- 11 6月, 2019 8 次提交
-
-
由 Nikolay Korovaiko 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21381 Differential Revision: D15724412 Pulled By: Krovatkin fbshipit-source-id: 18e4a1916c7cd1baea76953d0087d6257e58c55b
-
由 Vishwak Srinivasan 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21590 Differential Revision: D15742549 Pulled By: ezyang fbshipit-source-id: fd5b2cbce86e5f229c2ffba114ef362934296d07
-
由 Elias Ellison 提交于
Summary: test that wasn't on the CI, but is tested internally. Pull Request resolved: https://github.com/pytorch/pytorch/pull/21594 Differential Revision: D15742157 Pulled By: eellison fbshipit-source-id: 11fc82d1fc0281ffedd674ed96100e0c783c0599
-
由 fehiepsi 提交于
Summary: This PR addresses some numerical issues of Sigmoid/StickBreakingTransform, where these transforms give +-inf when the unconstrained values move to +-20 areas. For example, with ``` t = torch.distributions.SigmoidTransform() x = torch.tensor(20.) t.inv(t(x)), t.log_abs_det_jacobian(x, t(x)) ``` current behaviour the inverse will return `inf` and logdet return `-inf` while this PR makes it to `15.9424` and `-15.9424`. And for ``` t = torch.distributions.StickBreakingTransform() x = torch.tensor([20., 20.]) t.inv(t(x)), t.log_abs_det_jacobian(x, t(x)) ``` current value is `(inf, nan)` and `-inf` for logdet, while this PR makes it `[16.6355, 71.3942]` and `-47.8272` for logdet. Although these finite values are wrong and seems unavoidable, it is better than returning `inf` or `nan` in my opinion. This is useful in HMC where despite that the grad will be zero when the unconstrained parameter moves to unstable area (due to clipping), velocity variable will force the parameter move to another area which by chance can move the parameter out of unstable area. But inf/nan can be useful to stop doing inference early. So the changes in this PR might be inappropriate. I also fix some small issues of `_Simplex` and `_RealVector` constraints where batch shape of the input is not respected when checking validation. Pull Request resolved: https://github.com/pytorch/pytorch/pull/20288 Differential Revision: D15742047 Pulled By: ezyang fbshipit-source-id: b427ed1752c41327abb3957f98d4b289307a7d17
-
由 Haixin Liu 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21532 Add python binding to deserialize blob Reviewed By: yinghai Differential Revision: D15706816 fbshipit-source-id: f498c7e0f7392f055b13810bbf81cba59f25e1d2
-
由 Elias Ellison 提交于
Summary: This changes our compiler so it first emits Loads & Stores, and then transforms the graph to SSA in a follow up pass. When a variable is set, we emit a prim::Store, and when a variable is referenced, we emit a prim::Load. ``` a = 1 print(a) ``` becomes: ``` %a.1 : int = prim::Constant[value=1]() prim::Store[name="a"](%a.1) %a : int = prim::Load[name="a"]() prim::Print(%a) ``` In the follow up pass, convertToSSA, the values are turned into SSA form with the Loads & Stores removed. This change will enable breaks and continues because you can transform the graph with the variable naming information still intact. There are still some remaining jitter and edge cases issues that I have to look through, but I think is still ready for eview. Pull Request resolved: https://github.com/pytorch/pytorch/pull/21101 Differential Revision: D15723353 Pulled By: eellison fbshipit-source-id: 3269934d4bc24ddaf3a87fdd20620b0f954d83d0
-
由 Ailing Zhang 提交于
Summary: update doc as pointed out in https://github.com/pytorch/hub/pull/22 Pull Request resolved: https://github.com/pytorch/pytorch/pull/21568 Differential Revision: D15732927 Pulled By: ailzhang fbshipit-source-id: 78ab026539e5ee59e7c3a8144e2c9fcbbc225733
-
由 mal 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21438 ghimport-source-id: 33f145f5b3508163365442c22a223c4a44e677d8 Differential Revision: D15738856 fbshipit-source-id: 656e8d0e3d0d22f116e3ab66bf0282608d6f1a76
-
- 10 6月, 2019 18 次提交
-
-
