{"payload":{"feedbackUrl":"https://github.com/orgs/community/discussions/53140","repo":{"id":187099494,"defaultBranch":"master","name":"pytorch","ownerLogin":"yifuwang","currentUserCanPush":false,"isFork":true,"isEmpty":false,"createdAt":"2019-05-16T20:50:24.000Z","ownerAvatar":"https://avatars.githubusercontent.com/u/4156752?v=4","public":true,"private":false,"isOrgOwned":false},"refInfo":{"name":"","listCacheKey":"v0:1721857269.0","currentOid":""},"activityList":{"items":[{"before":null,"after":"16d7b59eb1a0fc25c956e679684986a594823090","ref":"refs/heads/reduce_scatter","pushedAt":"2024-07-24T21:41:09.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"},"commit":{"message":"one_shot_reduce_scatter","shortMessageHtmlLink":"one_shot_reduce_scatter"}},{"before":null,"after":"176c4342daca063112c0ec91ee3d89738e1c4657","ref":"refs/heads/cuStreamWaitValue","pushedAt":"2024-07-24T21:41:09.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"},"commit":{"message":"test cuStreamWaitValue","shortMessageHtmlLink":"test cuStreamWaitValue"}},{"before":null,"after":"1f610876a92de61ec8326bd423204dcf1b2d88cf","ref":"refs/heads/symm_mem_cuda_graph","pushedAt":"2024-07-24T21:41:09.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"},"commit":{"message":"symm_mem_cuda_graph","shortMessageHtmlLink":"symm_mem_cuda_graph"}},{"before":null,"after":"29a856a2a1bc8d97455d05f557a07474ff586e5a","ref":"refs/heads/nccl_mem_alloc","pushedAt":"2024-07-24T21:41:09.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"},"commit":{"message":"nccl 2.22.3-1","shortMessageHtmlLink":"nccl 2.22.3-1"}},{"before":null,"after":"4a663633ee8bfb0afdf437ecc8c4c5f22fe8ff82","ref":"refs/heads/will","pushedAt":"2024-07-24T21:41:09.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"},"commit":{"message":"fix reorder pass algo after rebase\n\nghstack-source-id: 58f6048e0b4295a006b51e97ec5c082480e1dac3\nPull Request resolved: https://github.com/pytorch/pytorch/pull/130276","shortMessageHtmlLink":"fix reorder pass algo after rebase"}},{"before":null,"after":"7c3f2425189b29690cd67a47b12882dbdc11d7a9","ref":"refs/heads/fp8","pushedAt":"2024-07-24T21:41:09.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"},"commit":{"message":"fp8 pass","shortMessageHtmlLink":"fp8 pass"}},{"before":null,"after":"91ebc78d81cc6f064e76d065136decd773424a8e","ref":"refs/heads/memory_planning","pushedAt":"2024-07-24T21:41:09.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"},"commit":{"message":"Proof-of-concept: manage registered communication buffers with Inductor\n\nghstack-source-id: 674419cefdf071b40b78b4444a5ac96814864435\nPull Request resolved: https://github.com/pytorch/pytorch/pull/129186","shortMessageHtmlLink":"Proof-of-concept: manage registered communication buffers with Inductor"}},{"before":null,"after":"958a8dec08135f891167be5461fb86387894e698","ref":"refs/heads/subgroup_fix","pushedAt":"2024-07-24T21:41:09.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"},"commit":{"message":"Replace manual parsing of \"TMPDIR\", \"TMP\", \"TEMP\" and \"TEMPDIR\" with std::filesystem::temp_directory_path()\n\nghstack-source-id: a19173e5a31c9a001259e922668bb10fc42ec101\nPull Request resolved: https://github.com/pytorch/pytorch/pull/130842","shortMessageHtmlLink":"Replace manual parsing of \"TMPDIR\", \"TMP\", \"TEMP\" and \"TEMPDIR\" with …"}},{"before":null,"after":"95a353465a108e09bdba1d7f3df2711a59c39317","ref":"refs/heads/main","pushedAt":"2024-07-24T21:41:09.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"},"commit":{"message":"[IntraNodeComm] fix an issue where input check fails when running all-reduce on sub groups\n\nghstack-source-id: 218c718ade2473fbf6e93baaf8ae8a265d00386c\nPull Request resolved: https://github.com/pytorch/pytorch/pull/130492","shortMessageHtmlLink":"[IntraNodeComm] fix an issue where input check fails when running all…"}},{"before":null,"after":"ef3fd5bf73e65e8760a519efd7ddf88fb76e1c3b","ref":"refs/heads/py_work","pushedAt":"2024-07-24T21:41:09.