torch-mlir

Commit Graph

Author	SHA1	Message	Date
Maksim Levental	d46f169c1a	Fix kwarg annotation in eager (#747 )	2022-04-11 17:35:42 -05:00
Maksim Levental	66de821eaf	small framework plus build_script_function (#745 )	2022-04-11 16:53:52 -05:00
Maksim Levental	18ef40acaf	Fixes a bug in use of upstream `normalize_function` in our `normalize_args_kwargs` (in eager mode) and introduces unit tests. (#740 ) NB: `shouldnt_normalize2` and `shouldnt_normalize3` currently XPASS i.e., args will successfully normalize despite being incorrect due to an [upstream bug](https://github.com/pytorch/pytorch/issues/75342).	2022-04-11 16:17:44 -05:00
Sean Silva	14cf87633c	Add link to forum post describing `__torch_dispatch__`	2022-04-01 10:10:43 -07:00
Maksim Levental	3e999beaea	Small bug fixes in eager mode (#691 )	2022-03-28 13:31:07 -05:00
max	fe8ac57e6d	This PR implements an eager mode backend for PyTorch through the torch-mlir framework. This is accomplished by overriding the `__torch_dispatch__` class method on wrapper subclass `TorchMLIRTensor(torch.Tensor)`. Effectively, this mode works by compiling op by op as the NN is eagerly executed by PyTorch. Entailed in that compilation is building a representation of the op that can be `torch.jit.script`ed, importing using `ModuleBuilder`, and then executing (e.g., with `RefBackendLinalgOnTensorsBackend`). This mode includes a fallback to conventional PyTorch if anything in the torch-mlir compilation process fails (e.g., unsupported op). Currently, all e2e tests pass execpt for two that involve an upstream PyTorch bug (https://github.com/pytorch/pytorch/issues/74400). High priority next steps: 1. A compile cache in order to speed up reruns of the same NN. 2. Integration with IREE (though not in this repo). 3. Integration with `torch.distributed`.	2022-03-22 14:42:57 -07:00

6 Commits (a893c7d5cffc35e376767d959063fe4c1553909d)