Migraphx support for gfx12

[aarushjain29]

rocBLAS error: Cannot read /opt/rocm/lib/rocblas/library/TensileLibrary.dat: Illegal seek for GPU arch : gfx1201
List of available TensileLibrary Files :
"/opt/rocm/lib/rocblas/library/TensileLibrary_lazy_gfx908.dat"
"/opt/rocm/lib/rocblas/library/TensileLibrary_lazy_gfx1102.dat"
"/opt/rocm/lib/rocblas/library/TensileLibrary_lazy_gfx906.dat"
"/opt/rocm/lib/rocblas/library/TensileLibrary_lazy_gfx1010.dat"
"/opt/rocm/lib/rocblas/library/TensileLibrary_lazy_gfx90a.dat"
"/opt/rocm/lib/rocblas/library/TensileLibrary_lazy_gfx1100.dat"
"/opt/rocm/lib/rocblas/library/TensileLibrary_lazy_gfx900.dat"
"/opt/rocm/lib/rocblas/library/TensileLibrary_lazy_gfx1101.dat"
"/opt/rocm/lib/rocblas/library/TensileLibrary_lazy_gfx1012.dat"
"/opt/rocm/lib/rocblas/library/TensileLibrary_lazy_gfx1030.dat"
"/opt/rocm/lib/rocblas/library/TensileLibrary_lazy_gfx942.dat"
"/opt/rocm/lib/rocblas/library/TensileLibrary_lazy_gfx941.dat"
"/opt/rocm/lib/rocblas/library/TensileLibrary_lazy_gfx940.dat"

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 92.16%. Comparing base (0fb79b4) to head (49152cb).
Report is 132 commits behind head on develop.

Additional details and impacted files

@@           Coverage Diff            @@
##           develop    #3517   +/-   ##
========================================
  Coverage    92.16%   92.16%           
========================================
  Files          512      512           
  Lines        21401    21401           
========================================
  Hits         19724    19724           
  Misses        1677     1677           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[ahsan-ca]

LGTM besides formatting.

[migraphx-bot]

Test	Batch	Rate new 49152c	Rate old b73def	Diff	Compare
torchvision-resnet50	64	3,260.90	3,257.93	0.09%	✅
torchvision-resnet50_fp16	64	6,986.88	6,992.99	-0.09%	✅
torchvision-densenet121	32	2,439.15	2,432.26	0.28%	✅
torchvision-densenet121_fp16	32	4,086.78	4,038.39	1.20%	✅
torchvision-inceptionv3	32	1,640.38	1,638.89	0.09%	✅
torchvision-inceptionv3_fp16	32	2,760.29	2,761.69	-0.05%	✅
cadene-inceptionv4	16	776.08	776.39	-0.04%	✅
cadene-resnext64x4	16	808.08	811.37	-0.41%	✅
slim-mobilenet	64	7,534.77	7,532.73	0.03%	✅
slim-nasnetalarge	64	211.52	211.42	0.05%	✅
slim-resnet50v2	64	3,506.02	3,507.25	-0.04%	✅
bert-mrpc-onnx	8	1,150.45	1,147.76	0.23%	✅
bert-mrpc-tf	1	467.70	469.91	-0.47%	✅
pytorch-examples-wlang-gru	1	514.97	514.96	0.00%	✅
pytorch-examples-wlang-lstm	1	376.05	386.61	-2.73%	✅
torchvision-resnet50_1	1	768.67	772.05	-0.44%	✅
cadene-dpn92_1	1	397.79	398.73	-0.24%	✅
cadene-resnext101_1	1	383.40	383.67	-0.07%	✅
onnx-taau-downsample	1	343.17	342.33	0.25%	✅
dlrm-criteoterabyte	1	33.36	33.33	0.09%	✅
dlrm-criteoterabyte_fp16	1	52.74	52.70	0.08%	✅
agentmodel	1	8,478.53	8,056.20	5.24%	🔆
unet_fp16	2	58.94	58.92	0.03%	✅
resnet50v1_fp16	1	937.46	950.32	-1.35%	✅
resnet50v1_int8	1	1,013.15	1,000.02	1.31%	✅
bert_base_cased_fp16	64	1,171.39	1,169.24	0.18%	✅
bert_large_uncased_fp16	32	363.89	363.69	0.06%	✅
bert_large_fp16	1	200.53	198.89	0.82%	✅
distilgpt2_fp16	16	2,202.16	2,203.09	-0.04%	✅
yolov5s	1	530.44	540.85	-1.92%	✅
tinyllama	1	43.43	43.43	-0.00%	✅
vicuna-fastchat	1	176.99	170.64	3.73%	🔆
whisper-tiny-encoder	1	418.28	418.21	0.02%	✅
whisper-tiny-decoder	1	428.27	426.10	0.51%	✅

Check results before merge 🔆

[migraphx-bot]

     ✅ 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

     ✅ torchvision-resnet50_1: PASSED: MIGraphX meets tolerance

     ✅ cadene-dpn92_1: PASSED: MIGraphX meets tolerance

     ✅ cadene-resnext101_1: 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