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"},"commit":{"message":"SymmetricMemory-based, low contention intra-node all-gather and reduce-scatter\n\nghstack-source-id: d46ebfd04de86fda844c553869cc23cbd2751acb\nPull Request resolved: https://github.com/pytorch/pytorch/pull/130583","shortMessageHtmlLink":"SymmetricMemory-based, low contention intra-node all-gather and reduc…"}},{"before":null,"after":"f22c17fce640a5f5dbf8c78135155194570bf059","ref":"refs/heads/reduce_scatter_2","pushedAt":"2024-07-24T21:41:09.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"},"commit":{"message":"all-to-all + reduction","shortMessageHtmlLink":"all-to-all + reduction"}},{"before":null,"after":"f7edf056f8f2d0a2ded177755bf8a306a89c8041","ref":"refs/heads/coll_selection","pushedAt":"2024-07-24T21:41:09.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"},"commit":{"message":"foo","shortMessageHtmlLink":"foo"}},{"before":"76bd8b02b427c29654872ddf0ee5428cb2ba250e","after":"651e42da678bae10e35190fa6b93b837137588c3","ref":"refs/heads/export-D55960644","pushedAt":"2024-04-10T19:12:17.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"},"commit":{"message":"[multi_tensor_apply] temporarily disable the small buffer optimization (#123712)\n\nSummary:\n\nIntroduced in https://github.com/pytorch/pytorch/pull/119764\n\nThe small buffer optimization caused a regression described in D55721766. The cause is likely that older versions of nvcc aren't able to generate efficient code for the logic (happens in fbcode but not reproducible with nvcc 12.2). Disabling it for now.\n\nWill follow up to find a way for the small buffer optimization to play well with the compilers in question, or remove the optimization all together.\n\nTest Plan: Existing tests in CI\n\nDifferential Revision: D55960644","shortMessageHtmlLink":"[multi_tensor_apply] temporarily disable the small buffer optimization ("}},{"before":null,"after":"76bd8b02b427c29654872ddf0ee5428cb2ba250e","ref":"refs/heads/export-D55960644","pushedAt":"2024-04-10T09:19:15.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"}},{"before":"5361c369014ae2015b8fa532ce9ebfcd9d8e3c04","after":"758bf514d33f3b1f0347d7b44175cb0aa8bbaf74","ref":"refs/heads/export-D54355011","pushedAt":"2024-02-29T10:10:55.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"},"commit":{"message":"[functional collecitve] don't import torchdynamo when running torchdeploy\n\nSummary: Importing torchdynamo in `functional_collective_impl.py` seems to break loading of torchdeploy models.\n\nTest Plan: CI\n\nDifferential Revision: D54355011","shortMessageHtmlLink":"[functional collecitve] don't import torchdynamo when running torchde…"}},{"before":"eb9d56a14111d2b1e0e407360a1d097c3ee967fa","after":"5361c369014ae2015b8fa532ce9ebfcd9d8e3c04","ref":"refs/heads/export-D54355011","pushedAt":"2024-02-29T10:10:21.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"}},{"before":"1ec597b8bc5009755490d49aa0c99ff774af86c5","after":"eb9d56a14111d2b1e0e407360a1d097c3ee967fa","ref":"refs/heads/export-D54355011","pushedAt":"2024-02-29T09:48:17.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"}},{"before":null,"after":"1ec597b8bc5009755490d49aa0c99ff774af86c5","ref":"refs/heads/export-D54355011","pushedAt":"2024-02-29T09:36:56.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"}},{"before":"61df4815bf60562804f24e361e9b631298dbccc7","after":"2d3300fa0ad7c827f0e4aa8099d83fcbb9a11606","ref":"refs/heads/export-D44843942","pushedAt":"2023-04-10T22:10:40.