistributions`. It would be good to test this with both `Normal` and `MultivariateNormal`
- [x] add a `test_mixture_same_family_log_prob()` to `TestDistributions`.
- [x] add a `test_mixture_same_family_sample()` to `TestDistributions`.
- [x] add a `test_mixture_same_family_shape()` to `TestDistributionShapes`

### Triaged for follup-up PR?
- support batch shape
- implement `.expand()`
- implement `kl_divergence()` in torch/distributions/kl.py
Pull Request resolved: https://github.com/pytorch/pytorch/pull/22742

Differential Revision: D19899726

Pulled By: ezyang

fbshipit-source-id: 9c816e83a2ef104fe3ea3117c95680b51c7a2fa4
 2020-02-14 10:31:56 -08:00
George Guanheng Zhang
0c98939b7b Revert D19899550: [pytorch][PR] Second try on Von Mises: Make it JIT compatible 
Test Plan: revert-hammer

Differential Revision:
D19899550

Original commit changeset: fbcdd9bc9143

fbshipit-source-id: c8a675a8b53f884acd0e6c57bc7aa15faf83d5d6
 2020-02-14 08:42:16 -08:00
Ahmad Salim Al-Sibahi
b1583ceb1e Second try on Von Mises: Make it JIT compatible (#33177) 
Summary:
Follow up from https://github.com/pytorch/pytorch/issues/17168 .
Pull Request resolved: https://github.com/pytorch/pytorch/pull/33177

Differential Revision: D19899550

Pulled By: ezyang

fbshipit-source-id: fbcdd9bc91438164bcb2b1cbc314c765520754e1
 2020-02-14 07:16:41 -08:00
Michael Carilli
40246fa63c Gradient scaling API (#26512) 
Summary:
This PR implements the gradient scaling API that mruberry, jjsjann123, ngimel, zdevito, gchanan and I have been discussing.  Relevant issue/RFC: https://github.com/pytorch/pytorch/issues/25081.

Volume-wise, this PR is mostly documentation and tests.  The Python API (found entirely in `torch/cuda/amp/amp_scaler.py`) is lightweight .  The exposed functions are intended to make the implementation and control flow of gradient scaling convenient, intuitive, and performant.

The API is probably easiest to digest by looking at the documentation and examples. `docs/source/amp.rst` is the homepage for the Automatic Mixed Precision package.  `docs/source/notes/amp_examples.rst` includes several examples demonstrating common but not-immediately-obvious use cases.  Examples are backed by tests in `test_cuda.py` (and thankfully the tests pass :P).

Two small utility kernels have been added in `native/cuda/AmpKernels.cu` to improve performance and avoid host-device synchronizations wherever possible.

Existing optimizers, both in the wild and in Pytorch core, do not need to change to use the scaling API.

However, the API was also designed to establish a contract between user scripts and optimizers such that writers of _new_ custom optimizers have the control points they need to implement fast, optionally sync-free updates.  User scripts that obey the scaling API can drop such custom optimizers in and reap performance benefits without having to change anything aside from the optimizer constructor itself.  [I know what the contract with custom optimizers should be](35829f24ef/torch/cuda/amp/amp_scaler.py (L179-L184)), but I'm waiting for review on the rest of the API before I go about documenting it (it will be given a dedicated section in `docs/source/notes/amp_examples.rst`.

Currently, the gradient scaling examples do not include the auto-casting API as discussed in https://github.com/pytorch/pytorch/issues/25081.  The gradient scaling API is intended to be orthogonal/modular relative to autocasting.  Without auto-casting the gradient scaling API is fully use-_able_, but not terribly use-_ful_, so it's up to you guys whether you want to wait until auto-casting is ready before merging the scaling API as well.

### Todo
- [ ] How do I get c10 registered status for my two custom kernels?  They're very simple.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/26512

Differential Revision: D19859905

Pulled By: mruberry

fbshipit-source-id: bb8ae6966214718dfee11345db824389e4286923
 2020-02-13 11:06:06 -08:00
Ilia Cherniavskii
04829e924a Update CPU threading doc (#33083) 
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/33083

Added more recommendations, some notes and warning

Test Plan: cd docs ; make html

Differential Revision: D19829133

Pulled By: ilia-cher

fbshipit-source-id: b9fbd89f5875b3ce35cc42ba75a3b44bb132c506
 2020-02-11 14:13:51 -08:00
Shinichiro Hamaji
478356aeec Fix broken links in governance.rst 
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/30815

Differential Revision: D19697401

Pulled By: ezyang

fbshipit-source-id: d7e1a1b54039624f471b6cfb568428feb73060f4
 2020-02-04 14:26:09 -08:00
Shinichiro Hamaji
67706187fb Fix a broken link in contribution_guide.rst 
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/30814

Differential Revision: D19697403

Pulled By: ezyang

fbshipit-source-id: b01fd0e189b3bc7ccaa197c9c64e12fee70a6310
 2020-02-04 14:14:25 -08:00
BowenBao
10183061eb [ONNX] Update ONNX landing page since 1.3 (#32805) 
Summary:
* New ops supported for exporting.
* Updates on support for tensor indexing and dynamic list of tensors.
* lara-hdr, spandantiwari Should we also include updates on torchvision support in this page?

cc houseroad, neginraoof Please review if I have missed anything.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/32805

Reviewed By: hl475

Differential Revision: D19635699

Pulled By: houseroad

fbshipit-source-id: b6be4fce641f852dcbceed20b4433f4037d8024a
 2020-02-03 10:38:29 -08:00
Edward Z. Yang
1177191c8e Synchronize with ShipIt. 
Signed-off-by: Edward Z. Yang <ezyang@fb.com>
 2020-01-21 13:39:28 -05:00
Brian Wignall
f326045b37 Fix typos, via a Levenshtein-type corrector (#31523) 
Summary:
Should be non-semantic.

