Driver quantize fp8 update

[CharlieL7]

• Updates the quantization to always quantize fp8 to the OCP fp8e4m3fn type
• Removes running simplify_qdq and optimize_module during the quantization so that the ocp_to_fnuz conversion pass can work properly
• Don't merge this until FP8 OCP to FP8 FNUZ on hardware with only FP8 FNUZ support #3684 is done.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 92.29%. Comparing base (976ae75) to head (9fc2e97).
Report is 3 commits behind head on develop.

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #3715      +/-   ##
===========================================
+ Coverage    92.28%   92.29%   +0.01%     
===========================================
  Files          519      519              
  Lines        22222    22216       -6     
===========================================
- Hits         20507    20504       -3     
+ Misses        1715     1712       -3     

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

[pfultz2]

Why is this removed on serialization?

[CharlieL7]

I added this earlier when getting FP8 OCP in to query the gfx number from the driver. We could keep it, but it would not be used anywhere anymore.

[migraphx-bot]

Test	Batch	Rate new 9fc2e9	Rate old f36eba	Diff	Compare
torchvision-resnet50	64	3,251.40	3,253.72	-0.07%	✅
torchvision-resnet50_fp16	64	6,915.73	6,920.21	-0.06%	✅
torchvision-densenet121	32	2,452.99	2,455.81	-0.11%	✅
torchvision-densenet121_fp16	32	4,158.21	4,158.69	-0.01%	✅
torchvision-inceptionv3	32	1,630.71	1,630.15	0.03%	✅
torchvision-inceptionv3_fp16	32	2,716.42	2,714.74	0.06%	✅
cadene-inceptionv4	16	763.47	762.38	0.14%	✅
cadene-resnext64x4	16	813.01	812.83	0.02%	✅
slim-mobilenet	64	7,461.91	7,460.11	0.02%	✅
slim-nasnetalarge	64	208.59	208.61	-0.01%	✅
slim-resnet50v2	64	3,448.16	3,446.57	0.05%	✅
bert-mrpc-onnx	8	1,149.10	1,148.57	0.05%	✅
bert-mrpc-tf	1	485.55	486.00	-0.09%	✅
pytorch-examples-wlang-gru	1	483.78	512.26	-5.56%	🔴
pytorch-examples-wlang-lstm	1	442.47	489.70	-9.65%	🔴
torchvision-resnet50_1	1	813.52	808.03	0.68%	✅
cadene-dpn92_1	1	424.15	440.63	-3.74%	🔴
cadene-resnext101_1	1	383.83	385.04	-0.31%	✅
onnx-taau-downsample	1	373.00	373.31	-0.08%	✅
dlrm-criteoterabyte	1	33.30	33.30	-0.00%	✅
dlrm-criteoterabyte_fp16	1	52.70	52.42	0.53%	✅
agentmodel	1	8,672.50	8,664.84	0.09%	✅
unet_fp16	2	58.47	58.49	-0.03%	✅
resnet50v1_fp16	1	1,046.35	1,032.87	1.31%	✅
resnet50v1_int8	1	1,028.24	1,034.09	-0.57%	✅
bert_base_cased_fp16	64	1,180.56	1,180.18	0.03%	✅
bert_large_uncased_fp16	32	365.26	365.26	-0.00%	✅
bert_large_fp16	1	199.07	198.79	0.14%	✅
distilgpt2_fp16	16	2,227.41	2,225.55	0.08%	✅
yolov5s	1	530.54	524.79	1.10%	✅
tinyllama	1	43.78	43.59	0.44%	✅
vicuna-fastchat	1	173.88	174.11	-0.13%	✅
whisper-tiny-encoder	1	417.75	417.02	0.17%	✅
whisper-tiny-decoder	1	432.99	430.09	0.67%	✅

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