{"payload":{"feedbackUrl":"https://github.com/orgs/community/discussions/53140","repo":{"id":503218726,"defaultBranch":"main","name":"detectron2","ownerLogin":"darrenchang","currentUserCanPush":false,"isFork":true,"isEmpty":false,"createdAt":"2022-06-14T05:19:30.000Z","ownerAvatar":"https://avatars.githubusercontent.com/u/10385811?v=4","public":true,"private":false,"isOrgOwned":false},"refInfo":{"name":"","listCacheKey":"v0:1723777464.0","currentOid":""},"activityList":{"items":[{"before":null,"after":"5d7eae12be4bd2cb86cc883da32511e57b19b10a","ref":"refs/heads/torch_load_wegihts_only_flase","pushedAt":"2024-08-16T03:04:24.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"},"commit":{"message":"Add weights_only to torch.loads","shortMessageHtmlLink":"Add weights_only to torch.loads"}},{"before":"133f83bed41c1e09710b162bc1fd149917e9c8c8","after":"d09e11c915c054062b5bf625e749ea35b958f258","ref":"refs/heads/compatible_with_torch2trt","pushedAt":"2024-08-16T02:57:43.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"},"commit":{"message":"Get rid of some Tensor.to methods and add weigths_only=False","shortMessageHtmlLink":"Get rid of some Tensor.to methods and add weigths_only=False"}},{"before":"34c78ef5c60ada1a39cabe97e42ba18c60507106","after":"133f83bed41c1e09710b162bc1fd149917e9c8c8","ref":"refs/heads/compatible_with_torch2trt","pushedAt":"2024-08-16T02:16:26.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"},"commit":{"message":"Change all .to(dtype) to cast directly","shortMessageHtmlLink":"Change all .to(dtype) to cast directly"}},{"before":"ce721422d5167f81b58a1920092a716a7aabfbc7","after":"34c78ef5c60ada1a39cabe97e42ba18c60507106","ref":"refs/heads/compatible_with_torch2trt","pushedAt":"2024-08-15T13:17:10.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"},"commit":{"message":"Cast to uint8 inplace","shortMessageHtmlLink":"Cast to uint8 inplace"}},{"before":"08a95b744ff235f2776a2da841f08e10f2c09fbc","after":"ce721422d5167f81b58a1920092a716a7aabfbc7","ref":"refs/heads/compatible_with_torch2trt","pushedAt":"2024-08-15T12:43:59.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"},"commit":{"message":"Fix: change int64 to long\n\n- int64 is not a method","shortMessageHtmlLink":"Fix: change int64 to long"}},{"before":null,"after":"08a95b744ff235f2776a2da841f08e10f2c09fbc","ref":"refs/heads/compatible_with_torch2trt","pushedAt":"2024-08-15T08:58:56.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"},"commit":{"message":"Convert all layered operations to torch.<type>","shortMessageHtmlLink":"Convert all layered operations to torch.&lt;type&gt;"}},{"before":"a885bb734b9927d35ffc75ba6b2047cf71c955b7","after":null,"ref":"refs/heads/ppwwyyxx-patch-3","pushedAt":"2024-08-15T08:43:26.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"}},{"before":"a03262181a7019c4775c87214f5b9256a262d926","after":null,"ref":"refs/heads/yuxinwu","pushedAt":"2024-08-15T08:43:20.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"}},{"before":"6dfe0e56b5803dce2f4033b7aac9e1ed21f2737e","after":null,"ref":"refs/heads/yuxin","pushedAt":"2024-08-15T08:43:18.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"}},{"before":"f89e39ab8220a2981157f5060b5c73058c163516","after":null,"ref":"refs/heads/tmp","pushedAt":"2024-08-15T08:43:18.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"}},{"before":"1f7fa2c9a0cc85f6b938303dbbd794087e314552","after":null,"ref":"refs/heads/test-ci","pushedAt":"2024-08-15T08:43:16.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"}},{"before":"b391be5db74afca774ec2fd0021eda8a2dc8c7c5","after":null,"ref":"refs/heads/ppwwyyxx-patch-4","pushedAt":"2024-08-15T08:43:13.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"}},{"before":"c1e082d62fd0315077f9ac3ca76049a927a0c7aa","after":null,"ref":"refs/heads/ppwwyyxx-patch-2","pushedAt":"2024-08-15T08:43:12.