Uses https://en.wikipedia.org/wiki/Wikipedia:Lists_of_common_misspellings/For_machines to find likely typos, with https://github.com/bwignall/typochecker to help automate the checking.

Uses an updated version of the tool used in https://github.com/pytorch/pytorch/pull/30606 .
Pull Request resolved: https://github.com/pytorch/pytorch/pull/31523

Differential Revision: D19216749

Pulled By: mrshenli

fbshipit-source-id: 7fd489cb9a77cd7e4950c1046f925d57524960ea
 2020-01-17 16:03:19 -08:00
anjali411
5b815d980e Added cummin 
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/32238

Differential Revision: D19416791

Pulled By: anjali411

fbshipit-source-id: 5aadc0a7a55af40d76f444ab7d7d47ec822f55a5
 2020-01-17 10:51:58 -08:00
Shen Li
322f34b245 Adding DDP Design Note 
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/32158

Test Plan: Imported from OSS

Differential Revision: D19405980

Pulled By: mrshenli

fbshipit-source-id: 808ef1c71b637546f8872375bf1828967b1a5a60
 2020-01-15 14:10:45 -08:00
Vamshi Chowdary
05088da8e9 [pytorch][PR] Fixed error in sample code of documentation (#31682) 
Summary:
"in_features" and "out_features" are not defined. Possibly a typo. They should be "input_features" and "output_features" instead
Pull Request resolved: https://github.com/pytorch/pytorch/pull/31682

Differential Revision: D19251685

Pulled By: zou3519

fbshipit-source-id: ac9e524e792a1853a16e8876d76b908495d8f35e
 2020-01-15 10:34:07 -08:00
anjali411
8dc67a014f Add cummax 
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/32169

Differential Revision: D19393236

Pulled By: anjali411

fbshipit-source-id: 5dac6b0a4038eb48458d4a0b253418daeccbb6bc
 2020-01-14 17:19:10 -08:00
Zafar Takhirov
701ca68882 Docs entry for the is_quantized 
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/32075

Test Plan: Imported from OSS

Differential Revision: D19353861

Pulled By: z-a-f

fbshipit-source-id: 4249216ac9a4af354a251c62181d65bc14cbfd3e
 2020-01-13 13:54:35 -08:00
Shen Li
62f93443e5 Explain RPC behavior when using Tensor as arg or return value 
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/31968

Test Plan: Imported from OSS

Differential Revision: D19321380

Pulled By: mrshenli

fbshipit-source-id: e3431f1f02963cc8d8266a420ab03866106f26ac
 2020-01-09 16:42:24 -08:00
Bram Wasti
021e1e20c1 Revert D19320493: Javadoc changes 
Test Plan: revert-hammer

Differential Revision:
D19320493

Original commit changeset: cc76b2a2acbe

fbshipit-source-id: 3b36dd2d2591acc60a06a421dd625c21adbe578a
 2020-01-09 14:23:30 -08:00
Jessica Lin
26f552a3d1 Javadoc changes (#31956) 
Summary:
- Add Javadoc url in index.rst
- Delete no longer needed java rst files
- Remove intersphinx extension from conf.oy
- Remove javasphinx from docs/requirements.txt
Pull Request resolved: https://github.com/pytorch/pytorch/pull/31956

Differential Revision: D19320493

Pulled By: jlin27

fbshipit-source-id: cc76b2a2acbe2ecdabcd3339e1cc3182f0c906ae
 2020-01-09 10:55:24 -08:00
xiaobing.zhang
9ba6a768de Add op bitwise_or (#31559) 
Summary:
ezyang ,  this PR add bitwise_or operator as https://github.com/pytorch/pytorch/pull/31104 .
Benchmark script :
```
import timeit
import torch
torch.manual_seed(1)

for n, t in [(10, 100000),(1000, 10000)]:
    print('__or__ (a.numel() == {}) for {} times'.format(n, t))
    for device in ('cpu', 'cuda'):
        for dtype in ('torch.int8', 'torch.uint8', 'torch.int16', 'torch.int32', 'torch.int64'):
            print(f'device: {device}, dtype: {dtype}, {t} times', end='\t\t')
            print(timeit.timeit(f'a | b\nif "{device}" == "cuda": torch.cuda.synchronize()', setup=f'import torch; a = torch.randint(0, 10, ({n},), dtype = {dtype}, device="{device}"); b = torch.randint(0, 10, ({n},), dtype = {dtype}, device="{device}")', number=t))

for n, t in [(10, 100000),(1000, 10000)]:
    print('__ior__ (a.numel() == {}) for {} times'.format(n, t))
    for device in ('cpu', 'cuda'):
        for dtype in ('torch.int8', 'torch.uint8', 'torch.int16', 'torch.int32', 'torch.int64'):
            print(f'device: {device}, dtype: {dtype}, {t} times', end='\t\t')
            print(timeit.timeit(f'a | b\nif "{device}" == "cuda": torch.cuda.synchronize()', setup=f'import torch; a = torch.randint(0, 10, ({n},), dtype = {dtype}, device="{device}"); b = torch.tensor(5, dtype = {dtype}, device="{device}")', number=t))
```
Device: **Tesla P100, skx-8180**
Cuda verison: **9.0.176**

