{"payload":{"feedbackUrl":"https://github.com/orgs/community/discussions/53140","repo":{"id":33015583,"defaultBranch":"master","name":"keras","ownerLogin":"keras-team","currentUserCanPush":false,"isFork":false,"isEmpty":false,"createdAt":"2015-03-28T00:35:42.000Z","ownerAvatar":"https://avatars.githubusercontent.com/u/34455048?v=4","public":true,"private":false,"isOrgOwned":true},"refInfo":{"name":"","listCacheKey":"v0:1717592286.0","currentOid":""},"activityList":{"items":[{"before":"77ef1792201cd30b52146c71dd0380786512ac84","after":"a4e85547b2f85283aab53f35f84c77d026ab2102","ref":"refs/heads/master","pushedAt":"2024-06-13T01:00:48.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Fix bugs with `Mean`, `Accuracy` and `BinaryAccuracy` metrics. (#19847)\n\n- `reduce_to_samplewise_values` would not reduce `sample_weights` correctly because the number of dimensions of `values` was checked.\r\n- `reduce_to_samplewise_values` needs to explicitely broadcast `sample_weights`. Before, it was implicitly broadcast in the multiplication with `values`. However, the explicit broadcast is needed for the computation of `num_samples` for the averaging to be correct. This causes a bug when `sample_weights` is of rank 2 or more and a broadcast happens when doing the multiplication. This logic existed in `tf_keras`: https://github.com/keras-team/tf-keras/blob/master/tf_keras/metrics/base_metric.py#L508\r\n- `Accuracy` and `BinaryAccuracy` were doing a mean reduction too early, before multiplying by `sample_weights`. This matters when the rank of `sample_weights` is the same as `y_true` and `y_pred`.","shortMessageHtmlLink":"Fix bugs with Mean, Accuracy and BinaryAccuracy metrics. (#19847)"}},{"before":"4551644005364fbeb4757344f6f909efee8d9de1","after":"77ef1792201cd30b52146c71dd0380786512ac84","ref":"refs/heads/master","pushedAt":"2024-06-12T22:13:02.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Fix inconsistent behavior of `losses.sparse_categorical_crossentropy`… (#19838)\n\n* Fix inconsistent behavior of `losses.sparse_categorical_crossentropy` with and without `ignore_class`\r\n\r\n* Test\r\n\r\n* chore(format)\r\n\r\n* Fix tests in `losses`","shortMessageHtmlLink":"Fix inconsistent behavior of losses.sparse_categorical_crossentropy… ("}},{"before":"46df34110339ae3dfbc896f446ace9f33da64838","after":"4551644005364fbeb4757344f6f909efee8d9de1","ref":"refs/heads/master","pushedAt":"2024-06-12T15:59:19.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Fix the compatibility issues of `Orthogonal` and `GRU` (#19844)\n\n* Add legacy `Orthogonal` class name\r\n\r\n* Add legacy `implementation` arg to `GRU`","shortMessageHtmlLink":"Fix the compatibility issues of Orthogonal and GRU (#19844)"}},{"before":"9ad3ca0ba5a77b4616b223481fb43898e31d272e","after":"46df34110339ae3dfbc896f446ace9f33da64838","ref":"refs/heads/master","pushedAt":"2024-06-12T15:58:16.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Add `training` argument to `Model.compute_loss()`. (#19840)\n\nThis allows models to perform different computations during training and evaluation. For instance, some expensive to compute metrics can be skipped during training and only computed during evaluation.\r\n\r\nNote that backwards compatibility with overrides that do not have the `training` argument is maintained.","shortMessageHtmlLink":"Add training argument to Model.compute_loss(). (#19840)"}},{"before":"d1b404df9fb518c24dd30d0d903692c27ff272db","after":"9ad3ca0ba5a77b4616b223481fb43898e31d272e","ref":"refs/heads/master","pushedAt":"2024-06-12T15:55:22.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Adjust code example","shortMessageHtmlLink":"Adjust code example"}},{"before":"101119fd2f0e2ee29364e31c212a3aa2fd66c91a","after":"d1b404df9fb518c24dd30d0d903692c27ff272db","ref":"refs/heads/master","pushedAt":"2024-06-12T15:54:09.