Fix: Driver --batch option sets Window Dimensions.

[lakhinderwalia]

[causten]

Take a look at the models that are failing in CI. You likely have caught some input parameter assumptions.

[codecov[bot]]

Codecov Report

Attention: Patch coverage is 95.45455% with 1 line in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
src/onnx/onnx_parser.cpp	93.75%	1 Missing :warning:

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #3770      +/-   ##
===========================================
- Coverage    92.41%   92.41%   -0.00%     
===========================================
  Files          520      520              
  Lines        22485    22494       +9     
===========================================
+ Hits         20778    20786       +8     
- Misses        1707     1708       +1     

Files with missing lines	Coverage Δ
src/onnx/include/migraphx/onnx/onnx_parser.hpp	100.00% <ø> (ø)
src/onnx/onnx.cpp	96.77% <100.00%> (+0.22%)	:arrow_up:
src/program.cpp	70.25% <100.00%> (+0.05%)	:arrow_up:
src/onnx/onnx_parser.cpp	89.20% <93.75%> (-0.13%)	:arrow_down:

:rocket: New features to boost your workflow:

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

[lakhinderwalia]

vicuna-fastchat model is failing due to its unspecified dynamic dimensions for input_ids: vicuna/encoder_model.onnx --exhaustive-tune --fill1 input_ids --input-dim @input_ids. Also, attention_mask should be specified. Reference: https://github.com/ROCm/AMDMIGraphX/pull/3770#issuecomment-2609723783

[lakhinderwalia]

These tests pass the perf tests above, but not the other test listed below, in the same check-in, which is strange.

Test Batch Rate new d25a96 Rate old 5dc019 Diff Compare dlrm-criteoterabyte 1 30.55 30.53 0.07% ✅ dlrm-criteoterabyte_fp16 1 49.13 49.09 0.07% ✅ vicuna-fastchat 1 43.77 43.81 -0.10% ✅

❌dlrm-criteoterabyte: ERROR - check error output ❌vicuna-fastchat: ERROR - check error output

[lakhinderwalia]

These tests pass the perf tests above, but not the other test listed below, in the same check-in, which is strange.

Test Batch Rate new d25a96 Rate old 5dc019 Diff Compare dlrm-criteoterabyte 1 30.55 30.53 0.07% ✅ dlrm-criteoterabyte_fp16 1 49.13 49.09 0.07% ✅ vicuna-fastchat 1 43.77 43.81 -0.10% ✅

❌dlrm-criteoterabyte: ERROR - check error output ❌vicuna-fastchat: ERROR - check error output

Test scripts are updated.

[migraphx-bot]

Test	Batch	Rate new 9acc38	Rate old ea9d54	Diff	Compare
torchvision-resnet50	64	3,237.87	3,237.03	0.03%	:white_check_mark:
torchvision-resnet50_fp16	64	6,881.86	6,880.50	0.02%	:white_check_mark:
torchvision-densenet121	32	2,436.54	2,438.10	-0.06%	:white_check_mark:
torchvision-densenet121_fp16	32	4,186.41	4,201.36	-0.36%	:white_check_mark:
torchvision-inceptionv3	32	1,615.22	1,613.90	0.08%	:white_check_mark:
torchvision-inceptionv3_fp16	32	2,678.79	2,679.12	-0.01%	:white_check_mark:
cadene-inceptionv4	16	751.12	750.97	0.02%	:white_check_mark:
cadene-resnext64x4	16	809.57	809.24	0.04%	:white_check_mark:
slim-mobilenet	64	6,665.86	6,663.22	0.04%	:white_check_mark:
slim-nasnetalarge	64	198.59	198.52	0.03%	:white_check_mark:
slim-resnet50v2	64	3,427.34	3,426.77	0.02%	:white_check_mark:
bert-mrpc-onnx	8	1,139.76	1,139.70	0.01%	:white_check_mark:
bert-mrpc-tf	1	484.35	484.39	-0.01%	:white_check_mark:
pytorch-examples-wlang-gru	1	483.58	477.62	1.25%	:white_check_mark:
pytorch-examples-wlang-lstm	1	452.01	439.47	2.85%	:white_check_mark:
torchvision-resnet50_1	1	807.87	807.21	0.08%	:white_check_mark:
cadene-dpn92_1	1	431.14	428.04	0.72%	:white_check_mark:
cadene-resnext101_1	1	390.77	391.14	-0.10%	:white_check_mark:
onnx-taau-downsample	1	372.23	372.67	-0.12%	:white_check_mark:
dlrm-criteoterabyte	1	31.81	31.79	0.06%	:white_check_mark:
dlrm-criteoterabyte_fp16	1	50.95	50.93	0.03%	:white_check_mark:
agentmodel	1	8,760.79	8,853.71	-1.05%	:white_check_mark:
unet_fp16	2	58.27	58.35	-0.14%	:white_check_mark:
resnet50v1_fp16	1	1,042.67	1,050.98	-0.79%	:white_check_mark:
resnet50v1_int8	1	807.54	812.20	-0.57%	:white_check_mark:
bert_base_cased_fp16	64	1,164.98	1,164.77	0.02%	:white_check_mark:
bert_large_uncased_fp16	32	361.70	361.75	-0.01%	:white_check_mark:
bert_large_fp16	1	200.84	201.50	-0.33%	:white_check_mark:
distilgpt2_fp16	16	2,217.93	2,218.64	-0.03%	:white_check_mark:
yolov5s	1	533.88	528.92	0.94%	:white_check_mark:
tinyllama	1	43.61	43.62	-0.02%	:white_check_mark:
vicuna-fastchat	1	43.98	43.80	0.39%	:white_check_mark:
whisper-tiny-encoder	1	411.91	412.07	-0.04%	:white_check_mark:
whisper-tiny-decoder	1	410.10	411.63	-0.37%	:white_check_mark:

This build is OK for merge :white_check_mark:

[migraphx-bot]

     :white_check_mark: bert-mrpc-onnx: PASSED: MIGraphX meets tolerance

     :white_check_mark: bert-mrpc-tf: PASSED: MIGraphX meets tolerance

     :white_check_mark: pytorch-examples-wlang-gru: PASSED: MIGraphX meets tolerance

     :white_check_mark: pytorch-examples-wlang-lstm: PASSED: MIGraphX meets tolerance

     :white_check_mark: torchvision-resnet50_1: PASSED: MIGraphX meets tolerance

     :white_check_mark: cadene-dpn92_1: PASSED: MIGraphX meets tolerance

     :white_check_mark: cadene-resnext101_1: PASSED: MIGraphX meets tolerance

     :white_check_mark: dlrm-criteoterabyte: PASSED: MIGraphX meets tolerance

     :white_check_mark: agentmodel: PASSED: MIGraphX meets tolerance

     :white_check_mark: unet: PASSED: MIGraphX meets tolerance

     :white_check_mark: resnet50v1: PASSED: MIGraphX meets tolerance

     :white_check_mark: bert_base_cased_fp16: PASSED: MIGraphX meets tolerance

:red_circle:bert_large_uncased_fp16: FAILED: MIGraphX is not within tolerance - check verbose output

     :white_check_mark: bert_large: PASSED: MIGraphX meets tolerance

     :white_check_mark: yolov5s: PASSED: MIGraphX meets tolerance

     :white_check_mark: tinyllama: PASSED: MIGraphX meets tolerance

     :white_check_mark: vicuna-fastchat: PASSED: MIGraphX meets tolerance

     :white_check_mark: whisper-tiny-encoder: PASSED: MIGraphX meets tolerance

     :white_check_mark: whisper-tiny-decoder: PASSED: MIGraphX meets tolerance

     :white_check_mark: distilgpt2_fp16: PASSED: MIGraphX meets tolerance

[jayhawk-commits]

/AzurePipelines run

[azure-pipelines[bot]]

Azure Pipelines successfully started running 1 pipeline(s).

[pfultz2]

The driver is for developers. So the batch is not a priority to fix.