Before:
```
__or__ (a.numel() == 10) for 100000 times
device: cpu, dtype: torch.int8, 100000 times            0.17616272252053022
device: cpu, dtype: torch.uint8, 100000 times           0.17148233391344547
device: cpu, dtype: torch.int16, 100000 times           0.17616403382271528
device: cpu, dtype: torch.int32, 100000 times           0.17717823758721352
device: cpu, dtype: torch.int64, 100000 times           0.1801931718364358
device: cuda, dtype: torch.int8, 100000 times           1.270583058707416
device: cuda, dtype: torch.uint8, 100000 times          1.2636413089931011
device: cuda, dtype: torch.int16, 100000 times          1.2839747751131654
device: cuda, dtype: torch.int32, 100000 times          1.2548385225236416
device: cuda, dtype: torch.int64, 100000 times          1.2650810535997152
__or__ (a.numel() == 1000) for 10000 times
device: cpu, dtype: torch.int8, 10000 times             0.031136621721088886
device: cpu, dtype: torch.uint8, 10000 times            0.030786747112870216
device: cpu, dtype: torch.int16, 10000 times            0.02391665056347847
device: cpu, dtype: torch.int32, 10000 times            0.024147341027855873
device: cpu, dtype: torch.int64, 10000 times            0.024414129555225372
device: cuda, dtype: torch.int8, 10000 times            0.12741921469569206
device: cuda, dtype: torch.uint8, 10000 times           0.1249831635504961
device: cuda, dtype: torch.int16, 10000 times           0.1283819805830717
device: cuda, dtype: torch.int32, 10000 times           0.12591975275427103
device: cuda, dtype: torch.int64, 10000 times           0.12655890546739101
__ior__ (a.numel() == 10) for 100000 times
device: cpu, dtype: torch.int8, 100000 times            0.3908365070819855
device: cpu, dtype: torch.uint8, 100000 times           0.38267823681235313
device: cpu, dtype: torch.int16, 100000 times           0.38239253498613834
device: cpu, dtype: torch.int32, 100000 times           0.3817988149821758
device: cpu, dtype: torch.int64, 100000 times           0.3901665909215808
device: cuda, dtype: torch.int8, 100000 times           1.4211318120360374
device: cuda, dtype: torch.uint8, 100000 times          1.4215159295126796
device: cuda, dtype: torch.int16, 100000 times          1.4307750314474106
device: cuda, dtype: torch.int32, 100000 times          1.4123614141717553
device: cuda, dtype: torch.int64, 100000 times          1.4480243818834424
__ior__ (a.numel() == 1000) for 10000 times
device: cpu, dtype: torch.int8, 10000 times             0.06468924414366484
device: cpu, dtype: torch.uint8, 10000 times            0.06442475505173206
device: cpu, dtype: torch.int16, 10000 times            0.05267547257244587
device: cpu, dtype: torch.int32, 10000 times            0.05286940559744835
device: cpu, dtype: torch.int64, 10000 times            0.06211103219538927
device: cuda, dtype: torch.int8, 10000 times            0.15332304500043392
device: cuda, dtype: torch.uint8, 10000 times           0.15353196952492
device: cuda, dtype: torch.int16, 10000 times           0.15300503931939602
device: cuda, dtype: torch.int32, 10000 times           0.15274472255259752
device: cuda, dtype: torch.int64, 10000 times           0.1512152962386608
```
After:
```
__or__ (a.numel() == 10) for 100000 times
device: cpu, dtype: torch.int8, 100000 times            0.2465507509186864
device: cpu, dtype: torch.uint8, 100000 times           0.2472386620938778
device: cpu, dtype: torch.int16, 100000 times           0.2469814233481884
device: cpu, dtype: torch.int32, 100000 times           0.2535214088857174
device: cpu, dtype: torch.int64, 100000 times           0.24855613708496094
device: cuda, dtype: torch.int8, 100000 times           1.4351346511393785
device: cuda, dtype: torch.uint8, 100000 times          1.4434308474883437
device: cuda, dtype: torch.int16, 100000 times          1.4520929995924234
device: cuda, dtype: torch.int32, 100000 times          1.4456610176712275
device: cuda, dtype: torch.int64, 100000 times          1.4580101007595658
__or__ (a.numel() == 1000) for 10000 times
device: cpu, dtype: torch.int8, 10000 times             0.029985425993800163
device: cpu, dtype: torch.uint8, 10000 times            0.03024935908615589
device: cpu, dtype: torch.int16, 10000 times            0.026356655173003674
device: cpu, dtype: torch.int32, 10000 times            0.027377349324524403
device: cpu, dtype: torch.int64, 10000 times            0.029163731262087822
device: cuda, dtype: torch.int8, 10000 times            0.14540370367467403
device: cuda, dtype: torch.uint8, 10000 times           0.1456305105239153
device: cuda, dtype: torch.int16, 10000 times           0.1450125053524971
device: cuda, dtype: torch.int32, 10000 times           0.1472016740590334
device: cuda, dtype: torch.int64, 10000 times           0.14709716010838747
__ior__ (a.numel() == 10) for 100000 times
device: cpu, dtype: torch.int8, 100000 times            0.27195510920137167
device: cpu, dtype: torch.uint8, 100000 times           0.2692424338310957
device: cpu, dtype: torch.int16, 100000 times           0.27726674638688564
device: cpu, dtype: torch.int32, 100000 times           0.2815811652690172
device: cpu, dtype: torch.int64, 100000 times           0.2852728571742773
device: cuda, dtype: torch.int8, 100000 times           1.4743850827217102
device: cuda, dtype: torch.uint8, 100000 times          1.4766502184793353
device: cuda, dtype: torch.int16, 100000 times          1.4774163831025362
device: cuda, dtype: torch.int32, 100000 times          1.4749693805351853
device: cuda, dtype: torch.int64, 100000 times          1.5772947426885366
__ior__ (a.numel() == 1000) for 10000 times
device: cpu, dtype: torch.int8, 10000 times             0.03614502027630806
device: cpu, dtype: torch.uint8, 10000 times            0.03619729354977608
device: cpu, dtype: torch.int16, 10000 times            0.0319912089034915
device: cpu, dtype: torch.int32, 10000 times            0.03319283854216337
device: cpu, dtype: torch.int64, 10000 times            0.0343862259760499
device: cuda, dtype: torch.int8, 10000 times            0.1581476852297783
device: cuda, dtype: torch.uint8, 10000 times           0.15974601730704308
device: cuda, dtype: torch.int16, 10000 times           0.15957212820649147
device: cuda, dtype: torch.int32, 10000 times           0.16002820804715157
device: cuda, dtype: torch.int64, 10000 times           0.16129320487380028
```

