Peephole LSTM Error

[richagadgil]

The peephole is causing this error in line 1095

auto spph = m.insert_instruction(ins, make_op("squeeze", {{"axes", {0}}}), pph);

Pass: rewrite_rnn
terminate called after throwing an instance of 'migraphx::version_2_12_0::exception'
  what():  /code/AMDMIGraphX/src/normalize_attributes.cpp:120: tune_attribute: squeeze: axes: value out of range!
Aborted (core dumped)

Printing out the pph instruction.get_shape() value gives me: @literal half_type, {}, {}

Quantization is causing this to be an empty parameter instead of undefined.

[codecov]

Codecov Report

Attention: Patch coverage is 50.00000% with 1 line in your changes missing coverage. Please review.

Project coverage is 92.29%. Comparing base (810b3f2) to head (7bc259c).

Files with missing lines	Patch %	Lines
src/propagate_constant.cpp	50.00%	1 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #3768      +/-   ##
===========================================
- Coverage    92.29%   92.29%   -0.01%     
===========================================
  Files          519      519              
  Lines        22217    22219       +2     
===========================================
+ Hits         20505    20506       +1     
- Misses        1712     1713       +1     

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

[migraphx-bot]

Test	Batch	Rate new a9919e	Rate old 43488c	Diff	Compare
torchvision-resnet50	64	3,253.77	3,254.25	-0.01%	✅
torchvision-resnet50_fp16	64	6,920.21	6,916.65	0.05%	✅
torchvision-densenet121	32	2,455.98	2,455.95	0.00%	✅
torchvision-densenet121_fp16	32	4,171.47	4,183.26	-0.28%	✅
torchvision-inceptionv3	32	1,629.63	1,629.97	-0.02%	✅
torchvision-inceptionv3_fp16	32	2,713.11	2,715.98	-0.11%	✅
cadene-inceptionv4	16	763.10	762.93	0.02%	✅
cadene-resnext64x4	16	812.75	812.98	-0.03%	✅
slim-mobilenet	64	7,456.27	7,459.33	-0.04%	✅
slim-nasnetalarge	64	208.65	208.65	0.00%	✅
slim-resnet50v2	64	3,444.90	3,445.08	-0.01%	✅
bert-mrpc-onnx	8	1,148.07	1,147.48	0.05%	✅
bert-mrpc-tf	1	481.16	481.97	-0.17%	✅
pytorch-examples-wlang-gru	1	481.46	478.37	0.65%	✅
pytorch-examples-wlang-lstm	1	444.23	439.41	1.10%	✅
torchvision-resnet50_1	1	806.56	803.28	0.41%	✅
cadene-dpn92_1	1	425.37	427.49	-0.49%	✅
cadene-resnext101_1	1	385.94	376.23	2.58%	✅
onnx-taau-downsample	1	373.67	373.50	0.05%	✅
dlrm-criteoterabyte	1	33.30	33.31	-0.03%	✅
dlrm-criteoterabyte_fp16	1	52.64	52.64	0.00%	✅
agentmodel	1	8,929.37	8,546.35	4.48%	🔆
unet_fp16	2	58.46	58.43	0.06%	✅
resnet50v1_fp16	1	1,039.87	1,033.48	0.62%	✅
resnet50v1_int8	1	1,020.62	1,018.13	0.24%	✅
bert_base_cased_fp16	64	1,180.91	1,181.00	-0.01%	✅
bert_large_uncased_fp16	32	365.25	365.36	-0.03%	✅
bert_large_fp16	1	200.94	201.35	-0.21%	✅
distilgpt2_fp16	16	2,225.22	2,228.65	-0.15%	✅
yolov5s	1	521.75	527.51	-1.09%	✅
tinyllama	1	43.59	43.62	-0.05%	✅
vicuna-fastchat	1	172.03	173.15	-0.65%	✅
whisper-tiny-encoder	1	418.04	416.80	0.30%	✅
whisper-tiny-decoder	1	431.47	429.87	0.37%	✅

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

[richagadgil]

This fix works to compile the model and run perf, but there is still an issue with the .mxr files once compiled:

FP32:

[ MIGraphX Version: 2.12.0.a9919ec08 ] Complete: /code/AMDMIGraphX/build/bin/migraphx-driver compile model.onnx --input-dim @symbol 1 1 @hidden_in_1 2 1 320 @hidden_in_2 2 1 320 --fill1 symbol -o model_fp32.mxr

[ MIGraphX Version: 2.12.0.a9919ec08 ] Complete: /code/AMDMIGraphX/build/bin/migraphx-driver perf model_fp32.mxr --fill1 symbol

FP16:

[ MIGraphX Version: 2.12.0.a9919ec08 ] Complete: /code/AMDMIGraphX/build/bin/migraphx-driver compile model.onnx --input-dim @symbol 1 1 @hidden_in_1 2 1 320 @hidden_in_2 2 1 320 --fill1 symbol --fp16 -o model_fp16.mxr

Running [ MIGraphX Version: 2.12.0.a9919ec08 ]: /code/AMDMIGraphX/build/bin/migraphx-driver perf model_fp16.mxr --fill1 symbol
Compiling ... 
Reading: model_fp16.mxr
terminate called after throwing an instance of 'msgpack::v1::insufficient_bytes'
  what():  insufficient bytes
Aborted (core dumped)

[richagadgil]

log.txt

MIGRAPHX_TRACE_COMPILE=1 /code/AMDMIGraphX/build/bin/migraphx-driver perf model.onnx --input-dim @symbol 1 1 @hidden_in_1 2 1 320 @hidden_in_2 2 1 320 --fill1 symbol --fp16 -o model_fp16.mxr

[richagadgil]

Changed the PR to skip const prop over undefined instructions instead.