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"}},{"before":"214a48a7a88470afe6ef34449ef3be122d2fc72e","after":null,"ref":"refs/heads/ppwwyyxx-patch-1","pushedAt":"2024-08-15T08:43:10.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"}},{"before":"afe9eb920646102f7e6bf0cd2115841cea2aca13","after":"bcfd464d0c810f0442d91a349c0f6df945467143","ref":"refs/heads/main","pushedAt":"2024-08-15T08:42:14.000Z","pushType":"push","commitsCount":21,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"},"commit":{"message":"upgrade pyre version in `fbcode/vision` - batch 1\n\nDifferential Revision: D60992191\n\nfbshipit-source-id: f826042c9d5b4f9b72b142fcef13f5772e3b9a8d","shortMessageHtmlLink":"upgrade pyre version in <code>fbcode/vision</code> - batch 1"}},{"before":"ce79b37711dcc29e99a4f2a3b40035a59bb992f3","after":"c3a786acde2a51f8c35868c256d8787c2c74eeb9","ref":"refs/heads/soft-nms","pushedAt":"2024-04-03T09:02:54.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"},"commit":{"message":"Added default value for inference def\n\nMake it easier with TensorRT","shortMessageHtmlLink":"Added default value for inference def"}},{"before":"f8b38b69ed8e1acba73443520ec435456db3cae1","after":"ce79b37711dcc29e99a4f2a3b40035a59bb992f3","ref":"refs/heads/soft-nms","pushedAt":"2024-04-03T07:51:20.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"},"commit":{"message":"Revert \"Enable soft nms by default\"\n\nThis reverts commit f8b38b69ed8e1acba73443520ec435456db3cae1.","shortMessageHtmlLink":"Revert \"Enable soft nms by default\""}},{"before":"f8b38b69ed8e1acba73443520ec435456db3cae1","after":null,"ref":"refs/heads/soft-nms-6","pushedAt":"2024-04-03T07:49:32.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"}},{"before":null,"after":"f8b38b69ed8e1acba73443520ec435456db3cae1","ref":"refs/heads/soft-nms","pushedAt":"2024-04-03T07:49:19.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"},"commit":{"message":"Enable soft nms by default","shortMessageHtmlLink":"Enable soft nms by default"}},{"before":"fd8925e64f78aef4ea032c5abbe66baf003f4e36","after":null,"ref":"refs/heads/soft_nms_5","pushedAt":"2024-04-03T07:48:51.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"}},{"before":"d4a5f28e01b2babbaba9f90198fb95f5c661cccd","after":null,"ref":"refs/heads/soft_nms_3","pushedAt":"2024-04-03T07:48:47.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"}},{"before":"4ed6134a96ecd765f2c52abb991c3b5fe82e8657","after":null,"ref":"refs/heads/soft_nms_2","pushedAt":"2024-04-03T07:48:43.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"}},{"before":"eec28810cc377b0b0914dbbd547dacad00cd6e16","after":null,"ref":"refs/heads/soft-nms","pushedAt":"2024-04-03T07:48:24.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"}},{"before":"f2989deabed2a27bc243a1be05ded92fe2bdfd5e","after":"f8b38b69ed8e1acba73443520ec435456db3cae1","ref":"refs/heads/soft-nms-6","pushedAt":"2024-04-02T18:57:28.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"},"commit":{"message":"Enable soft nms by default","shortMessageHtmlLink":"Enable soft nms by default"}},{"before":"d4a5f28e01b2babbaba9f90198fb95f5c661cccd","after":"f2989deabed2a27bc243a1be05ded92fe2bdfd5e","ref":"refs/heads/soft-nms-6","pushedAt":"2024-04-02T17:07:08.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"},"commit":{"message":"Implemented soft nms - by alekseynp\n\nhttps://github.com/facebookresearch/detectron2/pull/1183","shortMessageHtmlLink":"Implemented soft nms - by alekseynp"}},{"before":null,"after":"d4a5f28e01b2babbaba9f90198fb95f5c661cccd","ref":"refs/heads/soft-nms-6","pushedAt":"2024-04-02T14:07:01.