Fix  https://github.com/pytorch/pytorch/issues/24511, https://github.com/pytorch/pytorch/issues/24515, https://github.com/pytorch/pytorch/issues/24658, https://github.com/pytorch/pytorch/issues/24662.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/31559

Differential Revision: D19315875

Pulled By: ezyang

fbshipit-source-id: 4a3ca88fdafbeb796079687e676228111eb44aad
 2020-01-08 15:06:30 -08:00
Jessica Lin
c888473b57 Restructure docs organization and naming (#31849) 
Summary:
* Rename “Other Languages” → “Language Bindings”
* Move the Community section to the bottom
* Move "Language Bindings" above "Python API"
Pull Request resolved: https://github.com/pytorch/pytorch/pull/31849

Differential Revision: D19290966

Pulled By: jlin27

fbshipit-source-id: 30b579e032a9fb1636e4afc7bbbd85a2708f637d
 2020-01-07 11:16:53 -08:00
Rohan Varma
a561a8448b minor doc tweak to use mp.spawn in example (#30381) 
Summary:
Per pietern's comment in https://github.com/pytorch/pytorch/issues/30022, we can make this example launcher a bit simpler by using `torch.multiprocessing`.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/30381

Differential Revision: D19292080

Pulled By: rohan-varma

fbshipit-source-id: 018ace945601166ef3af05d8c3e69d900bd77c3b
 2020-01-06 22:19:01 -08:00
xiaobing.zhang
b47e9b97a2 Add op bitwise_and (#31104) 
Summary:
Refer to https://github.com/pytorch/pytorch/pull/25665,  add `bitwise_and` operator.
Benchmark script :
```
import timeit
#for __and__
for n, t in [(10, 100000),(1000, 10000)]:
    print('__and__ (a.numel() == {}) for {} times'.format(n, t))
    for device in ('cpu', 'cuda'):
        for dtype in ('torch.int8', 'torch.uint8', 'torch.int16', 'torch.int32', 'torch.int64'):
            print(f'device: {device}, dtype: {dtype}, {t} times', end='\t\t')
            print(timeit.timeit(f'a & b\nif "{device}" == "cuda": torch.cuda.synchronize()', setup=f'import torch; a = torch.randint(0, 10, ({n},), dtype = {dtype}, device="{device}"); b = torch.randint(0, 10, ({n},), dtype = {dtype}, device="{device}")', number=t))
#for __iand__
for n, t in [(10, 100000),(1000, 10000)]:
    print('__iand__ (a.numel() == {}) for {} times'.format(n, t))
    for device in ('cpu', 'cuda'):
        for dtype in ('torch.int8', 'torch.uint8', 'torch.int16', 'torch.int32', 'torch.int64'):
            print(f'device: {device}, dtype: {dtype}, {t} times', end='\t\t')
            print(timeit.timeit(f'a & b\nif "{device}" == "cuda": torch.cuda.synchronize()', setup=f'import torch; a = torch.randint(0, 10, ({n},), dtype = {dtype}, device="{device}"); b = torch.tensor(5, dtype = {dtype}, device="{device}")', number=t))
```
Device: **Tesla P100, skx-8180**
Cuda verison: **9.0.176**

