{"payload":{"feedbackUrl":"https://github.com/orgs/community/discussions/53140","repo":{"id":416162048,"defaultBranch":"main","name":"dfdx","ownerLogin":"coreylowman","currentUserCanPush":false,"isFork":false,"isEmpty":false,"createdAt":"2021-10-12T02:58:54.000Z","ownerAvatar":"https://avatars.githubusercontent.com/u/7787278?v=4","public":true,"private":false,"isOrgOwned":false},"refInfo":{"name":"","listCacheKey":"v0:1698329953.0","currentOid":""},"activityList":{"items":[{"before":"e04dd4fea568bcc6ee5e3f536a57488cdba0d772","after":"4722a99d303f347d6088d95867d007c75ca6dd78","ref":"refs/heads/main","pushedAt":"2024-01-25T19:46:50.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"feat(wgpu): add to_dtype kernel (#906)\n\n* feat(wgpu): add to_dtype kernel\r\n\r\n* fix: add WebGPUNativeType\r\n\r\n* style: clippy fix\r\n\r\n---------\r\n\r\nCo-authored-by: Corey Lowman ","shortMessageHtmlLink":"feat(wgpu): add to_dtype kernel (#906)"}},{"before":"a4a271f8bb958b45a015c203f5c89b024eecee59","after":"e04dd4fea568bcc6ee5e3f536a57488cdba0d772","ref":"refs/heads/main","pushedAt":"2024-01-25T19:46:30.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"make non-const conv1d available to stable rust (#911)\n\n* make device capable of Conv1DKernel\r\n\r\n* make non-const conv1d available to stable rust\r\n\r\n---------\r\n\r\nCo-authored-by: Corey Lowman ","shortMessageHtmlLink":"make non-const conv1d available to stable rust (#911)"}},{"before":"f4f7165da521966aea5c655d32a0a7c5637b7cff","after":"a4a271f8bb958b45a015c203f5c89b024eecee59","ref":"refs/heads/main","pushedAt":"2024-01-25T19:46:11.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"add missing ` to optimizer section in crate-level documentation (#909)","shortMessageHtmlLink":"add missing ` to optimizer section in crate-level documentation (#909)"}},{"before":"630514f402f4340421aa4a3746ec99f29120923a","after":"f4f7165da521966aea5c655d32a0a7c5637b7cff","ref":"refs/heads/main","pushedAt":"2024-01-03T23:50:32.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"fix shapes mod typo (#901)","shortMessageHtmlLink":"fix shapes mod typo (#901)"}},{"before":"4615ac105a16f45e45d6ccfd403099daf437a944","after":"630514f402f4340421aa4a3746ec99f29120923a","ref":"refs/heads/main","pushedAt":"2024-01-03T23:50:15.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"Implement AbsKernelOp for WebGPU backend (#896)\n\n* Removed some of the more low level commands in favor of a wrapper struct\r\n\r\nAlso added tests for higher code coverage.\r\n\r\n* AtomicPtr unsound fix\r\n\r\n* Partial implementation of `Device` for Webgpu\r\n\r\n* Remove foolish Mutex\r\n\r\n* Add Mutex back, since evidently it was causing issues.\r\n\r\nHopefully I can figure out a way to remove it again.\r\n\r\n* Removed `num_traits::Num` requirement from Zeros.\r\n\r\nHad to figure out a way to store zeros in place\r\n\r\n* Implement abs kernel, and use broken unary operation for all the compiler errors\r\n\r\n* cargo fmt\r\n\r\n* disable f16, since we don't support it yet\r\n\r\n* no-std\r\n\r\n* Added test for abs on webgpu. Also added `backward` implementation,\r\nthough I won't be able to test that until I fix `mean`.\r\n\r\n* cargo fmt\r\n\r\n* Managed to get built spirv working as long as we go through the\r\nnon-passthrough route.\r\n\r\nCan't get sum_to working until wgpu supports atomic operations. Which is\r\nsuper unfortunate.\r\n\r\nMaybe I'll work on that soon...\r\n\r\n* Have the code work correctly, almost got sum_to working, too\r\n\r\nWeird magic number issue that I can't figure out...