Enable standalone convolution for mlir by default

[pfultz2]

Only use miopen for windograd.

[umangyadav]

Only use miopen for windograd.

There are still some checks here.
https://github.com/ROCmSoftwarePlatform/AMDMIGraphX/blob/eafd55de0b4bebc26cd090c0d10a3b5d739d6a56/src/targets/gpu/fuse_mlir.cpp#L210

Therefore description looks misleading a bit. It would still call MIOpen for more cases in addition to winograd.

[migraphx-bot]

Test	Batch	Rate new 872f72	Rate old b381cb	Diff	Compare
torchvision-resnet50	64	2,829.73	2,830.27	-0.02%	✅
torchvision-resnet50_fp16	64	6,500.01	6,500.13	-0.00%	✅
torchvision-densenet121	32	2,087.52	2,096.37	-0.42%	✅
torchvision-densenet121_fp16	32	3,661.32	3,666.22	-0.13%	✅
torchvision-inceptionv3	32	1,598.19	1,584.51	0.86%	✅
torchvision-inceptionv3_fp16	32	2,575.35	2,570.45	0.19%	✅
cadene-inceptionv4	16	724.38	692.71	4.57%	🔆
cadene-resnext64x4	16	688.87	691.65	-0.40%	✅
slim-mobilenet	64	8,339.38	8,332.86	0.08%	✅
slim-nasnetalarge	64	230.52	225.39	2.27%	✅
slim-resnet50v2	64	2,665.68	2,650.43	0.58%	✅
bert-mrpc-onnx	8	823.19	810.59	1.55%	✅
bert-mrpc-tf	1	390.41	385.17	1.36%	✅
pytorch-examples-wlang-gru	1	299.59	295.61	1.35%	✅
pytorch-examples-wlang-lstm	1	312.94	302.11	3.59%	🔆
torchvision-resnet50_1	1	607.14	592.11	2.54%	✅
torchvision-inceptionv3_1	1	342.87	339.33	1.04%	✅
cadene-dpn92_1	1	404.69	377.19	7.29%	🔆
cadene-resnext101_1	1	328.47	328.14	0.10%	✅
slim-vgg16_1	1	458.34	454.87	0.76%	✅
slim-mobilenet_1	1	2,124.76	2,020.85	5.14%	🔆
slim-inceptionv4_1	1	213.37	214.06	-0.32%	✅
onnx-taau-downsample	1	304.79	304.11	0.22%	✅
dlrm-criteoterabyte	1	21.61	21.62	-0.03%	✅
dlrm-criteoterabyte_fp16	1	40.65	40.17	1.20%	✅
agentmodel	1	5,889.94	4,540.21	29.73%	🔆
unet_fp16	2	54.77	54.68	0.17%	✅
resnet50v1_fp16	1	933.56	952.82	-2.02%	✅
bert_base_cased_fp16	64	903.13	901.06	0.23%	✅
bert_large_uncased_fp16	32	285.59	191.55	49.09%	🔆
bert_large_fp16	1	166.73	166.45	0.17%	✅
distilgpt2_fp16	16	1,281.08	1,278.49	0.20%	✅

Check results before merge 🔆

[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:torchvision-inceptionv3_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:slim-vgg16_1: PASSED: MIGraphX meets tolerance

    :white_check_mark:slim-mobilenet_1: PASSED: MIGraphX meets tolerance

    :white_check_mark:slim-inceptionv4_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

🔴bert_base_cased_fp16: FAILED: MIGraphX is not within tolerance - check verbose output

    :white_check_mark:bert_large_uncased_fp16: PASSED: MIGraphX meets tolerance

    :white_check_mark:bert_large: PASSED: MIGraphX meets tolerance

🔴distilgpt2_fp16: FAILED: MIGraphX is not within tolerance - check verbose output