Hidet v0.2.1

What's Changed

• [Version] Bump version to 0.2.1.dev by @yaoyaoding in #73
• [CI] Prevent fork repos from running workflow by @yaoyaoding in #74
• [Fixbug] Fix a bug in trace_from when the inputs are directly used as outputs by @yaoyaoding in #76
• [Operator] Add reduce_f16 and squeeze as Reduce's resolve variants by @hjjq in #75
• [IR] Input specification assertion message for valid IR check by @AndreSlavescu in #78
• [Operator] Add conv3d, max_pool3d, avg_pool3d by @hjjq in #79
• [Dynamo] Add the entry point registration for dynamo by @yaoyaoding in #80
• [Fix] Update shape utility functions to expect Sequence instead of List by @yaoyaoding in #86
• [Bugfix] 'double'->'float64' in onnx dtype conversion by @soodoshll in #88
• [Fix] Mark the reduce fp16 operator not fusible by @yaoyaoding in #100
• [Fixbug] Use uint64_t instead of unsigned long long for literals by @yaoyaoding in #101
• [Fixbug] Fix a bug in the minimum and maximum operator by @yaoyaoding in #102
• [Dynamo] Update dynamo registration after pytorch refactored that part by @yaoyaoding in #84
• [Fixbug] Fix bugs in binary_arithmetic op and swizzle layout by @hjjq in #104
• [Fixbug] Call fuse in reduce_fp16 operator by @yaoyaoding in #105
• [ONNX] Fix the out of bound error in onnx slice function during importing by @yaoyaoding in #106
• [Fixbug] Reverse map of binary operator by @yaoyaoding in #107
• [Fixbug] Add attributes to Clip operator by @yaoyaoding in #108
• [Fixbug] Binary arthmatic ops raise error when one is scalar on GPU by @yaoyaoding in #109
• [Graph] Refactor forward function of FlowGraph by @yaoyaoding in #110
• [Fixbug] Use int64 as the output of arg-reduce by @yaoyaoding in #111
• [README] Update readme by @yaoyaoding in #114
• [Fixbug] Fix a bug when an graph output is constant by @yaoyaoding in #113
• [Community] Create CODE_OF_CONDUCT.md by @yaoyaoding in #115
• [Community] Update issue templates by @yaoyaoding in #116
• [Fixbug] Resolve the min/max function according to compute capability by @yaoyaoding in #112
• [Workflow] Update workflow by @yaoyaoding in #117
• [Workflow] Update publish workflow by @yaoyaoding in #119

New Contributors

• @soodoshll made their first contribution in #88

Full Changelog: v0.2.0...v0.2.1

Hidet v0.2.0

What's Changed

• [Version] Bump version to 0.2.dev by @yaoyaoding in #60
• [Frontend] Add torch.tensor binding by @yaoyaoding in #61
• [Version] Add version to root namespace by @yaoyaoding in #62
• [FFI] Add SharedLibrary class to track the usage of dynamic library by @yaoyaoding in #63
• [Operator] Fix a bug in resize2d operator defintion by @yaoyaoding in #64
• [CI] Update scripts to build wheel by @yaoyaoding in #65
• [CI] Remove docs workflow by @yaoyaoding in #66
• [Docs] Update README.md and require cuda-python>=11.6.1 by @yaoyaoding in #67
• [Docs] Add instructions for installing nightly version of hidet by @yaoyaoding in #68
• [Docs] Fixed typo in docs by @AndreSlavescu in #69
• [Operator] Add dilation support for conv2d by @hjjq in #71
• [Fixbug] Cast back to original data type in the mix precision pass: by @yaoyaoding in #72
• [CI] Add automatic publish workflow (PyPI) by @yaoyaoding in #70

New Contributors

• @AndreSlavescu made their first contribution in #69

Full Changelog: v0.1...v0.2.0

Hidet v0.1

This is the first release of hidet.