\r\n\r\n* Cargo fmt\r\n\r\n* Do we need to skip webgpu features?","shortMessageHtmlLink":"Implement AbsKernelOp for WebGPU backend (#896)"}},{"before":"beee7a1be0e6e9803c3aab561fbce2d45c84caec","after":"4615ac105a16f45e45d6ccfd403099daf437a944","ref":"refs/heads/main","pushedAt":"2023-12-04T22:18:33.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"Implementations for u8->u64 and i8->i64 (#873)","shortMessageHtmlLink":"Implementations for u8->u64 and i8->i64 (#873)"}},{"before":"dda8daaef5ad36036ec5330efc69a227b7ca63ea","after":"beee7a1be0e6e9803c3aab561fbce2d45c84caec","ref":"refs/heads/main","pushedAt":"2023-12-04T13:48:25.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"Split `TryConcatAlong` into different traits (#892)\n\n* Added test (fails to compile)\r\n\r\n* Split `TryConcatAlong` into two traits\r\n\r\n- Deprecated `TryConcatAlong` in favor of `TryConcatTensorAlong` or `TryConcatShapeAlong`.\r\n- Created `concat_tensor_along/` and `concat_shape_along/`.\r\n - Copied relevant sections and files from `concat_along`, adjusting where necessary.\r\n - Moved `concat_along/` kernels to `concat_tensor_along/`.\r\n- Adjusted the issue's integration test to the new trait, which runs successfully.\r\n\r\n* Update dfdx-core/src/tensor_ops/concat_shape_along/mod.rs\r\n\r\nCo-authored-by: Corey Lowman \r\n\r\n* rm unneeded bound\r\n\r\n---------\r\n\r\nCo-authored-by: Corey Lowman ","shortMessageHtmlLink":"Split TryConcatAlong into different traits (#892)"}},{"before":"2ff7b60a101e474125117dbb13d274b0be41b2c1","after":"dda8daaef5ad36036ec5330efc69a227b7ca63ea","ref":"refs/heads/main","pushedAt":"2023-12-03T16:00:16.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"Implement base Webgpu device (#893)\n\n* Implement Webgpu device.\r\n\r\nAll the tests pass, but it doesn't implement `Device` yet, because that\r\nis a lot more work.\r\n\r\n* Removed some of the more low level commands in favor of a wrapper struct\r\n\r\nAlso added tests for higher code coverage.\r\n\r\n* AtomicPtr unsound fix\r\n\r\n* Partial implementation of `Device` for Webgpu\r\n\r\n* Finished up with template tensor_ops\r\n\r\n* Remove foolish Mutex\r\n\r\n* Cargo fmt\r\n\r\n* Adding f16 to webgpu\r\n\r\n* Add Mutex back, since evidently it was causing issues.\r\n\r\nHopefully I can figure out a way to remove it again.\r\n\r\n* Removed `num_traits::Num` requirement from Zeros.\r\n\r\nHad to figure out a way to store zeros in place\r\n\r\n* Fixing no-std stuff","shortMessageHtmlLink":"Implement base Webgpu device (#893)"}},{"before":"a522c7ac1e10f2dd978005fbad9216092644001f","after":"2ff7b60a101e474125117dbb13d274b0be41b2c1","ref":"refs/heads/main","pushedAt":"2023-12-02T18:57:32.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"Add compatibility for hcos/hsin/hlog/hexp on older GPUs (#887)\n\n* On older GPU there are no hcos/hsin/hlog/hexp\r\n\r\n* Update dfdx-core/src/tensor_ops/utilities/compatibility.cuh\r\n\r\n---------\r\n\r\nCo-authored-by: Corey Lowman ","shortMessageHtmlLink":"Add compatibility for hcos/hsin/hlog/hexp on older GPUs (#887)"}},{"before":"4476b5ee19dc9cc446388545560c57e80cb086c8","after":"a522c7ac1e10f2dd978005fbad9216092644001f","ref":"refs/heads/main","pushedAt":"2023-12-02T18:44:21.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"Replace explicit features and paths on generated code (#886)\n\n* Disallow feature detection on generated code\r\n\r\n- Move feature detection to before generating code.\r\n- Only emits the `SaveSafeTensors` and `LoadSafeTensors` derivations and the `#[serialize]` attr if the `safetensors` feature is enabled.\r\n\r\n* Only use `::dfdx` as a path entrypoint on code generation\r\n\r\n- Re-export the `self` crate as `dfdx`.\r\n - See [related question](https://users.rust-lang.org/t/how-to-express-crate-path-in-procedural-macros/91274/10).\r\n - Does the same for `dfdx-core`.\r\n- Change the generated code so they always have the `::dfdx` path entrypoint.\r\n - User apps no longer need to depend on `dfdx-core` or `dfdx-derives` directly.\r\n- Re-export `safetensors`.\r\n - User apps no longer need to depend on `safetensors` directly (as in, this is no longer required by the `SaveSafeTensors` and `LoadSafeTensors` derivations).\r\n\r\n* Add safetensors feature to dfdx-derives\r\n\r\n* fix typo","shortMessageHtmlLink":"Replace explicit features and paths on generated code (#886)"}},{"before":"53469e93eeb30cfca8d5d709447a9924c87bc0b0","after":"4476b5ee19dc9cc446388545560c57e80cb086c8","ref":"refs/heads/main","pushedAt":"2023-11-06T13:58:13.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"make device capable of stacking and concatenating usize type, which are necessary for preparing tensors for embedding (#882)\n\nCo-authored-by: Corey Lowman ","shortMessageHtmlLink":"make device capable of stacking and concatenating usize type, which a…"}},{"before":"370334fda03057c0e62e442262cb00e87c06c4c8","after":"53469e93eeb30cfca8d5d709447a9924c87bc0b0","ref":"refs/heads/main","pushedAt":"2023-11-06T13:57:18.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"Add `forward_mut` in `Sequential` derive (#884)\n\n* add test for try_forward_mut (fails)\r\n\r\n* add forward_mut to Sequential derive\r\n\r\n* also use the _mut for tuple structs","shortMessageHtmlLink":"Add forward_mut in Sequential derive (#884)"}},{"before":"069633a771cd9ee60219607b2ce8328cd7366b4c","after":null,"ref":"refs/heads/bump-versions","pushedAt":"2023-10-26T14:19:13.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"}},{"before":"7ba3a8b6fc94394c60a35fbb6ebfb89b2e781e4a","after":"370334fda03057c0e62e442262cb00e87c06c4c8","ref":"refs/heads/main","pushedAt":"2023-10-26T14:19:10.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"Bump dependency versions (#877)","shortMessageHtmlLink":"Bump dependency versions (#877)"}},{"before":null,"after":"069633a771cd9ee60219607b2ce8328cd7366b4c","ref":"refs/heads/bump-versions","pushedAt":"2023-10-26T13:40:06.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"Bump dependency versions","shortMessageHtmlLink":"Bump dependency versions"}},{"before":"e08e312fe75b1f1d03444049759a7f83b5f76194","after":null,"ref":"refs/heads/fix-870","pushedAt":"2023-10-25T17:33:00.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"}},{"before":"7df12c0dd913a4257934fe9a7233ade30382bcd3","after":"7ba3a8b6fc94394c60a35fbb6ebfb89b2e781e4a","ref":"refs/heads/main","pushedAt":"2023-10-25T17:32:57.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"#870 allocate 1 byte for workspace instead of 0 (#876)","shortMessageHtmlLink":"#870 allocate 1 byte for workspace instead of 0 (#876)"}},{"before":null,"after":"e08e312fe75b1f1d03444049759a7f83b5f76194","ref":"refs/heads/fix-870","pushedAt":"2023-10-25T17:32:30.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"#870 allocate 1 byte for workspace instead of 0","shortMessageHtmlLink":"#870 allocate 1 byte for workspace instead of 0"}},{"before":"d36c60afcd92384c4b204cac1034058547d0dbce","after":null,"ref":"refs/heads/single-error","pushedAt":"2023-10-25T16:52:04.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"}},{"before":"5e0c3ddda5de3258d52f9d728426adc191ae1f5e","after":"7df12c0dd913a4257934fe9a7233ade30382bcd3","ref":"refs/heads/main","pushedAt":"2023-10-25T16:52:01.