Track broadcast axes in the shape_transform_descriptor

[pfultz2]

Although, this prevents simplifying as much, it does help preserve the permutation of the broadcasted axes.

So if we have a tensor of {2, 16, 10240} that goes into a reduction along the last axis it will output to {2, 16, 1}, which may be broadcasted back into {2, 16, 10240}, but there could be more shape transformations after the reduce but before an pointwise operator:

@1 = multibroadcast[out_lens={2, 16, 10240},out_dyn_dims={}](@0) -> int64_type, {2, 16, 10240}, {16, 1, 0}
@2 = reshape[dims={2, 160, 32, 32}](@1) -> int64_type, {2, 160, 32, 32}, {163840, 1024, 32, 1}
@3 = transpose[permutation={0, 2, 3, 1}](@2) -> int64_type, {2, 32, 32, 160}, {163840, 32, 1, 1024}

On develop this would be simplified to:

@1 = unsqueeze[axes={1, 2, 5},steps={}](@0) -> int64_type, {2, 1, 1, 16, 1, 1}, {16, 16, 16, 1, 1, 1}
@2 = multibroadcast[out_lens={2, 1, 1, 16, 1, 10},out_dyn_dims={}](@1) -> int64_type, {2, 1, 1, 16, 1, 10}, {16, 16, 16, 1, 1, 0}
@3 = reshape[dims={2, 1, 1, 160}](@2) -> int64_type, {2, 1, 1, 160}, {160, 160, 160, 1}
@4 = multibroadcast[out_lens={2, 32, 32, 160},out_dyn_dims={}](@3) -> int64_type, {2, 32, 32, 160}, {160, 0, 0, 1}

Ideally, we would want to apply these transformations without the broadcast before the reduction but if it simplified like above because the shape_transform_descriptor doesnt track the permutation of the the broadcasted axes. With this PR, it will simplify to:

@1 = unsqueeze[axes={3, 4},steps={}](@0) -> int64_type, {2, 16, 1, 1, 1}, {16, 1, 1, 1, 1}
@2 = transpose[permutation={0, 3, 4, 1, 2}](@1) -> int64_type, {2, 1, 1, 16, 1}, {16, 1, 1, 1, 1}
@3 = multibroadcast[out_lens={2, 1, 1, 16, 10},out_dyn_dims={}](@2) -> int64_type, {2, 1, 1, 16, 10}, {16, 1, 1, 1, 0}
@4 = reshape[dims={2, 1, 1, 160}](@3) -> int64_type, {2, 1, 1, 160}, {160, 160, 160, 1}
@5 = multibroadcast[out_lens={2, 32, 32, 160},out_dyn_dims={}](@4) -> int64_type, {2, 32, 32, 160}, {160, 0, 0, 1}

This has a transpose because the shape_transform_descriptor understands how it will output in NHWC, which means we can make the input to the reduction NHWC layout as well. This PR doesn't enable such rewriting, it only modifies the shape_transform descriptor to track such layouts.

Also, there is some updates to the tests as well:

• Validate that a simplified transformation produces the same result
• Check that the simplification cannot be simplified further

[TedThemistokleous]

Fix CI but otherwise looks good

[codecov]

Codecov Report

Attention: Patch coverage is 96.03960% with 4 lines in your changes missing coverage. Please review.

Project coverage is 92.23%. Comparing base (241e24e) to head (64a50f7).
Report is 5 commits behind head on develop.

Files with missing lines	Patch %	Lines
src/shape_transform_descriptor.cpp	96.03%	4 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #3610      +/-   ##
===========================================
+ Coverage    92.20%   92.23%   +0.02%     
===========================================
  Files          513      513              
  Lines        21658    21726      +68     
===========================================
+ Hits         19970    20038      +68     
  Misses        1688     1688              

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

[migraphx-bot]

Test	Batch	Rate new 64a50f	Rate old 241e24	Diff	Compare
torchvision-resnet50	64	3,255.03	3,253.39	0.05%	✅
torchvision-resnet50_fp16	64	6,989.80	6,971.83	0.26%	✅
torchvision-densenet121	32	2,431.22	2,432.57	-0.06%	✅
torchvision-densenet121_fp16	32	4,092.43	4,102.36	-0.24%	✅
torchvision-inceptionv3	32	1,628.97	1,627.63	0.08%	✅
torchvision-inceptionv3_fp16	32	2,750.44	2,741.25	0.34%	✅
cadene-inceptionv4	16	765.03	763.96	0.14%	✅
cadene-resnext64x4	16	810.16	809.72	0.05%	✅
slim-mobilenet	64	7,462.81	7,381.84	1.10%	✅
slim-nasnetalarge	64	208.47	208.40	0.03%	✅
slim-resnet50v2	64	3,435.87	3,439.64	-0.11%	✅
bert-mrpc-onnx	8	1,143.63	1,145.80	-0.19%	✅
bert-mrpc-tf	1	460.49	492.93	-6.58%	🔴
pytorch-examples-wlang-gru	1	527.91	426.82	23.69%	🔆
pytorch-examples-wlang-lstm	1	401.55	392.81	2.23%	✅
torchvision-resnet50_1	1	787.71	777.08	1.37%	✅
cadene-dpn92_1	1	402.67	429.78	-6.31%	🔴
cadene-resnext101_1	1	382.29	383.15	-0.23%	✅
onnx-taau-downsample	1	345.95	346.23	-0.08%	✅
dlrm-criteoterabyte	1	33.33	33.33	-0.01%	✅
dlrm-criteoterabyte_fp16	1	52.76	52.75	0.02%	✅
agentmodel	1	8,359.46	8,083.75	3.41%	🔆
unet_fp16	2	58.91	58.78	0.22%	✅
resnet50v1_fp16	1	1,020.82	943.59	8.18%	🔆
resnet50v1_int8	1	1,008.13	1,004.54	0.36%	✅
bert_base_cased_fp16	64	1,169.80	1,169.61	0.02%	✅
bert_large_uncased_fp16	32	363.53	363.60	-0.02%	✅
bert_large_fp16	1	199.20	200.33	-0.56%	✅
distilgpt2_fp16	16	2,202.65	2,196.55	0.28%	✅
yolov5s	1	540.70	526.38	2.72%	✅
tinyllama	1	43.63	43.64	-0.02%	✅
vicuna-fastchat	1	174.80	173.50	0.75%	✅
whisper-tiny-encoder	1	417.72	417.83	-0.03%	✅
whisper-tiny-decoder	1	427.53	427.29	0.06%	✅

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