Reconcile PyTorch pass pipelines with those upstream in torch-mlir (if needed) #17021
Labels
integrations/pytorch
PyTorch integration work
Comments
|
Related: nod-ai/SHARK-TestSuite#47 |
|
Related nod-ai/SHARK-TestSuite#225 |
saienduri
added a commit
that referenced
this issue
May 23, 2024
This commit updates `iree-import-onnx` so that it behaves the same as torch-mlir's version (https://github.com/llvm/torch-mlir/blob/main/python/torch_mlir/tools/import_onnx/__main__.py). Specifically, enabling the data propagation improves shape inference and is leading to more models passing. Related to #17021 --------- Signed-off-by: saienduri <saimanas.enduri@amd.com>
gglangg
pushed a commit
to gglangg/iree
that referenced
this issue
Jun 4, 2024
This commit updates `iree-import-onnx` so that it behaves the same as torch-mlir's version (https://github.com/llvm/torch-mlir/blob/main/python/torch_mlir/tools/import_onnx/__main__.py). Specifically, enabling the data propagation improves shape inference and is leading to more models passing. Related to iree-org#17021 --------- Signed-off-by: saienduri <saimanas.enduri@amd.com>
gglangg
pushed a commit
to gglangg/iree
that referenced
this issue
Jun 4, 2024
This commit updates `iree-import-onnx` so that it behaves the same as torch-mlir's version (https://github.com/llvm/torch-mlir/blob/main/python/torch_mlir/tools/import_onnx/__main__.py). Specifically, enabling the data propagation improves shape inference and is leading to more models passing. Related to iree-org#17021 --------- Signed-off-by: saienduri <saimanas.enduri@amd.com>
bangtianliu
pushed a commit
to bangtianliu/iree
that referenced
this issue
Jun 5, 2024
This commit updates `iree-import-onnx` so that it behaves the same as torch-mlir's version (https://github.com/llvm/torch-mlir/blob/main/python/torch_mlir/tools/import_onnx/__main__.py). Specifically, enabling the data propagation improves shape inference and is leading to more models passing. Related to iree-org#17021 --------- Signed-off-by: saienduri <saimanas.enduri@amd.com>
LLITCHEV
pushed a commit
to LLITCHEV/iree
that referenced
this issue
Jul 30, 2024
This commit updates `iree-import-onnx` so that it behaves the same as torch-mlir's version (https://github.com/llvm/torch-mlir/blob/main/python/torch_mlir/tools/import_onnx/__main__.py). Specifically, enabling the data propagation improves shape inference and is leading to more models passing. Related to iree-org#17021 --------- Signed-off-by: saienduri <saimanas.enduri@amd.com> Signed-off-by: Lubo Litchev <lubol@google.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
We have some people running "e2e" test workflows in torch-mlir that end up differing when running through IREE. I suspect the pipelines may have drifted a bit.
linalg:
onnx:
The text was updated successfully, but these errors were encountered: