Add unique op #1547
Add unique op #1547
Conversation
![github-advanced-security[bot]](https://avatars.githubusercontent.com/in/57789?s=60&v=4){kind=small}
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #1547 +/- ##
==========================================
- Coverage 72.28% 72.25% -0.03%
==========================================
Files 217 217
Lines 29097 29138 +41
Branches 3455 3462 +7
==========================================
+ Hits 21034 21055 +21
- Misses 6935 6953 +18
- Partials 1128 1130 +2 ☔ View full report in Codecov by Sentry. |
| except Exception as e: | ||
| # try to provide a more informative error message | ||
| if _NOT_IMPLEMENTED_UNIQUE.search(str(e)) is not None: | ||
| raise NotImplementedError( | ||
| f"'onnxruntime' does not yet support Unique(11) operator with dtype={self.dtype}'" | ||
| ) from e |
There was a problem hiding this comment.
I would remove this try-catch as the function here is symbolic; we don't expect them to raise any errors
justinchuby
added
the
module: torchlib
Yea, I may have jumped the gun a bit. Working on officially getting permission from my employer. |
@microsoft-github-policy-service agree [company="Radiance Technologies"] @microsoft-github-policy-service agree company="Radiance Technologies" |
|
@microsoft-github-policy-service agree company="Radiance Technologies" |
a-gardner1
force-pushed
the
wip-113118-add-unique-ops
branch
from
453783f to
b528a6a
Compare
|
Thanks for completing the CLA. I will take a look next week |
a-gardner1
force-pushed
the
wip-113118-add-unique-ops
branch
from
56c06cf to
7e6d906
Compare
Follow-up to #113118 and #124306. Developed in coordination with the solution to microsoft/onnxscript#1547 This PR adds the missing fake tensor implementation for `aten.unique_dim`, thus enabling tracing and compilation of `torch.unique` when `dim` is not None. Local testing has proceeded with the following simple script (provided that one has checked out the changes in microsoft/onnxscript#1547): ```python import onnx import onnxruntime as ort import logging import numpy as np onnx_program = torch.onnx.dynamo_export( lambda x: torch.unique(x, dim=0, return_inverse=True), torch.arange(10), export_options=torch.onnx.ExportOptions( dynamic_shapes=True, diagnostic_options=torch.onnx.DiagnosticOptions( verbosity_level=logging.DEBUG))) onnx_program.save("torch_unique.onnx") onnx_inputs = onnx_program.adapt_torch_inputs_to_onnx(torch.arange(10)) onnx_outputs = onnx_program(*onnx_inputs) loaded_onnx_program = onnx.load("torch_unique.onnx") onnx.checker.check_model(loaded_onnx_program) ort_session = ort.InferenceSession("torch_unique.onnx") inputs = np.random.randint(0, 10, 10) print(f"Inputs: {inputs}") outputs = ort_session.run(None, { "l_x_": inputs }) print(f"Outputs: {outputs}") print("Success") ``` Co-authored-by: Edward Z. Yang <ezyang@meta.com> Pull Request resolved: #126561 Approved by: https://github.com/ezyang
Follow-up to pytorch#113118 and pytorch#124306. Developed in coordination with the solution to microsoft/onnxscript#1547 This PR adds the missing fake tensor implementation for `aten.unique_dim`, thus enabling tracing and compilation of `torch.unique` when `dim` is not None. Local testing has proceeded with the following simple script (provided that one has checked out the changes in microsoft/onnxscript#1547): ```python import onnx import onnxruntime as ort import logging import numpy as np onnx_program = torch.onnx.dynamo_export( lambda x: torch.unique(x, dim=0, return_inverse=True), torch.arange(10), export_options=torch.onnx.ExportOptions( dynamic_shapes=True, diagnostic_options=torch.onnx.DiagnosticOptions( verbosity_level=logging.DEBUG))) onnx_program.save("torch_unique.onnx") onnx_inputs = onnx_program.adapt_torch_inputs_to_onnx(torch.arange(10)) onnx_outputs = onnx_program(*onnx_inputs) loaded_onnx_program = onnx.load("torch_unique.onnx") onnx.checker.check_model(loaded_onnx_program) ort_session = ort.InferenceSession("torch_unique.onnx") inputs = np.random.randint(0, 10, 10) print(f"Inputs: {inputs}") outputs = ort_session.run(None, { "l_x_": inputs }) print(f"Outputs: {outputs}") print("Success") ``` Co-authored-by: Edward Z. Yang <ezyang@meta.com> Pull Request resolved: pytorch#126561 Approved by: https://github.com/ezyang
|
Circling back around to this @justinchuby. At the time, I had been waiting for you to resolve the bug that required a hacky workaround, but I realize that might not be clear. There were also a couple of other potential unresolved bugs outside the scope of this PR, e.g., this comment. How would you like to proceed? |
|
Sorry for missing the clarity. I would suggest that you remove all the hacks so that the code is at its desirable state. If tests fail because of that, that’s ok. I will then go ahead to fix what’s needed. (After I’m back from vacation) |
Sounds good. FYI, the hacks were removed in b8b4cb1. As a reminder, the unit tests within If I get a chance, I'll try to rebase this PR and resolve conflicts first. |
|
I am implementing a CTC decoder class in pytorch - I'm not able to export this class to onnx. Here is my error - How can I resolve this? |
|
@a-gardner1 sorry for the delay. I think this PR is now good to merge (started the merge) |
Add support for exporting
torch.uniquefollowing the conclusion of pytorch/pytorch#113118.