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"},"commit":{"message":"Initialize FrozenBatchNorm2D with num_batches_tracked\n\nSummary: When using D2Go to freeze layers with `FrozenBatchNorm2D`, the function `FrozenBatchNorm2D.convert_frozen_batchnorm` is utilized to replace the original `BatchNorm`. However, during conversion, the attribute `num_batches_tracked` is not copied, which can lead to issues when the model is unfrozen during TorchScript tracking due to conditionals in `FrozenBatchNorm2D.forward`. To avoid these problems, it is necessary to include a copy of the attribute `num_batches_tracked` in `FrozenBatchNorm2D.convert_frozen_batchnorm`, as presented in this diff.\n\nReviewed By: FarooqZuberi\n\nDifferential Revision: D45593607\n\nfbshipit-source-id: 3a119e2741b80c24cdb24a637491c6c64e5dfeb0","shortMessageHtmlLink":"Initialize FrozenBatchNorm2D with num_batches_tracked"}},{"before":null,"after":"fd8925e64f78aef4ea032c5abbe66baf003f4e36","ref":"refs/heads/soft_nms_5","pushedAt":"2024-04-02T13:10:07.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"},"commit":{"message":"Bugfixes.","shortMessageHtmlLink":"Bugfixes."}},{"before":null,"after":"d4a5f28e01b2babbaba9f90198fb95f5c661cccd","ref":"refs/heads/soft_nms_3","pushedAt":"2024-04-02T12:33:09.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"},"commit":{"message":"Initialize FrozenBatchNorm2D with num_batches_tracked\n\nSummary: When using D2Go to freeze layers with `FrozenBatchNorm2D`, the function `FrozenBatchNorm2D.convert_frozen_batchnorm` is utilized to replace the original `BatchNorm`. However, during conversion, the attribute `num_batches_tracked` is not copied, which can lead to issues when the model is unfrozen during TorchScript tracking due to conditionals in `FrozenBatchNorm2D.forward`. To avoid these problems, it is necessary to include a copy of the attribute `num_batches_tracked` in `FrozenBatchNorm2D.convert_frozen_batchnorm`, as presented in this diff.\n\nReviewed By: FarooqZuberi\n\nDifferential Revision: D45593607\n\nfbshipit-source-id: 3a119e2741b80c24cdb24a637491c6c64e5dfeb0","shortMessageHtmlLink":"Initialize FrozenBatchNorm2D with num_batches_tracked"}},{"before":"bd0254964eab59b8b0a1402a7e55c5bb9ce2c7a7","after":"4ed6134a96ecd765f2c52abb991c3b5fe82e8657","ref":"refs/heads/soft_nms_2","pushedAt":"2024-04-02T11:28:26.000Z","pushType":"push","commitsCount":186,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"},"commit":{"message":"Merge commit 'd4a5f28e01b2babbaba9f90198fb95f5c661cccd' into soft_nms_2","shortMessageHtmlLink":"Merge commit 'd4a5f28e01b2babbaba9f90198fb95f5c661cccd' into soft_nms_2"}},{"before":"64f8b12182da0ca2cbdc86c394960e8eba150bde","after":"afe9eb920646102f7e6bf0cd2115841cea2aca13","ref":"refs/heads/main","pushedAt":"2024-04-02T11:23:48.000Z","pushType":"push","commitsCount":134,"pusher":{"login":"darrenchang","name":"Darren Chang","path":"/darrenchang","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/10385811?s=80&v=4"},"commit":{"message":"add DATALOADER.REPEAT_SQRT\n\nSummary:\nPull Request resolved: https://github.com/facebookresearch/detectron2/pull/5245\n\nX-link: https://github.com/fairinternal/detectron2/pull/602\n\nFor sampler **RepeatFactorTrainingSampler**, current per-category weight is computed as **1/sqrt(frequency)**.\n\nThis works fine on LVIS but is not sufficient in highly imbalanced data we have for person segmentation.\n\nThus we add an argument **DATALOADER.REPEAT_SQRT**. If false, we compute per-category weight as **1/frequency**.\n\nThis change is entirely back-compatible.\n\nReviewed By: wat3rBro\n\nDifferential Revision:\nD55355021\n\nPrivacy Context Container: L1165023\n\nfbshipit-source-id: 6bca2eecc3b9a7b4693a288c5779627254cd5ec5","shortMessageHtmlLink":"add DATALOADER.REPEAT_SQRT"}}],"hasNextPage":true,"hasPreviousPage":false,"activityType":"all","actor":null,"timePeriod":"all","sort":"DESC","perPage":30,"cursor":"Y3Vyc29yOnYyOpK7MjAyNC0wOC0xNlQwMzowNDoyNC4wMDAwMDBazwAAAASbv71A","startCursor":"Y3Vyc29yOnYyOpK7MjAyNC0wOC0xNlQwMzowNDoyNC4wMDAwMDBazwAAAASbv71A","endCursor":"Y3Vyc29yOnYyOpK7MjAyNC0wNC0wMlQxMToyMzo0OC4wMDAwMDBazwAAAAQlcXQ8"}},"title":"Activity · darrenchang/detectron2"}