torch-mlir/test/Dialect/Torch
Ramiro Leal-Cavazos 273664ded6
[custom op] Replace `tanh` dtype function with `expm1` (#1769)
This commit replaces the `tanh` dtype function, which was being used
to test the implementation of dtype functions in
a710237437, with a dtype function for
`expm1`. The dtype function for `expm1` is identical to the `tanh`
one, so the same level of testing is maintained.

Currently, there are ops getting dtype information from the
`RefineTypes` pass and ops getting dtype information from the
`TorchDtypeRefinementPipeline`. Since each pass can only propagete
dtype information for the ops it knows how to handle, some models with
many ops handled in both passes require the two dtype propagation
passes to execute many times, reaching the iteration limit set in the
`LowerToBackendContractPass`. To temporarily avoid this issue while
the migration to `TorchDtypeRefinementPipeline` is finished, this
commit switches `tanh` to `expm1`, since the latter is used a lot less
in large models.
2023-01-03 14:18:26 -08:00
..
GlobalizeObjectGraph Rework how global slot initializers work. 2022-08-08 18:12:06 -07:00
adjust-calling-conventions.mlir Move single-tensor-tuple-return test to mlir unit test. 2022-11-10 09:23:53 -05:00
canonicalize.mlir Add aten.slice.Tensor & aten.cat folders (#1691) 2022-12-13 13:02:47 -08:00
decompose-complex-ops-legal.mlir Add a concept of "backend legal ops". 2022-08-18 11:46:50 -07:00
decompose-complex-ops.mlir Allow running DecomposeComplexOps more than once (#1671) 2022-12-08 09:26:38 -08:00
drop-abstract-interp-calculations.mlir [custom op] Generalize shape library logic to work with dtypes (#1594) 2022-12-13 08:25:41 -08:00
erase-module-initializer.mlir Iteratively run the main simplification pipeline. 2022-08-17 14:54:33 -07:00
inline-global-slots-analysis.mlir Rework how global slot initializers work. 2022-08-08 18:12:06 -07:00
inline-global-slots-transform.mlir Rework how global slot initializers work. 2022-08-08 18:12:06 -07:00
invalid.mlir Rework how global slot initializers work. 2022-08-08 18:12:06 -07:00
lower-to-backend-contract-error.mlir Allow running DecomposeComplexOps more than once (#1671) 2022-12-08 09:26:38 -08:00
maximize-value-semantics.mlir build: update llvm tag to 798fa4b4 (#1684) 2022-12-07 12:20:41 -08:00
ops.mlir [custom op] Generalize shape library logic to work with dtypes (#1594) 2022-12-13 08:25:41 -08:00
prepare-for-globalize-object-graph.mlir mlir: bump llvm tag to 5380e3 (#856) 2022-05-16 12:54:35 -07:00
reduce-op-variants-error.mlir mlir: bump llvm tag to 5380e3 (#856) 2022-05-16 12:54:35 -07:00
reduce-op-variants.mlir Remove all but one of valsem ops + move fill.Scalar to elementwise (#1531) 2022-10-28 15:06:11 +00:00
refine-public-return.mlir Add support for returning more than one copy of the same tensor (#1228) 2022-08-18 22:41:45 +00:00
refine-types-branch.mlir [custom op] Generalize shape library logic to work with dtypes (#1594) 2022-12-13 08:25:41 -08:00
refine-types-ops.mlir mlir: bump llvm tag to 5380e3 (#856) 2022-05-16 12:54:35 -07:00
refine-types.mlir [custom op] Generalize shape library logic to work with dtypes (#1594) 2022-12-13 08:25:41 -08:00
reify-dtype-calculations.mlir [custom op] Replace `tanh` dtype function with `expm1` (#1769) 2023-01-03 14:18:26 -08:00
reify-shape-calculations.mlir Remove all but one of valsem ops + move fill.Scalar to elementwise (#1531) 2022-10-28 15:06:11 +00:00
simplify-dtype-calculations.mlir [TORCH] Add Complex Number support (#1673) 2022-12-15 21:40:01 +00:00
simplify-shape-calculations.mlir canonicalizer: [nfc] update LIT variable names for consistency (#1051) 2022-07-13 12:28:25 -07:00
torch-function-to-torch-backend-pipeline.mlir update llvm tag to a2620e00. (#1567) 2022-11-10 18:39:28 +08:00
verify-backend-contract-error.mlir Pass op legality info to `verifyBackendContractPass` (#1705) 2022-12-15 08:32:52 -08:00