Before:
```
__and__ (a.numel() == 10) for 100000 times
device: cpu, dtype: torch.int8, 100000 times            0.1766007635742426
device: cpu, dtype: torch.uint8, 100000 times           0.17322628945112228
device: cpu, dtype: torch.int16, 100000 times           0.17650844901800156
device: cpu, dtype: torch.int32, 100000 times           0.17711848113685846
device: cpu, dtype: torch.int64, 100000 times           0.18240160401910543
device: cuda, dtype: torch.int8, 100000 times           1.273967768996954
device: cuda, dtype: torch.uint8, 100000 times          1.2778537990525365
device: cuda, dtype: torch.int16, 100000 times          1.2753686187788844
device: cuda, dtype: torch.int32, 100000 times          1.2797665279358625
device: cuda, dtype: torch.int64, 100000 times          1.2933144550770521
__and__ (a.numel() == 1000) for 10000 times
device: cpu, dtype: torch.int8, 10000 times             0.031139614060521126
device: cpu, dtype: torch.uint8, 10000 times            0.03091452084481716
device: cpu, dtype: torch.int16, 10000 times            0.022756479680538177
device: cpu, dtype: torch.int32, 10000 times            0.025045674294233322
device: cpu, dtype: torch.int64, 10000 times            0.024164282716810703
device: cuda, dtype: torch.int8, 10000 times            0.12820732593536377
device: cuda, dtype: torch.uint8, 10000 times           0.12775669433176517
device: cuda, dtype: torch.int16, 10000 times           0.12697868794202805
device: cuda, dtype: torch.int32, 10000 times           0.12832533661276102
device: cuda, dtype: torch.int64, 10000 times           0.1280576130375266
__iand__ (a.numel() == 10) for 100000 times
device: cpu, dtype: torch.int8, 100000 times            0.3687064303085208
device: cpu, dtype: torch.uint8, 100000 times           0.36253443732857704
device: cpu, dtype: torch.int16, 100000 times           0.362891579978168
device: cpu, dtype: torch.int32, 100000 times           0.37680106051266193
device: cpu, dtype: torch.int64, 100000 times           0.3689364707097411
device: cuda, dtype: torch.int8, 100000 times           1.419940729625523
device: cuda, dtype: torch.uint8, 100000 times          1.4247053815051913
device: cuda, dtype: torch.int16, 100000 times          1.4191444097086787
device: cuda, dtype: torch.int32, 100000 times          1.4305962566286325
device: cuda, dtype: torch.int64, 100000 times          1.4567416654899716
__iand__ (a.numel() == 1000) for 10000 times
device: cpu, dtype: torch.int8, 10000 times             0.06224383972585201
device: cpu, dtype: torch.uint8, 10000 times            0.06205617543309927
device: cpu, dtype: torch.int16, 10000 times            0.05016433447599411
device: cpu, dtype: torch.int32, 10000 times            0.05216377507895231
device: cpu, dtype: torch.int64, 10000 times            0.06139362137764692
device: cuda, dtype: torch.int8, 10000 times            0.14827249851077795
device: cuda, dtype: torch.uint8, 10000 times           0.14801877550780773
device: cuda, dtype: torch.int16, 10000 times           0.14952312968671322
device: cuda, dtype: torch.int32, 10000 times           0.14999118447303772
device: cuda, dtype: torch.int64, 10000 times           0.14951884001493454
```
After:
```
__and__ (a.numel() == 10) for 100000 times
device: cpu, dtype: torch.int8, 100000 times            0.23157884553074837
device: cpu, dtype: torch.uint8, 100000 times           0.23063660878688097
device: cpu, dtype: torch.int16, 100000 times           0.23005440644919872
device: cpu, dtype: torch.int32, 100000 times           0.23748818412423134
device: cpu, dtype: torch.int64, 100000 times           0.24106105230748653
device: cuda, dtype: torch.int8, 100000 times           1.4394256137311459
device: cuda, dtype: torch.uint8, 100000 times          1.4436759827658534
device: cuda, dtype: torch.int16, 100000 times          1.4631587155163288
device: cuda, dtype: torch.int32, 100000 times          1.459101552143693
device: cuda, dtype: torch.int64, 100000 times          1.4784048134461045
__and__ (a.numel() == 1000) for 10000 times
device: cpu, dtype: torch.int8, 10000 times             0.028442862443625927
device: cpu, dtype: torch.uint8, 10000 times            0.028130197897553444
device: cpu, dtype: torch.int16, 10000 times            0.025318274274468422
device: cpu, dtype: torch.int32, 10000 times            0.02519288007169962
device: cpu, dtype: torch.int64, 10000 times            0.028299466706812382
device: cuda, dtype: torch.int8, 10000 times            0.14342594426125288
device: cuda, dtype: torch.uint8, 10000 times           0.145280827768147
device: cuda, dtype: torch.int16, 10000 times           0.14673697855323553
device: cuda, dtype: torch.int32, 10000 times           0.14499565307050943
device: cuda, dtype: torch.int64, 10000 times           0.14582364354282618
__iand__ (a.numel() == 10) for 100000 times
device: cpu, dtype: torch.int8, 100000 times            0.25548241566866636
device: cpu, dtype: torch.uint8, 100000 times           0.2552562616765499
device: cpu, dtype: torch.int16, 100000 times           0.25905191246420145
device: cpu, dtype: torch.int32, 100000 times           0.26635489892214537
device: cpu, dtype: torch.int64, 100000 times           0.26269810926169157
device: cuda, dtype: torch.int8, 100000 times           1.485458506271243
device: cuda, dtype: torch.uint8, 100000 times          1.4742380809038877
device: cuda, dtype: torch.int16, 100000 times          1.507783885113895
device: cuda, dtype: torch.int32, 100000 times          1.4926990242674947
device: cuda, dtype: torch.int64, 100000 times          1.519851053133607
__iand__ (a.numel() == 1000) for 10000 times
device: cpu, dtype: torch.int8, 10000 times             0.03425929415971041
device: cpu, dtype: torch.uint8, 10000 times            0.03293587639927864
device: cpu, dtype: torch.int16, 10000 times            0.029559112153947353
device: cpu, dtype: torch.int32, 10000 times            0.030915481969714165
device: cpu, dtype: torch.int64, 10000 times            0.03292469773441553
device: cuda, dtype: torch.int8, 10000 times            0.15792148280888796
device: cuda, dtype: torch.uint8, 10000 times           0.16000914946198463
device: cuda, dtype: torch.int16, 10000 times           0.1600684942677617
device: cuda, dtype: torch.int32, 10000 times           0.16162546630948782
device: cuda, dtype: torch.int64, 10000 times           0.1629159888252616
```
Fix  https://github.com/pytorch/pytorch/issues/24508, https://github.com/pytorch/pytorch/issues/24509,  https://github.com/pytorch/pytorch/issues/24655, https://github.com/pytorch/pytorch/issues/24656.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/31104