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"},"commit":{"message":"[spmd compile api] run gm_transforms before running the first iteration (#98788)\n\nSummary:\nPull Request resolved: https://github.com/pytorch/pytorch/pull/98788\n\nThe non-transformed graph module contains functionalized optimizer which, in a memory constraint environment, needs to be defunctionalized (via fx transformation or lowering to Inductor) before running the first iteration. Otherwise OOM may occur.\n\nTest Plan: Manually tested.\n\nReviewed By: mrshenli\n\nDifferential Revision: D44843942\n\nfbshipit-source-id: ccd91f39c844a7e38bbdbb7b81fd4a2e30a31ad8","shortMessageHtmlLink":"[spmd compile api] run gm_transforms before running the first iterati…"}},{"before":"6ae7a9b15759c2411ce83ff755e7e797754121f5","after":"a92f161a4e0d3443c743d8ac5909475c482c963b","ref":"refs/heads/export-D44845043","pushedAt":"2023-04-10T22:02:28.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"},"commit":{"message":"[spmd compile api] use fake tensor for DTensor propagation (#98789)\n\nSummary:\nPull Request resolved: https://github.com/pytorch/pytorch/pull/98789\n\nWhen using real tensors for DTensor propagation, functionalized _fuse_adam causes a memory spike of size(params + optim_state), which causes OOM on memory constrained environments.\n\nTest Plan: Tested manually.\n\nDifferential Revision: D44845043\n\nfbshipit-source-id: 5476ff2118fda2ed727c0812bb6cecba6b3556ec","shortMessageHtmlLink":"[spmd compile api] use fake tensor for DTensor propagation (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1661418272\" data-permission-text=\"Title is private\" data-url=\"https://github.com/pytorch/pytorch/issues/98789\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/pytorch/pytorch/pull/98789/hovercard\" href=\"https://github.com/pytorch/pytorch/pull/98789\">pytorch#9…</a>"}},{"before":"7008d6a43c29bffecce6b21e519dc7cf6d5f0067","after":"61df4815bf60562804f24e361e9b631298dbccc7","ref":"refs/heads/export-D44843942","pushedAt":"2023-04-10T22:02:06.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"}},{"before":null,"after":"6ae7a9b15759c2411ce83ff755e7e797754121f5","ref":"refs/heads/export-D44845043","pushedAt":"2023-04-10T21:47:12.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"}},{"before":null,"after":"7008d6a43c29bffecce6b21e519dc7cf6d5f0067","ref":"refs/heads/export-D44843942","pushedAt":"2023-04-10T21:45:40.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"}},{"before":null,"after":"324dbb4da33b1b45f64c092c9d7a02dc64a6358f","ref":"refs/heads/export-D44615037","pushedAt":"2023-04-02T21:53:26.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"yifuwang","name":"Yifu Wang","path":"/yifuwang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/4156752?s=80&v=4"},"commit":{"message":"[spmd expansion] support torch.ops.aten.sym_numel (#98177)\n\nSummary:\nPull Request resolved: https://github.com/pytorch/pytorch/pull/98177\n\nThe current logic assumes non-overload ops takes two arguments however torch.ops.aten.sym_numel takes one.\nghstack-source-id: 184813957\n\nTest Plan: Manual testing.\n\nReviewed By: mrshenli\n\nDifferential Revision: D44615037\n\nfbshipit-source-id: 6c9fe56e6face4887a99b60f3bffc8df9ec038a7","shortMessageHtmlLink":"[spmd expansion] support torch.ops.aten.sym_numel (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1651105476\" data-permission-text=\"Title is private\" data-url=\"https://github.com/pytorch/pytorch/issues/98177\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/pytorch/pytorch/pull/98177/hovercard\" href=\"https://github.com/pytorch/pytorch/pull/98177\">pytorch#98177</a>)"}}],"hasNextPage":false,"hasPreviousPage":false,"activityType":"all","actor":null,"timePeriod":"all","sort":"DESC","perPage":30,"cursor":"Y3Vyc29yOnYyOpK7MjAyNC0wNy0yNFQyMTo0MTowOS4wMDAwMDBazwAAAASIZmf9","startCursor":"Y3Vyc29yOnYyOpK7MjAyNC0wNy0yNFQyMTo0MTowOS4wMDAwMDBazwAAAASIZmf9","endCursor":"Y3Vyc29yOnYyOpK7MjAyMy0wNC0wMlQyMTo1MzoyNi4wMDAwMDBazwAAAAMQkDoi"}},"title":"Activity · yifuwang/pytorch"}