For the usage of hidet, please visit: https://docs.hidet.org

What's Changed

• [Docs] Update documentation by @yaoyaoding in #2
• [Operator] Add leaky_relu and conv2d_transpose operator by @yaoyaoding in #3
• [Doc] Add doc on how to define operator computation by @yaoyaoding in #4
• [Bug] fix bugs in reshape and conv2d_transpose by @yaoyaoding in #5
• [Option] Add option module by @yaoyaoding in #6
• [Docs] Add documentation on how to add new operators by @yaoyaoding in #7
• [Operator] Add PRelu op by @hjjq in #8
• [Docs] Add documentation for operator cache & fix a typo by @yaoyaoding in #9
• [Operator] Add Abs and And operator by @hjjq in #10
• [CI] Update github workflow by @yaoyaoding in #11
• [CI] Update docs workflow, not delete remote dest dir by @yaoyaoding in #12
• [Operator] Add conv2d_transpose_gemm operator & fix a bug by @yaoyaoding in #13
• [Runtime] force to use gpu tensor buffer in cuda graph by @yaoyaoding in #14
• [Functor] Fix a bug in IR functor by @yaoyaoding in #15
• [Graph] Force users to give an input order when multiple symbolic inputs are found in traced graph by @yaoyaoding in #17
• [Operator] Add BitShift, Bitwise*, Ceil Operators by @hjjq in #19
• [IR] Refactor scalar type system by @yaoyaoding in #18
• [IR] Refactoring math functions by @yaoyaoding in #20
• [Operator] Fix a bug when resolve matmul to batch_matmul by @yaoyaoding in #21
• [Operator] Add cubic interpolation to Resize Operator by @hjjq in #22
• [Packfunc] Refactor packed func & add vector type by @yaoyaoding in #23
• [Pass] Add lower_special_cast pass and refactor resolve rule registration by @yaoyaoding in #24
• [Docs] Change github repo url by @yaoyaoding in #25
• [Operator] Add float16 precision matrix multiplication by @yaoyaoding in #26
• [Docs] Add a guide on operator resolving by @yaoyaoding in #27
• [CI] Avoid interactive query in apt installation of tzdata by @yaoyaoding in #28
• [Docs] Add sub-graph rewrite tutorial by @yaoyaoding in #29
• [Tensor] Implement dlpack tensor exchange protocol by @yaoyaoding in #30
• [Frontend] Add a torch dynamo backend based on hidet "onnx2hidet" by @yaoyaoding in #31
• [Frontend] Add hidet dynamo backend based on torch.fx by @yaoyaoding in #32
• [Frontend] Make onnx dependency optional by @yaoyaoding in #33
• [Frontend] Add more operator mappings for pytorch frontend by @yaoyaoding in #34
• [Opeartor] Fix a bug in take (index can be in [-r, r-1]) by @yaoyaoding in #35
• [Frontend] Add an option to print correctness report in hidet backend of torch dynamo by @yaoyaoding in #36
• [IR] Refactor the attribute 'dtype' of hidet.Tensor from 'str' to 'DataType' by @yaoyaoding in #37
• [Operator] Add a constant operator and deprecates manually implemented fill cuda kernel by @yaoyaoding in #38
• [ONNX] Add reduce l2 onnx operator by @yaoyaoding in #40
• [CLI] Add the 'hidet' command line interface by @yaoyaoding in #39
• [Codegen] Add explicit conversion type for float16 by @yaoyaoding in #41
• [Docs] Add the documentation for 'hidet' backend of PyTorch dynamo by @yaoyaoding in #42
• [Runtime] Refactor the cuda runtime api used in hidet by @yaoyaoding in #43
• [Testing] Remove redundant models in hidet.testing by @yaoyaoding in #44
• [Runtime][IR] Refactor the device attribute of Tensor object by @yaoyaoding in #45
• [Array-API][Phase 0] Adding the declarations of missing operators in Array API by @yaoyaoding in #46
• [Operator] Add arange and linspace operator by @yaoyaoding in #47
• [Bug] Fix a bug related to memset by @yaoyaoding in #49
• [Docs] Add and update documentation by @yaoyaoding in #48
• [Docs][Opeartor] Add more pytorch operator bindings and docs by @yaoyaoding in #50
• [License][Docs] Add license header and update README.md by @yaoyaoding in #51
• [Docs] Update docs by @yaoyaoding in #52
• [IR] Add LaunchKernelStmt by @yaoyaoding in #53
• [Operator] Add some torch operator mapping by @yaoyaoding in #54
• [Bug] Fix a bug in hidet dynamo backend when cuda graph is not used by @yaoyaoding in #55
• [Dynamo] Allow torch dynamo backend to accepts non-contiguous input by @yaoyaoding in #56
• [Graph] Add to_cuda() for Module class by @hjjq in #57
• [Bug] Fix a bug where the shared memory becomes zero in LaunchKernelStmt by @yaoyaoding in #58
• [Release] Prepare to release the first version of hidet to public by @yaoyaoding in #59

New Contributors

• @hjjq made their first contribution in #8

Full Changelog: https://github.com/hidet-org/hidet/commits/v0.1