Differential Revision: D18938930

Pulled By: VitalyFedyunin

fbshipit-source-id: a77e805a0b84e8ace16c6e648c2f67dad44f2e44
 2020-01-03 10:32:36 -08:00
vishwakftw
22d84204f7 Expose torch.poisson in documentation (#31667) 
Summary:
Changelog:
- Add doc string for torch.poisson briefing current behavior
- Check for non-positive entries in the tensor passed as input to torch.poisson

Closes https://github.com/pytorch/pytorch/issues/31646
Pull Request resolved: https://github.com/pytorch/pytorch/pull/31667

Differential Revision: D19247371

Pulled By: ngimel

fbshipit-source-id: b53d105e73bf59a45beeb566f47365c3eb74efca
 2019-12-28 21:32:26 -08:00
davidriazati
ec4e347744 Add Python language reference docs (#30686) 
Summary:
This exposes our audit of https://docs.python.org/3/reference/ with descriptions for each line item.

To generate the `.rst` from the Quip:

```bash
pip install m2r
m2r jit_language_reference.md
```

https://driazati.github.io/pytorch_doc_previews/30686/jit.html#python-functions-and-modules
Pull Request resolved: https://github.com/pytorch/pytorch/pull/30686

Pulled By: driazati

Differential Revision: D19219587

fbshipit-source-id: 249db9b5ee20e38804d4302bbfeca7d54f27d0bd
 2019-12-26 13:21:36 -08:00
Martin Yuan
11854bcd38 Add test to torch.jit.export_opnames, make the _C function private 
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/31446

Test Plan: Imported from OSS

Differential Revision: D19172851

Pulled By: iseeyuan

fbshipit-source-id: f06d8766ed73c9abe4ebf41c402ee64880d745be
 2019-12-20 13:38:43 -08:00
Elias Ellison
779b128872 add back in reference to jit_unsupported section (#31486) 
Summary:
It was added in https://github.com/pytorch/pytorch/pull/31329 and removed in a bad merge in https://github.com/pytorch/pytorch/pull/31138/
Pull Request resolved: https://github.com/pytorch/pytorch/pull/31486

Differential Revision: D19181967

Pulled By: eellison

fbshipit-source-id: 7e4b4a9b2042c30ec18f7f737bc4a9a56fac7d92
 2019-12-19 12:44:16 -08:00
davidriazati
503a4e9019 Cleanup after moving language reference (#31146) 
Summary:
Stacked PRs
 * **#31146 - [jit] Cleanup after moving language reference**
 * #31138 - [jit] Move TorchScript language reference to its own page

Preview: https://driazati.github.io/pytorch_doc_previews/jit.html#torchscript-language

Pull Request resolved: https://github.com/pytorch/pytorch/pull/31146

Pulled By: driazati

Differential Revision: D19167390

fbshipit-source-id: f28daed36754a553264fc8ac142ed22c3e26d63e
 2019-12-18 15:09:35 -08:00
davidriazati
ae2487bf4d Move TorchScript language reference to its own page (#31138) 
Summary:
Stacked PRs
 * #31146 - [jit] Cleanup after moving language reference
 * **#31138 - [jit] Move TorchScript language reference to its own page**

Preview: https://driazati.github.io/pytorch_doc_previews/jit.html#torchscript-language

Pull Request resolved: https://github.com/pytorch/pytorch/pull/31138

Pulled By: driazati

Differential Revision: D19167375

fbshipit-source-id: d37110d85fc8b8d2c741be49846e873de1357c2a
 2019-12-18 15:09:31 -08:00
Elias Ellison
fb30a48b4e add unsupported section (#31329) 
Summary:
Add a section for unsupported ops, and modules. Automatically generate the properties and attributes that aren't bound, and for ops that have semantic mismatches set up tests so the docs stay up to date.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/31329

Differential Revision: D19164472

Pulled By: eellison

fbshipit-source-id: 46290bb8a64d9de928cfb1eda5ff4558c3799c88
 2019-12-18 13:56:02 -08:00
Elliot Waite
c63f8e5ebe Fix typo in data.rst docs 
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/31395