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"[Breaking] Combining separate device errors into single `dfdx::tensor::Error` enum (#875)\n\n* [Breaking] Adding single Error enum\r\n\r\n* Fixing example\r\n\r\n* Fixing cuda kernels\r\n\r\n* Fixing no-std","shortMessageHtmlLink":"[Breaking] Combining separate device errors into single `dfdx::tensor…"}},{"before":"9a2564a61e954288cb350ec225e57e77d525db21","after":"d36c60afcd92384c4b204cac1034058547d0dbce","ref":"refs/heads/single-error","pushedAt":"2023-10-25T16:39:57.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"Fixing no-std","shortMessageHtmlLink":"Fixing no-std"}},{"before":"ca9548b986d0356670dc8a4b46516b1e66645854","after":"9a2564a61e954288cb350ec225e57e77d525db21","ref":"refs/heads/single-error","pushedAt":"2023-10-25T16:29:49.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"Fixing cuda kernels","shortMessageHtmlLink":"Fixing cuda kernels"}},{"before":null,"after":"ca9548b986d0356670dc8a4b46516b1e66645854","ref":"refs/heads/single-error","pushedAt":"2023-10-25T16:21:30.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"Fixing example","shortMessageHtmlLink":"Fixing example"}},{"before":"aef72b0e8d2e5d21603a4c112484dea787b0abea","after":null,"ref":"refs/heads/nn-rewrite","pushedAt":"2023-10-25T15:14:46.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"}},{"before":"9a208d928b92dd77d4b2fa79144bdc8b872aa7f0","after":"5e0c3ddda5de3258d52f9d728426adc191ae1f5e","ref":"refs/heads/main","pushedAt":"2023-10-25T15:14:42.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"[Breaking] Rewrite of `nn` to enable runtime layer sizes, proc macro declarations, and more (#854)","shortMessageHtmlLink":"[Breaking] Rewrite of nn to enable runtime layer sizes, proc macro …"}},{"before":"a4db3d5d07872a9835399c161e9c4229db132bba","after":"aef72b0e8d2e5d21603a4c112484dea787b0abea","ref":"refs/heads/nn-rewrite","pushedAt":"2023-10-25T14:58:41.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"Update no-std","shortMessageHtmlLink":"Update no-std"}},{"before":"2ac9dc96fc21325f51e577175a03d06e869429ee","after":"a4db3d5d07872a9835399c161e9c4229db132bba","ref":"refs/heads/nn-rewrite","pushedAt":"2023-10-25T14:54:58.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"Moving feature flags & top level documnetation","shortMessageHtmlLink":"Moving feature flags & top level documnetation"}},{"before":"929602eb25719699f53bdb65b0a27759cd8a787c","after":"2ac9dc96fc21325f51e577175a03d06e869429ee","ref":"refs/heads/nn-rewrite","pushedAt":"2023-10-25T14:51:32.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"Fixing cargo doc","shortMessageHtmlLink":"Fixing cargo doc"}},{"before":"f27a69243a5668b3660e4276556385c10b6fcdb5","after":"929602eb25719699f53bdb65b0a27759cd8a787c","ref":"refs/heads/nn-rewrite","pushedAt":"2023-10-25T14:44:42.000Z","pushType":"push","commitsCount":7,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"Fixing doctests","shortMessageHtmlLink":"Fixing doctests"}},{"before":"16fec9b25d502b2fa7fec72e47af7139cb38799b","after":"f27a69243a5668b3660e4276556385c10b6fcdb5","ref":"refs/heads/nn-rewrite","pushedAt":"2023-09-14T17:36:00.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"coreylowman","name":"Corey Lowman","path":"/coreylowman","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7787278?s=80&v=4"},"commit":{"message":"Fixing nightly feature propagation","shortMessageHtmlLink":"Fixing nightly feature propagation"}}],"hasNextPage":true,"hasPreviousPage":false,"activityType":"all","actor":null,"timePeriod":"all","sort":"DESC","perPage":30,"cursor":"djE6ks8AAAAD6UjZ5wA","startCursor":null,"endCursor":null}},"title":"Activity · coreylowman/dfdx"}