Skip to content

Skip constant folding for non-deterministic Bernoulli #3072

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

Open
Mohammed Alkindi (MohammedAlkindi) wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Skip constant folding for non-deterministic Bernoulli #3072

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
1 change: 1 addition & 0 deletions onnxscript/optimizer/_constant_folding.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -53,6 +53,7 @@
"RandomUniformLike",
"RandomNormalLike",
"Multinomial",
"Bernoulli",
}
)

Expand Down
15 changes: 15 additions & 0 deletions onnxscript/optimizer/_constant_folding_test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -869,6 +869,21 @@ def test_dequantize_linear_is_not_folded(self):
# DequantizeLinear should not be folded even when all inputs are constants
self.assertEqual(ops, ["DequantizeLinear"])

def test_bernoulli_is_not_folded(self):
model_text = """
<ir_version: 10, opset_import: [ "" : 18]>
agraph () => (float[4] z)
<float[4] p = {0.5, 0.5, 0.5, 0.5}>
{
z = Bernoulli (p)
}
"""
model = ir.from_onnx_text(model_text)
optimized = self._fold(model)
ops = [node.op_type for node in optimized.graph]
# Bernoulli is random, so folding it would freeze a single sample
self.assertEqual(ops, ["Bernoulli"])

def test_multi_graph_identity_output_preserves_output_name(self):
model = """
<ir_version: 10, opset_import: ["" : 20]>
Expand Down