Differential Revision: D19160010

Pulled By: zou3519

fbshipit-source-id: cbc4e719e69117e8747617729d240c72e7a4e3dd
 2019-12-18 09:52:10 -08:00
Vitaly Fedyunin
3e59e80429 Revert D18941024: Move TorchScript language reference to its own page 
Test Plan: revert-hammer

Differential Revision:
D18941024

Original commit changeset: d0ff600870a1

fbshipit-source-id: 01c0eac4c9741f27b91d710616e71a0d769f6f6a
 2019-12-18 08:55:50 -08:00
davidriazati
c05538b831 Move TorchScript language reference to its own page (#31138) 
Summary:
Preview: https://driazati.github.io/pytorch_doc_previews/jit.html#torchscript-language
](https://our.intern.facebook.com/intern/diff/18941024/)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/31138

Pulled By: driazati

Differential Revision: D18941024

fbshipit-source-id: d0ff600870a14c4a7c6ce54867d152072a12c48c
 2019-12-18 00:46:19 -08:00
Michael Suo
293a139d79 add a warning for script classes (#31069) 
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/31069

Just to clarify that they are still experimental.

Test Plan: Imported from OSS

Differential Revision: D18920496

Pulled By: suo

fbshipit-source-id: d2f3014592a01a21f7fc60a4ce46dd0bfe5e19e9
 2019-12-11 14:48:55 -08:00
Rohan Varma
dbc8b00816 Document WorkerInfo and RpcBackendOptions structures in RPC docs. (#31077) 
Summary:
We mention `WorkerInfo` and `RpcBackendOptions` in a couple of different locations in our docs, and these are public classes that the user may use, so we should add the class to the documentation.
<img width="978" alt="Screen Shot 2019-12-10 at 1 42 22 PM" src="https://user-images.githubusercontent.com/8039770/70571759-47db2080-1b53-11ea-9d61-c83985a29dd9.png">
Pull Request resolved: https://github.com/pytorch/pytorch/pull/31077

Differential Revision: D18928162

Pulled By: rohan-varma

fbshipit-source-id: 67f11eedd87523c469377b791a0ba23704ec3723
 2019-12-11 11:39:57 -08:00
Michael Suo
d02280b432 move migration guide to appendix (#31068) 
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/31068

Let's get it out of the early parts now that the recursive API has been
around for a while

Test Plan: Imported from OSS

Differential Revision: D18920498

Pulled By: suo

fbshipit-source-id: 6f4389739dd9e7e5f3014811b452249cc21d88e7
 2019-12-10 18:04:02 -08:00
TH3CHARLie
5edfe9cb80 add torch.square (#30719) 
Summary:
fixes https://github.com/pytorch/pytorch/issues/30524
This adds an new operator `torch.square` to PyTorch

I think it is ready for the first-time review now albanD
Pull Request resolved: https://github.com/pytorch/pytorch/pull/30719

Differential Revision: D18909268

Pulled By: albanD

fbshipit-source-id: 5626c445d8db20471a56fc1d7a3490e77812662b
 2019-12-10 15:22:46 -08:00
Elias Ellison
f48a8901c5 Add floor_divide function (#30493) 
Summary:
Adds `torch.floor_divide` following the numpy's `floor_divide` api. I only implemented the out-of-place version, I can add the inplace version if requested.

Also fixes  https://github.com/pytorch/pytorch/issues/27512
Pull Request resolved: https://github.com/pytorch/pytorch/pull/30493

Differential Revision: D18896211

Pulled By: eellison

fbshipit-source-id: ee401c96ab23a62fc114ed3bb9791b8ec150ecbd
 2019-12-10 07:51:39 -08:00
Joseph Spisak
7af9d77290
Update persons_of_interest.rst 
Updating to add POI for mobile, quantization and an addition to optimizers.
 2019-12-05 21:20:40 -08:00
davidriazati
2308a0ec1b Improve documentation around builtin functions (#30347) 
Summary:
This breaks the builtins page into some more sections and adds details about Python built-in functions
](https://our.intern.facebook.com/intern/diff/18718166/)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/30347

Pulled By: driazati

Reviewed By: wanchaol

Differential Revision: D18718166

fbshipit-source-id: bf43260ab7bcf92cccef684a5ce68cb16020771d
 2019-12-04 13:50:40 -08:00
Nathan Goldbaum
9d3402e4cb Add the __torch_function__ API override mechanism (#30730) 
Summary:
This is a re-do of https://github.com/pytorch/pytorch/issues/27064, which was reverted (b8792c0438). This was landed at the same time as other work that added new operators to the `torch` namespace so the check for whether the `torch` namespace is exhaustively checked for overridability was triggering test failures.

I've temporarily disabled that check and added an explanatory comment that the check will be re-enabled in a future PR that will be merged during a time when the commit velocity on PyTorch is lower.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/30730

Differential Revision: D18813270

Pulled By: ezyang

fbshipit-source-id: 70477c4656dca8fea6e7bc59259555041fcfbf68
 2019-12-04 13:19:07 -08:00
Tongzhou Wang
d0af07ca4c Fix capitalization inconsistency in optim.rst 
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/30608

Differential Revision: D18808516

Pulled By: ezyang

fbshipit-source-id: 4be68be9a8c8c3da7a0b98162bc1050b588fab43
 2019-12-04 08:17:03 -08:00
Edward Yang
b8792c0438 Revert D18645954: add __torch_function__ API override mechanism 
Test Plan: revert-hammer

Differential Revision:
D18645954

Original commit changeset: 54b5e4344d7a

fbshipit-source-id: 4a7aebb483e6b001130d6f384ccc53c5a808ab13
 2019-12-04 07:41:47 -08:00
Prasun Anand
d12786b24f add __torch_function__ API override mechanism (#27064) 
Summary:
Closes https://github.com/pytorch/pytorch/issues/24015 (see description of that issue for more details).