由 Jongsoo Park 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21382 Concat tensor inference function was not handling correctly the case where axis argument points to the last dimension so input tensors don't need to have the same number of dimensions. Split tensor inference function was not handling correctly the case where split information is provided as the second input tensor rather than as an argument. Reviewed By: mdschatz Differential Revision: D15633148 fbshipit-source-id: d566af44dc882457ee9efe83d2461b28408c2c5d
-
由 Malvika Joshi 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21548 Added documentation as titled. Reviewed By: ezyang Differential Revision: D15723146 fbshipit-source-id: fab4a35c62f07256673318c0874701f7628b2f7a
-
由 Richard Zou 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21373 ghimport-source-id: acab6d3a Differential Revision: D15717925 Pulled By: zou3519 fbshipit-source-id: 8515c448b368be79f71681833b5edf960da44fe8
-
由 Richard Zou 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21341 ghimport-source-id: 06021b06 Differential Revision: D15717907 Pulled By: zou3519 fbshipit-source-id: 48ee76cf2f11a8b092be75ecac8d5faee68ca0d9
-
由 peter 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21580 Differential Revision: D15738060 Pulled By: ezyang fbshipit-source-id: 05a2e97279816753d574678252bf9b35913c99b1
-
由 Xing Wang 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21319 Add assertion to raise Exception when Regularization is applied on FP16. Reviewed By: bddppq Differential Revision: D15528486 fbshipit-source-id: c887c90d1d9ccfdaded3b5fa16816c6f29910e2e
-
由 Iurii Zdebskyi 提交于
Summary: Following up on the feedback in this [PR](https://github.com/pytorch/pytorch/pull/21113/files?file-filters%5B%5D=.cwrap&owned-by%5B%5D=) Pull Request resolved: https://github.com/pytorch/pytorch/pull/21538 Differential Revision: D15721390 Pulled By: izdeby fbshipit-source-id: 1b5265bf8726c1051f306f7674d731e25a6c7d03
-
由 杨培文 (Yang Peiwen) 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21570 Differential Revision: D15732665 Pulled By: ezyang fbshipit-source-id: caa12a8619ad1396540f787b5c849d29cc5b03bd
-
由 Zachary DeVito 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21564 ghimport-source-id: b6f71e2238f6f7c8de6cfbf6969a5e08e07be46c Reviewed By: suo Differential Revision: D15729497 Pulled By: zdevito fbshipit-source-id: aacfea6058fadb572df692aa9ebd6cab0bcd03fc
-
由 Zachary DeVito 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21563 ghimport-source-id: de87e5e621da33326a9d2cb8a57d82d355166479 Reviewed By: suo Differential Revision: D15729499 Pulled By: zdevito fbshipit-source-id: 17b3e2e71d004f08c4413e80091388ae9ac2df2b
-
由 Zachary DeVito 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21562 ghimport-source-id: 17e5e183f730f50d97ef48973aafc6249d54978f Reviewed By: suo Differential Revision: D15729500 Pulled By: zdevito fbshipit-source-id: efa8a133b617b1498810392a8da6b513ce00b5eb
-
由 Zachary DeVito 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21561 ghimport-source-id: 4bf28d8140610a0cefef0c0a17f0a513ae855dde Reviewed By: suo Differential Revision: D15729498 Pulled By: zdevito fbshipit-source-id: b26458336da1efaba71d8a577c3917c6622dae0d
-
由 Zachary DeVito 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21560 ghimport-source-id: a555ca33fcd3efd1147aaf90f26a8e63da1c1a67 Reviewed By: suo Differential Revision: D15729502 Pulled By: zdevito fbshipit-source-id: d6c11472bfc791e2ad1e9aa695b0439d72b79681
-
由 Zachary DeVito 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21559 ghimport-source-id: 81ba4a5638577781e1ea706599966c033c37e814 Reviewed By: suo Differential Revision: D15729501 Pulled By: zdevito fbshipit-source-id: 3423bff61e89617c40078d5fab726b77d21bfa27
-
由 Zachary DeVito 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21558 ghimport-source-id: a8a19dbefea869ca1401e5afea6c02f31f95b99a Reviewed By: suo Differential Revision: D15729491 Pulled By: zdevito fbshipit-source-id: 9629664608a2379a2ddcafaf741fa8463c4fb917
