Enable attention by default on mi300

[pfultz2]

No description provided.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 92.17%. Comparing base (275f854) to head (3ce3bd0).
Report is 5 commits behind head on develop.

Additional details and impacted files

@@           Coverage Diff            @@
##           develop    #3463   +/-   ##
========================================
  Coverage    92.17%   92.17%           
========================================
  Files          512      512           
  Lines        21391    21391           
========================================
  Hits         19718    19718           
  Misses        1673     1673           

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

[migraphx-bot]

Test	Batch	Rate new 3ce3bd	Rate old ddc4c0	Diff	Compare
torchvision-resnet50	64	3,260.66	3,261.96	-0.04%	✅
torchvision-resnet50_fp16	64	6,994.71	6,986.50	0.12%	✅
torchvision-densenet121	32	2,435.87	2,435.05	0.03%	✅
torchvision-densenet121_fp16	32	4,099.37	4,098.64	0.02%	✅
torchvision-inceptionv3	32	1,639.16	1,637.72	0.09%	✅
torchvision-inceptionv3_fp16	32	2,764.55	2,760.21	0.16%	✅
cadene-inceptionv4	16	776.36	776.83	-0.06%	✅
cadene-resnext64x4	16	812.25	812.01	0.03%	✅
slim-mobilenet	64	7,537.45	7,539.31	-0.02%	✅
slim-nasnetalarge	64	211.56	211.58	-0.01%	✅
slim-resnet50v2	64	3,505.58	3,507.45	-0.05%	✅
bert-mrpc-onnx	8	1,147.39	1,148.38	-0.09%	✅
bert-mrpc-tf	1	465.83	462.39	0.74%	✅
pytorch-examples-wlang-gru	1	418.60	416.01	0.62%	✅
pytorch-examples-wlang-lstm	1	438.88	387.25	13.33%	🔆
torchvision-resnet50_1	1	774.28	799.30	-3.13%	🔴
cadene-dpn92_1	1	433.58	413.07	4.97%	🔆
cadene-resnext101_1	1	383.12	384.54	-0.37%	✅
onnx-taau-downsample	1	342.87	343.19	-0.09%	✅
dlrm-criteoterabyte	1	33.34	33.33	0.02%	✅
dlrm-criteoterabyte_fp16	1	52.77	52.75	0.04%	✅
agentmodel	1	8,200.22	8,254.64	-0.66%	✅
unet_fp16	2	58.93	58.74	0.33%	✅
resnet50v1_fp16	1	928.26	942.04	-1.46%	✅
resnet50v1_int8	1	1,015.86	993.52	2.25%	✅
bert_base_cased_fp16	64	1,170.51	1,171.05	-0.05%	✅
bert_large_uncased_fp16	32	363.63	363.48	0.04%	✅
bert_large_fp16	1	198.96	200.70	-0.87%	✅
distilgpt2_fp16	16	2,198.54	2,201.26	-0.12%	✅
yolov5s	1	543.79	537.45	1.18%	✅
tinyllama	1	43.45	43.68	-0.53%	✅
vicuna-fastchat	1	175.83	175.99	-0.09%	✅
whisper-tiny-encoder	1	418.06	418.69	-0.15%	✅
whisper-tiny-decoder	1	424.54	428.55	-0.93%	✅

This build is not recommended to 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