Skip to content

Fix breaking CIs #2906

@justinchuby

Description

@justinchuby
  1. ort-nightly ci
FAILED onnxscript/rewriter/rules/common/_basic_rules_test.py::ReshapeReshapeTest::test_reshape_reshape_dynamic_rule_1 - AssertionError: 
Not equal to tolerance rtol=0, atol=0

(shapes (1, 0, 6, 3), (1, 0, 0, 3) mismatch)
 ACTUAL: array([], shape=(1, 0, 6, 3), dtype=float32)
 DESIRED: array([], shape=(1, 0, 0, 3), dtype=float32)
  1. torch-nightly ci. OK to relax tolerance
SUBFAILED(sample_num=1, inputs="['Tensor<torch.Size([5, 5]), dtype=torch.float16>', (1,), True]", kwargs='{}') tests/function_libs/torch_lib/ops_test.py::TestOutputConsistencyFullGraphCPU::test_output_match_opinfo__logsumexp_cpu_float16 - AssertionError: Tensor-likes are not close!

Mismatched elements: 1 / 5 (20.0%)
Greatest absolute difference: 2.288818359375e-05 at index (1, 0) (up to 1e-05 allowed)
Greatest relative difference: 0.0022869110107421875 at index (1, 0) (up to 0.001 allowed)
SUBFAILED(sample_num=2, inputs="['Tensor<torch.Size([5, 5]), dtype=torch.float16>', (1,), False]", kwargs='{}') tests/function_libs/torch_lib/ops_test.py::TestOutputConsistencyFullGraphCPU::test_output_match_opinfo__logsumexp_cpu_float16 - AssertionError: Tensor-likes are not close!

Mismatched elements: 1 / 5 (20.0%)
Greatest absolute difference: 2.288818359375e-05 at index (1,) (up to 1e-05 allowed)
Greatest relative difference: 0.0022869110107421875 at index (1,) (up to 0.001 allowed)

Metadata

Metadata

Labels

No labels
No labels

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions