catch python buffer unsupported types

[shivadbhavsar]

Issue:
New datatypes (fp8 and bf16) are not natively supported by the python buffer protocol. See https://docs.python.org/3/library/struct.html#format-characters

This can lead to some silly behavior since all unsupported types seem to give format string "z". Example:

import migraphx
p = migraphx.program()
mm = p.get_main_module()
s = migraphx.shape(type="bf16_type", lens=[16])
a = migraphx.generate_argument(s) 
l = mm.add_literal(a)
print(a.get_shape()) # bf16_type, {16}, {1}
print(l.shape()) # fp8e5m2fnuz_type, {16}, {2} ??????????

Fix:

1. explicitly error out if the python buffer format string is "z".
2. add add_literal_from_argument method that allows the user to directly add an argument to some module they may or may not be a recognized type by py buffer protocol.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 92.23%. Comparing base (0860461) to head (fb0500a).
Report is 1 commits behind head on develop.

Additional details and impacted files

@@           Coverage Diff            @@
##           develop    #3701   +/-   ##
========================================
  Coverage    92.23%   92.23%           
========================================
  Files          514      514           
  Lines        21746    21746           
========================================
  Hits         20057    20057           
  Misses        1689     1689           

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

[migraphx-bot]

Test	Batch	Rate new fb0500	Rate old 086046	Diff	Compare
torchvision-resnet50	64	3,257.31	3,255.59	0.05%	✅
torchvision-resnet50_fp16	64	6,991.53	6,973.76	0.25%	✅
torchvision-densenet121	32	2,434.46	2,432.94	0.06%	✅
torchvision-densenet121_fp16	32	4,046.86	4,068.82	-0.54%	✅
torchvision-inceptionv3	32	1,627.31	1,627.80	-0.03%	✅
torchvision-inceptionv3_fp16	32	2,745.70	2,748.46	-0.10%	✅
cadene-inceptionv4	16	765.46	764.89	0.07%	✅
cadene-resnext64x4	16	812.00	813.14	-0.14%	✅
slim-mobilenet	64	7,460.36	7,464.91	-0.06%	✅
slim-nasnetalarge	64	209.02	208.99	0.01%	✅
slim-resnet50v2	64	3,441.25	3,438.12	0.09%	✅
bert-mrpc-onnx	8	1,145.53	1,144.42	0.10%	✅
bert-mrpc-tf	1	479.26	469.74	2.03%	✅
pytorch-examples-wlang-gru	1	415.68	423.59	-1.87%	✅
pytorch-examples-wlang-lstm	1	397.75	378.46	5.10%	🔆
torchvision-resnet50_1	1	814.68	753.15	8.17%	🔆
cadene-dpn92_1	1	398.54	434.07	-8.18%	🔴
cadene-resnext101_1	1	382.82	381.61	0.32%	✅
onnx-taau-downsample	1	345.38	345.60	-0.06%	✅
dlrm-criteoterabyte	1	33.33	33.33	-0.00%	✅
dlrm-criteoterabyte_fp16	1	52.75	52.71	0.07%	✅
agentmodel	1	8,258.39	8,466.76	-2.46%	✅
unet_fp16	2	58.74	58.76	-0.02%	✅
resnet50v1_fp16	1	935.49	923.27	1.32%	✅
resnet50v1_int8	1	1,017.41	1,005.45	1.19%	✅
bert_base_cased_fp16	64	1,169.08	1,168.88	0.02%	✅
bert_large_uncased_fp16	32	362.89	363.10	-0.06%	✅
bert_large_fp16	1	198.52	198.60	-0.05%	✅
distilgpt2_fp16	16	2,195.64	2,199.62	-0.18%	✅
yolov5s	1	528.08	528.00	0.02%	✅
tinyllama	1	43.60	43.67	-0.15%	✅
vicuna-fastchat	1	174.17	171.38	1.63%	✅
whisper-tiny-encoder	1	417.56	417.67	-0.03%	✅
whisper-tiny-decoder	1	427.17	428.88	-0.40%	✅

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