Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[torchlib] Fix _log_softmax #1789

Merged
merged 7 commits into from
Aug 8, 2024
Merged

[torchlib] Fix _log_softmax #1789

Changes from all commits
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
5235261
[torchlib] Specify the squeeze axis
justinchuby Aug 7, 2024
6f7cb50
Update core.py
justinchuby Aug 7, 2024
dad8f8c
Update core.py
justinchuby Aug 7, 2024
519b526
Update core.py
justinchuby Aug 7, 2024
eeb2e23
Update core.py
justinchuby Aug 7, 2024
bd7c050
Update core.py
justinchuby Aug 7, 2024
daeb3b3
Update onnxscript/function_libs/torch_lib/ops/core.py
justinchuby Aug 7, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 8 additions & 4 deletions onnxscript/function_libs/torch_lib/ops/core.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -82,11 +82,15 @@
) -> FLOAT:
"""_log_softmax(Tensor self, int dim, bool half_to_float) -> Tensor"""

# trace_only because we need to cast conditionally based on half_to_float
self_is_scalar = IsScalar(self)
if half_to_float:
self = op.Cast(self, to=FLOAT.dtype)

return aten__log_softmax(self, dim, half_to_float)
if self_is_scalar:
self = op.Unsqueeze(self, op.Constant(value_ints=[0]))

Check warning on line 89 in onnxscript/function_libs/torch_lib/ops/core.py

View check run for this annotation

Codecov / codecov/patch

onnxscript/function_libs/torch_lib/ops/core.py#L89 

Added line #L89 was not covered by tests
result = op.LogSoftmax(self, axis=dim)
if self_is_scalar:
result = op.Squeeze(result, op.Constant(value_ints=[0]))

Check warning on line 92 in onnxscript/function_libs/torch_lib/ops/core.py

View check run for this annotation

Codecov / codecov/patch

onnxscript/function_libs/torch_lib/ops/core.py#L92 

Added line #L92 was not covered by tests
return result


@torch_op("aten::_log_softmax", traceable=True)
Expand All @@ -101,7 +105,7 @@
if self_is_scalar:
self = op.Unsqueeze(self, op.Constant(value_ints=[0]))
result = op.LogSoftmax(self, axis=dim)
if self_is_scalar: # squeeze to scalar due to input is scalar
if self_is_scalar:
result = op.Squeeze(result)
return result

Expand Down
Loading