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"corrected the example code in unit_normalization.py (#19845)\n\nAdded missing closing bracket and exact output value in example code after replicating the code.","shortMessageHtmlLink":"corrected the example code in unit_normalization.py (#19845)"}},{"before":"1c60668f6bdd05dab619806e7b2dc25d3ed4ccbf","after":"101119fd2f0e2ee29364e31c212a3aa2fd66c91a","ref":"refs/heads/master","pushedAt":"2024-06-12T15:49:20.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"doc: `ops.slice` (#19843)","shortMessageHtmlLink":"doc: ops.slice (#19843)"}},{"before":"a8aac9720d166eb10c37bba292c21896ae9ca36d","after":"1c60668f6bdd05dab619806e7b2dc25d3ed4ccbf","ref":"refs/heads/master","pushedAt":"2024-06-11T23:18:30.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Fix code style","shortMessageHtmlLink":"Fix code style"}},{"before":"224de28928f46c017f00661513590e80bee47787","after":"a8aac9720d166eb10c37bba292c21896ae9ca36d","ref":"refs/heads/master","pushedAt":"2024-06-11T23:16:50.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Make batch norm mask shape error more descriptive (#19829)\n\n* Made batch norm mask shape error more descriptive\r\n\r\n* Added shape info in mask error message to help with degugging","shortMessageHtmlLink":"Make batch norm mask shape error more descriptive (#19829)"}},{"before":"26abe697a8802de40cb2761fc98b843fe1b2d5f6","after":"224de28928f46c017f00661513590e80bee47787","ref":"refs/heads/master","pushedAt":"2024-06-11T22:55:13.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Use `absl.testing.parameterized` for `tree_test.py`. (#19842)\n\nFor consistency, use `absl.testing.parameterized` instead of `parameterized` for `tree_test.py` since that is used for all other tests.\r\n\r\nIt's one less dependency. It also says `optree` or `dmtree` in each test name.","shortMessageHtmlLink":"Use absl.testing.parameterized for tree_test.py. (#19842)"}},{"before":"2c48f2e9aa29f81d2d8b23c380a85ca689ae9e30","after":"26abe697a8802de40cb2761fc98b843fe1b2d5f6","ref":"refs/heads/master","pushedAt":"2024-06-11T14:42:49.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Add `ops.switch` (#19834)\n\n* Add `ops.switch`\r\n\r\n* Update tests\r\n\r\n* Fix out-of-bound issue\r\n\r\n* Revert `torch.cond`","shortMessageHtmlLink":"Add ops.switch (#19834)"}},{"before":"47bedf1418135575a629990e7f1b7aeabc93cb69","after":"2c48f2e9aa29f81d2d8b23c380a85ca689ae9e30","ref":"refs/heads/master","pushedAt":"2024-06-11T14:34:23.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Fix `get_file` when the HTTP response has no `Content-Length` header (#19833)","shortMessageHtmlLink":"Fix get_file when the HTTP response has no Content-Length header (#…"}},{"before":"24060372be8b37fa35dd9580b375278ea05af924","after":"47bedf1418135575a629990e7f1b7aeabc93cb69","ref":"refs/heads/master","pushedAt":"2024-06-11T02:33:53.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"`Metric.variables` is now recursive. (#19830)\n\nThis allows it to surface variables from metrics nested at any depth.\r\n\r\nPreviously, metrics within metrics within metrics would not have their variables tracked in JAX, causing them to not be updated.","shortMessageHtmlLink":"Metric.variables is now recursive. (#19830)"}},{"before":"efaaf85e19113400f23462cbafcef433cd95ad9c","after":"24060372be8b37fa35dd9580b375278ea05af924","ref":"refs/heads/master","pushedAt":"2024-06-11T02:33:27.