-
Bug
-
Resolution: Unresolved
-
Major
-
None
-
None
-
False
-
-
False
-
-
Description
Summary: Multiple run2run determinism tests failing during PyTorch unit test execution on CPU and sGPU platforms.
Test Class: inductor/test_deterministic.py::DeterministicTest
Number of Failing Tests: 10
Platform: CPU, sGPU (CUDA)
Test Type: Unit Test
Version Information:
- PyTorch Commit: 6bdd8c9
- Branch: main
- Test Date: 2026-01-14
- Sprint: Sprint 24
Failure Pattern:
Tests failing with related import/pandas compatibility errors - likely common root cause
Common Error:
code
AssertionError: False is not true : stdout: , stderr: Traceback (most recent call last):
File "/pytorch/benchmarks/dynamo/huggingface.py", line 15, in <module>
from .common import (
ImportError: attempted relative import with no known parent package
During handling of the above exception, another exception occurred:
File "/pytorch/benchmarks/dynamo/common.py", line 32, in <module>
import pandas as pd
File "/miniconda/envs/cuda_torch_build/lib/python3.12/site-packages/pandas/_init_.py", line 22, in <module>
from pandas.compat import is_numpy_dev as _is_numpy_dev
code
Failing Tests:
1. test_run2run_determinism_model_name_BertForMaskedLM_training_or_inference_inference_precision_float16
2. test_run2run_determinism_model_name_DistillGPT2_training_or_inference_inference_precision_float16
3. test_run2run_determinism_model_name_DistillGPT2_training_or_inference_inference_precision_float32
4. test_run2run_determinism_model_name_GoogleFnet_training_or_inference_inference_precision_bfloat16
5. test_run2run_determinism_model_name_BertForMaskedLM_training_or_inference_inference_precision_float32
6. test_run2run_determinism_model_name_DistillGPT2_training_or_inference_inference_precision_amp
7. test_run2run_determinism_model_name_DistillGPT2_training_or_inference_inference_precision_bfloat16
8. test_run2run_determinism_model_name_GoogleFnet_training_or_inference_inference_precision_amp
9. test_run2run_determinism_model_name_GoogleFnet_training_or_inference_inference_precision_float16
10. test_run2run_determinism_model_name_GoogleFnet_training_or_inference_inference_precision_float32
Steps to Reproduce:
code
TEST_CONFIG=cpu python3 test/run_test.py -i inductor/test_deterministic
code
Expected Result:
Tests should pass without import errors
Actual Result:
Tests fail with pandas/numpy import compatibility errors
Priority: P2