Add support for the Einsum operator

[mirza-halilcevic]

• Implement ONNX parsing support for the Einsum operator
• Resolves https://github.com/migraphx-benchmark/AMDMIGraphX/issues/92

[migraphx-bot]

Test	Batch	Rate new 7140ae	Rate old dc028d	Diff	Compare
torchvision-resnet50	64	1,713.92	1,489.70	15.05%	:high_brightness:
torchvision-resnet50_fp16	64	3,810.84	1,346.10	183.10%	:high_brightness:
torchvision-densenet121	32	1,453.97	1,440.50	0.94%	:white_check_mark:
torchvision-densenet121_fp16	32	2,435.42	2,416.72	0.77%	:white_check_mark:
torchvision-inceptionv3	32	883.32	881.27	0.23%	:white_check_mark:
torchvision-inceptionv3_fp16	32	1,412.36	1,406.66	0.41%	:white_check_mark:
cadene-inceptionv4	16	407.47	404.25	0.80%	:white_check_mark:
cadene-resnext64x4	16	413.99	410.23	0.92%	:white_check_mark:
slim-mobilenet	64	3,823.89	3,794.28	0.78%	:white_check_mark:
slim-nasnetalarge	64	97.06	94.95	2.22%	:white_check_mark:
slim-resnet50v2	64	1,652.06	1,620.87	1.92%	:white_check_mark:
bert-mrpc-onnx	8	589.65	591.10	-0.25%	:white_check_mark:
bert-mrpc-tf	1	289.94	289.30	0.22%	:white_check_mark:
pytorch-examples-wlang-gru	1	333.82	378.53	-11.81%	:red_circle:
pytorch-examples-wlang-lstm	1	303.50	266.28	13.98%	:high_brightness:
torchvision-resnet50_1	1	453.79	369.37	22.86%	:high_brightness:
cadene-dpn92_1	1	245.35	233.66	5.00%	:high_brightness:
cadene-resnext101_1	1	189.30	189.16	0.07%	:white_check_mark:
onnx-taau-downsample	1	204.05	183.13	11.43%	:high_brightness:
dlrm-criteoterabyte	1	22.29	21.99	1.38%	:white_check_mark:
dlrm-criteoterabyte_fp16	1	41.63	41.43	0.47%	:white_check_mark:
agentmodel	1	6,211.60	6,337.70	-1.99%	:white_check_mark:
unet_fp16	2	33.74	33.63	0.33%	:white_check_mark:
resnet50v1_fp16	1	557.63	521.53	6.92%	:high_brightness:
resnet50v1_int8	1	463.51	452.53	2.43%	:white_check_mark:
bert_base_cased_fp16	64	620.97	620.67	0.05%	:white_check_mark:
bert_large_uncased_fp16	32	193.84	193.85	-0.01%	:white_check_mark:
bert_large_fp16	1	103.92	103.88	0.04%	:white_check_mark:
distilgpt2_fp16	16	1,187.47	1,187.83	-0.03%	:white_check_mark:
yolov5s	1	298.21	297.39	0.27%	:white_check_mark:
tinyllama	1	23.34	23.34	0.02%	:white_check_mark:
vicuna-fastchat	1	132.42	132.19	0.17%	:white_check_mark:
whisper-tiny-encoder	1	240.97	240.52	0.19%	:white_check_mark:
whisper-tiny-decoder	1	245.84	245.42	0.17%	:white_check_mark:

This build is not recommended to merge :red_circle:

[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

[codecov[bot]]

Codecov Report

Attention: Patch coverage is 98.66667% with 4 lines in your changes missing coverage. Please review.

Project coverage is 91.92%. Comparing base (dc028dd) to head (7140ae2). Report is 148 commits behind head on develop.

Files	Patch %	Lines
src/onnx/parse_einsum.cpp	98.91%	3 Missing :warning:
src/lexing.cpp	93.75%	1 Missing :warning:

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #2772      +/-   ##
===========================================
+ Coverage    91.82%   91.92%   +0.10%     
===========================================
  Files          486      489       +3     
  Lines        18993    19275     +282     
===========================================
+ Hits         17440    17719     +279     
- Misses        1553     1556       +3     

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[causten]

Windows CI is failing. Seems like... src\onnx\parse_einsum.cpp:516:21: error: use of undeclared identifier 'ssize_t'

[mirza-halilcevic]

Windows CI is failing. Seems like... src\onnx\parse_einsum.cpp:516:21: error: use of undeclared identifier 'ssize_t'

I replaced ssize_t with int64_t. Seems like it's still failing for some other reason. I don't have access to the logs though.

[causten]

I see the extended Jenkins tests run but they are not reflecting on this PR list. They all passed so I'm merging