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Fix `keras.ops.repeat` cannot return an expected shape when `x` is a … (#19826)\n\n* Fix `keras.ops.repeat` cannot return an expected shape when `x` is a `KerasTensor` and the `axis` is `None`\r\n\r\n* Test dynamic is still dynamic after repetition\r\n\r\n* Improve error messages","shortMessageHtmlLink":"Fix keras.ops.repeat cannot return an expected shape when x is a … ("}},{"before":"857a7acfc2496963dec777e13da8a6486ac50c38","after":"efaaf85e19113400f23462cbafcef433cd95ad9c","ref":"refs/heads/master","pushedAt":"2024-06-10T17:20:35.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"fixing dmtree.is_nested() and parameterized tree test (#19822)","shortMessageHtmlLink":"fixing dmtree.is_nested() and parameterized tree test (#19822)"}},{"before":"da72e2241ce4488efb036128d87d93e303f0a8f7","after":"857a7acfc2496963dec777e13da8a6486ac50c38","ref":"refs/heads/master","pushedAt":"2024-06-10T17:19:19.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Fix TypeError in `Lambda.from_config` (#19827)","shortMessageHtmlLink":"Fix TypeError in Lambda.from_config (#19827)"}},{"before":"2305fada8889e86463493bb4893b13ee8a8f0573","after":"da72e2241ce4488efb036128d87d93e303f0a8f7","ref":"refs/heads/master","pushedAt":"2024-06-10T17:18:44.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Fix CI","shortMessageHtmlLink":"Fix CI"}},{"before":"314fa3b9e38a73bd90872f22532efe44c3a9553d","after":"2305fada8889e86463493bb4893b13ee8a8f0573","ref":"refs/heads/master","pushedAt":"2024-06-09T17:34:58.000Z","pushType":"push","commitsCount":2,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Merge branch 'master' of github.com:keras-team/keras","shortMessageHtmlLink":"Merge branch 'master' of github.com:keras-team/keras"}},{"before":"13ac85cb2af7a9500d0ed6316265beee440ca4af","after":"314fa3b9e38a73bd90872f22532efe44c3a9553d","ref":"refs/heads/master","pushedAt":"2024-06-07T04:23:30.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Faster in_top_k implementation for Jax backend (#19814)\n\n* Faster in_top_k implementation.\r\n\r\n* Fix bug in rank computation.","shortMessageHtmlLink":"Faster in_top_k implementation for Jax backend (#19814)"}},{"before":"a2df0f9ac595639aa2d3a0359122b030d934389e","after":"13ac85cb2af7a9500d0ed6316265beee440ca4af","ref":"refs/heads/master","pushedAt":"2024-06-06T23:33:15.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Propagate kwargs through `keras.ops.isclose` (#19782)\n\n* propagate kwargs through isclose\r\nthis allows passing atol and rtol\r\n\r\n* switch isclose **kwargs to explicit kwargs\r\n\r\n* reduce line lengths\r\n\r\n* fix ops.isclose signature\r\n\r\n* fix ops.IsClose compute_output_spec signature\r\n\r\n* implement isclose rtol atol equal_nan args for all backends\r\n\r\n* shorten line lengths again\r\n\r\n* revert using tf.experimental.numpy.isclose\r\ntensorflow version now uses code inspired from tf.experimental.numpy.isclose\r\n\r\n* fix lint\r\n\r\n* add docs for new parameters","shortMessageHtmlLink":"Propagate kwargs through keras.ops.isclose (#19782)"}},{"before":"4b4cf604445810562272655163e643f601e16441","after":"a2df0f9ac595639aa2d3a0359122b030d934389e","ref":"refs/heads/master","pushedAt":"2024-06-05T16:27:35.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Fix `LayerNormalization.get_config` (#19807)","shortMessageHtmlLink":"Fix LayerNormalization.get_config (#19807)"}},{"before":"bf607fffbac61ab3bf509094c9b7a6a17853458c","after":"4b4cf604445810562272655163e643f601e16441","ref":"refs/heads/master","pushedAt":"2024-06-05T15:58:47.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"perf: remove a redundant call to `Functional.__init__` (#19805)","shortMessageHtmlLink":"perf: remove a redundant call to Functional.__init__ (#19805)"}},{"before":"ab8ddfce7757f20266598faad4c44b4a33655ad8","after":null,"ref":"refs/heads/sachinprasadhs-patch-1","pushedAt":"2024-06-05T12:58:06.