fix for intra op threads issue

[aarushjain29]

Motivation

Adds a new CLI option to the accuracy checker to control ONNX Runtime session threading, intended to mitigate intra-op thread affinity issues.

Changelog Category

Add a CHANGELOG.md entry for any option other than Not Applicable

• • Added: New functionality.
• • Changed: Changes to existing functionality.
• • Removed: Functionality or support that has been removed. (Compared to a previous release)
• • Optimized: Component performance that has been optimized or improved.
• • Resolved Issues: Known issues from a previous version that have been resolved.
• • Not Applicable: This PR is not to be included in the changelog.

[Copilot]

Pull request overview

Adds a new CLI option to the accuracy checker to control ONNX Runtime session threading, intended to mitigate intra-op thread affinity issues (notably in container/cgroup cpuset environments).

Changes:

• Add --ort-intra-op-num-threads CLI flag to set onnxruntime.SessionOptions.intra_op_num_threads.
• When the flag is provided, also set SessionOptions.inter_op_num_threads = 1 alongside intra-op threads.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

Additional details and impacted files

@@           Coverage Diff            @@
##           develop    #4810   +/-   ##
========================================
  Coverage    92.49%   92.49%           
========================================
  Files          583      583           
  Lines        29562    29562           
========================================
  Hits         27343    27343           
  Misses        2219     2219           

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[causten]

Test	Batch	New Rate (92020d)	Old Rate (0681c6)	Diff	Status
torchvision-resnet50	64	3,124.97	3,166.53	-1.31%	✅
torchvision-resnet50_fp16	64	6,510.15	6,820.89	-4.56%	✅
torchvision-densenet121	32	2,623.38	2,422.10	8.31%	🔆
torchvision-densenet121_fp16	32	4,378.43	4,086.15	7.15%	🔆
torchvision-inceptionv3	32	1,799.51	1,658.01	8.53%	🔆
torchvision-inceptionv3_fp16	32	2,774.00	2,677.22	3.61%	✅
cadene-inceptionv4	16	826.45	789.08	4.74%	✅
cadene-resnext64x4	16	779.02	790.86	-1.50%	✅
slim-mobilenet	64	8,322.37	8,521.12	-2.33%	✅
slim-nasnetalarge	64	228.14	218.37	4.48%	✅
slim-resnet50v2	64	3,271.98	3,288.39	-0.50%	✅
bert-mrpc-onnx	8	1,123.23	1,108.72	1.31%	✅
bert-mrpc-tf	1	454.57	456.31	-0.38%	✅
pytorch-examples-wlang-gru	1	359.95	338.72	6.27%	🔆
pytorch-examples-wlang-lstm	1	468.17	452.66	3.43%	✅
torchvision-resnet50_1	1	758.83	748.69	1.35%	✅
cadene-dpn92_1	1	438.41	403.71	8.60%	🔆
cadene-resnext101_1	1	366.33	323.10	13.38%	🔆
onnx-taau-downsample	1	397.99	391.86	1.57%	✅
dlrm-criteoterabyte	1	32.70	31.56	3.61%	✅
dlrm-criteoterabyte_fp16	1	51.98	50.03	3.92%	✅
agentmodel	1	11,618.03	10,347.89	12.27%	🔆
unet_fp16	2	56.22	55.66	1.01%	✅
resnet50v1_fp16	1	962.35	994.78	-3.26%	✅
resnet50v1_int8	1	926.81	943.47	-1.77%	✅
bert_base_cased_fp16	64	1,111.39	1,086.36	2.30%	✅
bert_large_uncased_fp16	32	345.47	338.39	2.09%	✅
bert_large_fp16	1	200.10	198.67	0.72%	✅
distilgpt2_fp16	16	2,103.80	2,066.44	1.81%	✅
yolov5s	1	553.28	549.43	0.70%	✅
tinyllama	1	45.44	43.80	3.75%	✅
vicuna-fastchat	1	43.66	42.63	2.43%	✅
whisper-tiny-encoder	1	413.58	406.13	1.83%	✅
whisper-tiny-decoder	1	406.92	395.21	2.96%	✅
llama2_7b	1	20.22	19.10	5.89%	🔆
qwen1.5-7b	1	23.24	23.42	-0.77%	✅
phi3-3.8b	1	26.41	26.63	-0.82%	✅
llama3-8b	1	21.51	21.68	-0.74%	✅
whisper-large-encoder	1	10.22	10.12	0.99%	✅
whisper-large-decoder	1	102.07	101.03	1.03%	✅
mistral-7b	1	23.80	23.64	0.70%	✅
FLUX.1-schnell	1	752.17	737.25	2.02%	✅

Check flagged results 🔆

[causten]

Test	Status	Result
bert-mrpc-onnx	✅	PASSED: MIGraphX meets tolerance
bert-mrpc-tf	✅	PASSED: MIGraphX meets tolerance
pytorch-examples-wlang-gru	✅	PASSED: MIGraphX meets tolerance
pytorch-examples-wlang-lstm	✅	PASSED: MIGraphX meets tolerance
dlrm-criteoterabyte	✅	PASSED: MIGraphX meets tolerance
agentmodel	✅	PASSED: MIGraphX meets tolerance
unet	✅	PASSED: MIGraphX meets tolerance
resnet50v1	✅	PASSED: MIGraphX meets tolerance
bert_base_cased_fp16	✅	PASSED: MIGraphX meets tolerance
bert_large_uncased_fp16	🔴	FAILED: MIGraphX is not within tolerance - check verbose output
bert_large	✅	PASSED: MIGraphX meets tolerance
yolov5s	✅	PASSED: MIGraphX meets tolerance
tinyllama	✅	PASSED: MIGraphX meets tolerance
vicuna-fastchat	✅	PASSED: MIGraphX meets tolerance
whisper-tiny-encoder	✅	PASSED: MIGraphX meets tolerance
whisper-tiny-decoder	✅	PASSED: MIGraphX meets tolerance
distilgpt2_fp16	✅	PASSED: MIGraphX meets tolerance
llama2_7b	✅	PASSED: MIGraphX meets tolerance
qwen1.5-7b	✅	PASSED: MIGraphX meets tolerance
phi3-3.8b	✅	PASSED: MIGraphX meets tolerance
llama3-8b	✅	PASSED: MIGraphX meets tolerance
whisper-large-encoder	❌	ERROR - check error output traceback Traceback (most recent call last): File "/src/AMDMIGraphX/tools/accuracy/accuracy_checker.py", line 377, in main() File "/src/AMDMIGraphX/tools/accuracy/accuracy_checker.py", line 224, in main model = migraphx.parse_onnx(model_name, default_dim_value=batch) RuntimeError: /src/AMDMIGraphX/src/include/migraphx/op/convolution.hpp:103: normalize_compute_shape: CONVOLUTION: mismatched channel numbers
whisper-large-decoder	✅	PASSED: MIGraphX meets tolerance
mistral-7b	✅	PASSED: MIGraphX meets tolerance
FLUX.1-schnell	✅	PASSED: MIGraphX meets tolerance

[causten]

The distilgpt2_fp16 accuracy check is a real accuracy failure. Can you review your PR