-
由 Charles Lovering 提交于
Summary: - Fixes a typo. Pull Request resolved: https://github.com/pytorch/pytorch/pull/21510 Differential Revision: D15731277 Pulled By: ezyang fbshipit-source-id: c3f8e110f5c61e797b857477b495168ea8d63cd5
-
由 Brennan Vincent 提交于
Summary: Should be self-explanatory. This `int` variable is overflowing. Reported in #21526 Pull Request resolved: https://github.com/pytorch/pytorch/pull/21530 Differential Revision: D15719275 Pulled By: umanwizard fbshipit-source-id: 24e917a00a5b78bc3af29ef3b8b72eea7e89d5d5
-
由 Xiaomeng Yang 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21556 Optimize batch mm op when broadcast the second input Reviewed By: houseroad Differential Revision: D15728914 fbshipit-source-id: c60441d69d4997dd32a3566780496c7ccda5e67a
-
- 08 6月, 2019 14 次提交
-
-
由 Zachary DeVito 提交于
Differential Revision: D15572818 Original commit changeset: 3a9b5f053664 fbshipit-source-id: b932411e8e88c7414c8db332d6049fe4e26bd83e
-
由 Zachary DeVito 提交于
Differential Revision: D15590900 Original commit changeset: 98829979feba fbshipit-source-id: eb7f1d396bb2b98d2852af81c69db81430eba33c
-
由 Zachary DeVito 提交于
Differential Revision: D15600068 Original commit changeset: 9b68e23d7f8b fbshipit-source-id: 45f36b3aaa4f1c457c27490579496456cbbc680b
-
由 Zachary DeVito 提交于
Differential Revision: D15600067 Original commit changeset: 82b7de458dd6 fbshipit-source-id: ca26a362cd73bdb9e8c4eba15dd5c10986fa79fe
-
由 Zachary DeVito 提交于
Differential Revision: D15618275 Original commit changeset: 038ae27e5416 fbshipit-source-id: 8dbe0f564ba103fe445dacc471085c659171705f
-
由 Zachary DeVito 提交于
Differential Revision: D15719982 Original commit changeset: 56bb021dd949 fbshipit-source-id: 2eb3e2c9745c35a84ebcc0fc7ac62b5f1fdd6437
-
由 Zachary DeVito 提交于
Differential Revision: D15719980 Original commit changeset: 3de2746c3f3c fbshipit-source-id: 4610e215936b2476a0271355ef3b8f1f480bdea8
-
由 Zachary DeVito 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21514 ghimport-source-id: 6a9b6fdd7e696ea29e8715482708efe897230e4d Reviewed By: jamesr66a Differential Revision: D15719980 Pulled By: zdevito fbshipit-source-id: 3de2746c3f3c3de4111b4cb73f4c4acedbf28862
-
由 Zachary DeVito 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21513 ghimport-source-id: 86278929818a8fc65684bd8f2ffac31460772fe9 Reviewed By: jamesr66a Differential Revision: D15719982 Pulled By: zdevito fbshipit-source-id: 56bb021dd949668562ea481c5ff0115a9ea2b02e
-
由 Zachary DeVito 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21325 ghimport-source-id: eeca1176f5e00c85a69cd016acccf5105e670e02 Reviewed By: jamesr66a Differential Revision: D15618275 Pulled By: zdevito fbshipit-source-id: 038ae27e5416f1ce338009627c839a4d61a00658
-
由 Zachary DeVito 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21273 ghimport-source-id: b92c1e07fbe4122467a21b98d29635295093e0c2 Reviewed By: jamesr66a Differential Revision: D15600067 Pulled By: zdevito fbshipit-source-id: 82b7de458dd65c175f55b0f383bfc3fcf4704032
-
由 Zachary DeVito 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21272 ghimport-source-id: 43e73d1b93ccbe0dd6845eb3f7444c9d0abd444b Reviewed By: jamesr66a Differential Revision: D15600068 Pulled By: zdevito fbshipit-source-id: 9b68e23d7f8b6046a5a0d6d9fd16138ac384b863
-
由 Zachary DeVito 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21240 ghimport-source-id: 5e9cbe8b3df3ac721135d2f652a420ae0b14ac55 Reviewed By: jamesr66a Differential Revision: D15590900 Pulled By: zdevito fbshipit-source-id: 98829979feba23685f0ba98ba3cb840157f7259a
-
由 Zachary DeVito 提交于
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/21185 ghimport-source-id: 6b9cb92d1f1f59bb980dcfa0d29dfe985ee955d1 Reviewed By: jamesr66a Differential Revision: D15572818 Pulled By: zdevito fbshipit-source-id: 3a9b5f053664c09212b97f1391d8d006337b5550
-