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"sachinprasadhs","name":"Sachin Prasad","path":"/sachinprasadhs","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/73069040?s=80&v=4"}},{"before":"33756c12ea19a4a61cfeb1e284b29a109bce522e","after":"bf607fffbac61ab3bf509094c9b7a6a17853458c","ref":"refs/heads/master","pushedAt":"2024-06-05T03:23:06.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Enable `Sequential` and `Functional` model to be used as a submodel (#19799)\n\n* Enable `Sequential` model to be submodel\r\n\r\n* Improve test coverage\r\n\r\n* Enable `Functional` model to be used as a submodel\r\n\r\n* Address comments\r\n\r\n* Update tests","shortMessageHtmlLink":"Enable Sequential and Functional model to be used as a submodel (#…"}},{"before":"41aff3dbf0e8bcb43f047062f32c4bc66b976e2d","after":"33756c12ea19a4a61cfeb1e284b29a109bce522e","ref":"refs/heads/master","pushedAt":"2024-06-04T16:58:28.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Fix `DynamicBackend` (#19800)\n\n* Fix `DynamicBackend`\r\n\r\n* Fix msg\r\n\r\n* Fix bug\r\n\r\n* Fix comments\r\n\r\n* Patch\r\n\r\n* Fix formatting","shortMessageHtmlLink":"Fix DynamicBackend (#19800)"}},{"before":"1eec3b8ed33c66f8877b7304785272e7dfe7ddc1","after":"41aff3dbf0e8bcb43f047062f32c4bc66b976e2d","ref":"refs/heads/master","pushedAt":"2024-06-04T16:56:06.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Make example in documentation more obvious (#19802)\n\nBy adding a third column in `dims` in the example, we avoid the misconception that the tuple could mean the exchange of the given pair, i.e. `(2, 3)` meaning to exchange the second and third column.","shortMessageHtmlLink":"Make example in documentation more obvious (#19802)"}},{"before":"c94663711d738b50af324214d89f895e046a2b66","after":"1eec3b8ed33c66f8877b7304785272e7dfe7ddc1","ref":"refs/heads/master","pushedAt":"2024-06-04T16:49:50.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Update auto-assignment.js (#19801)\n\nUpdate assignee name.","shortMessageHtmlLink":"Update auto-assignment.js (#19801)"}},{"before":null,"after":"ab8ddfce7757f20266598faad4c44b4a33655ad8","ref":"refs/heads/sachinprasadhs-patch-1","pushedAt":"2024-06-04T09:21:53.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"sachinprasadhs","name":"Sachin Prasad","path":"/sachinprasadhs","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/73069040?s=80&v=4"},"commit":{"message":"Update auto-assignment.js\n\nUpdate assignee name.","shortMessageHtmlLink":"Update auto-assignment.js"}},{"before":"26f438edd5516bf02504a5a0e13a20ceec69cb29","after":"c94663711d738b50af324214d89f895e046a2b66","ref":"refs/heads/master","pushedAt":"2024-06-03T18:46:08.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Fix `CategoryEncoding` generates output with 1 additional dimension in all backends except NumPy (#19724) (#19790)\n\n* Fix `CategoryEncoding` generates output with 1 additional dimension in all backends except NumPy (#19724)\r\n\r\n* Test one-hot with 1 extra dimension\r\n\r\n* chore(format)\r\n\r\n* chore(format)","shortMessageHtmlLink":"Fix CategoryEncoding generates output with 1 additional dimension i…"}},{"before":"8481973d90606dbff677737a9572532f84c22277","after":"26f438edd5516bf02504a5a0e13a20ceec69cb29","ref":"refs/heads/master","pushedAt":"2024-06-02T19:35:25.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"fchollet","name":"François Chollet","path":"/fchollet","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/710255?s=80&v=4"},"commit":{"message":"Reformat code","shortMessageHtmlLink":"Reformat code"}}],"hasNextPage":true,"hasPreviousPage":false,"activityType":"all","actor":null,"timePeriod":"all","sort":"DESC","perPage":30,"cursor":"djE6ks8AAAAEY-BtcwA","startCursor":null,"endCursor":null}},"title":"Activity · keras-team/keras"}