For a toy example, see the `DiagonalTensor` and `SubDiagonalTensor` class in test/test_overrides.py.

This PR currently contains:

* tests for `__torch_function__` behavior
* modification to `gen_python_functions` and `parse` function signatures and dispatched to correct overloaded argument.

This feature is inspired by and analogous to NumPy's `__array_function__` protocol ([see NumPy Enhancement Proposal 18](https://numpy.org/neps/nep-0018-array-function-protocol.html#trying-array-function-methods-until-the-right-one-works)).

### Benchmarks:
See Nathan's comment below: https://github.com/pytorch/pytorch/pull/27064#issuecomment-554601189
Pull Request resolved: https://github.com/pytorch/pytorch/pull/27064

Differential Revision: D18645954

Pulled By: ezyang

fbshipit-source-id: 54b5e4344d7afdbcf996bb57191b0bdadc7b1767
 2019-12-04 05:56:46 -08:00
Martin Yuan
b26401f965 Dump operator names of a script module (#30467) 
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/30467

Introduce function jit.export_opnames(module), which returns a list of all operator names used in the module and its submodules. One usage is to have mobile custom build to link only operators in the returned list to save the mobile size.

Example:
import torch
m = torch.jit.load("example.pt")
print(torch.jit.export_opnames(m))

The outputs are in alphabetical order:
['aten::_convolution', 'aten::add.Tensor', 'aten::add_.Tensor', 'aten::addmm', 'aten::append.Tensor', 'aten::cat', 'aten::dropout', 'aten::embedding', 'aten::matmul', 'aten::max.dim', 'aten::mul.Tensor', 'aten::permute', 'aten::relu', 'aten::t', 'aten::tanh', 'prim::ListConstruct', 'prim::TupleConstruct', 'prim::TupleUnpack']

Test Plan: Imported from OSS

Differential Revision: D18801619

Pulled By: iseeyuan

fbshipit-source-id: f9b198d3e82b095daf704ee595d8026ad889bb13
 2019-12-03 20:20:33 -08:00
Hong Xu
bb5dcaf24f Add logical_and and logical_or (#30521) 
Summary:
With the CI failure caused in 8bbafa0b32 fixed (incorrect return type of the lambdas in CUDA kernels)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/30521

Differential Revision: D18770151

Pulled By: ailzhang

fbshipit-source-id: 02f0fe1d5718c34d24da6dbb5884ee8b247ce39a
 2019-12-03 18:24:54 -08:00
Joseph Spisak
4d4d8e0dce Update persons_of_interest.rst (#30647) 
Summary:
Adding back the 3 names for the MSFT team - re: ONNX Governance.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/30647

Differential Revision: D18781163

Pulled By: jlin27

fbshipit-source-id: 7284ba29841ab41b9807c9d92694630b50de7b6a
 2019-12-03 14:46:15 -08:00
Brian Wignall
e7fe64f6a6 Fix typos (#30606) 
Summary:
Should be non-semantic.

Uses https://en.wikipedia.org/wiki/Wikipedia:Lists_of_common_misspellings/For_machines to find likely typos.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/30606

Differential Revision: D18763028

Pulled By: mrshenli

fbshipit-source-id: 896515a2156d062653408852e6c04b429fc5955c
 2019-12-02 20:17:42 -08:00
peterjc123
6deb41c88d Update magma to 2.5.1 for Windows and switch CUDA in CI to 9.2 
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/30513

Differential Revision: D18764184

Pulled By: ezyang

fbshipit-source-id: 4992869fd6a89471a5d25eb6a9b44ad8eceb480f
 2019-12-02 11:56:10 -08:00
Shen Li
ec5e471647 Reorganize rpc API doc and add introduction (#30491) 
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/30491

Our RPC API docs presents the APIs well but misses a general
introduction to the APIs. Readers might be a little lost the first
time landing this page. This commits reorganizes the APIs into
four components from user's perspective, RPC, RRef, dist autograd,
and dist optimizer. It also adds an intro to each and briefly
discribes why we provide those.

Test Plan: Imported from OSS

Differential Revision: D18723294

Pulled By: mrshenli

fbshipit-source-id: 4aced4ab537b070aa780aaaf9724659fd47cb3cb
 2019-11-28 15:34:18 -08:00
Rohan Varma
1350b99de4 Add local shutdown to process group agent (#30330) 
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/30330

This is now possible due to previous changes made in `gloo` and `ProcessGroupGloo`. We `abort` the listener thread that is waiting for a message, and join all other threads. The API is changed so that the previous `wait_all_workers` does not destroy the agent, and this is now done in a new `shutdown` method. All callsites are updated appropriately.

ghstack-source-id: 94673884
ghstack-source-id: 94673884

Test Plan: Unit tests pass.

Reviewed By: mrshenli

Differential Revision: D18661775

fbshipit-source-id: 5aaa7c14603e18253394224994f6cd43234301c2
 2019-11-27 22:34:08 -08:00