Weight stripping

[simberg-amd]

Base version of weight stripping implemented.

[codecov[bot]]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 92.04%. Comparing base (1cd2854) to head (6676521).

Additional details and impacted files

@@           Coverage Diff            @@
##           develop    #3416   +/-   ##
========================================
  Coverage    92.04%   92.04%           
========================================
  Files          506      506           
  Lines        20872    20873    +1     
========================================
+ Hits         19212    19213    +1     
  Misses        1660     1660           

Flag	Coverage Δ
	92.04% <ø> (+<0.01%)	:arrow_up:

Flags with carried forward coverage won't be shown. Click here to find out more.

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[simberg-amd]

Data from weight stripping: The literals combined is how it will work if the weights folder is not deleted, if not touched the weight folder will share all weights from all quantizations. Total size if total size that compilation took

[migraphx-bot]

Test	Batch	Rate new ff0182	Rate old b6b228	Diff	Compare
torchvision-resnet50	64	3,252.25	3,244.98	0.22%	:white_check_mark:
torchvision-resnet50_fp16	64	6,996.23	6,978.98	0.25%	:white_check_mark:
torchvision-densenet121	32	2,432.70	2,430.28	0.10%	:white_check_mark:
torchvision-densenet121_fp16	32	4,074.13	4,069.20	0.12%	:white_check_mark:
torchvision-inceptionv3	32	1,636.23	1,633.12	0.19%	:white_check_mark:
torchvision-inceptionv3_fp16	32	2,741.44	2,735.44	0.22%	:white_check_mark:
cadene-inceptionv4	16	775.52	776.01	-0.06%	:white_check_mark:
cadene-resnext64x4	16	808.53	807.20	0.16%	:white_check_mark:
slim-mobilenet	64	7,458.93	7,449.39	0.13%	:white_check_mark:
slim-nasnetalarge	64	207.97	207.58	0.19%	:white_check_mark:
slim-resnet50v2	64	3,343.73	3,339.82	0.12%	:white_check_mark:
bert-mrpc-onnx	8	1,152.83	1,151.95	0.08%	:white_check_mark:
bert-mrpc-tf	1	312.21	312.09	0.04%	:white_check_mark:
pytorch-examples-wlang-gru	1	418.97	426.46	-1.76%	:white_check_mark:
pytorch-examples-wlang-lstm	1	386.43	378.80	2.01%	:white_check_mark:
torchvision-resnet50_1	1	804.52	807.34	-0.35%	:white_check_mark:
cadene-dpn92_1	1	433.78	415.79	4.33%	:high_brightness:
cadene-resnext101_1	1	383.62	380.36	0.86%	:white_check_mark:
onnx-taau-downsample	1	344.80	344.49	0.09%	:white_check_mark:
dlrm-criteoterabyte	1	35.09	35.03	0.17%	:white_check_mark:
dlrm-criteoterabyte_fp16	1	58.13	58.10	0.05%	:white_check_mark:
agentmodel	1	8,201.23	8,096.18	1.30%	:white_check_mark:
unet_fp16	2	57.99	57.91	0.13%	:white_check_mark:
resnet50v1_fp16	1	925.68	928.31	-0.28%	:white_check_mark:
resnet50v1_int8	1	982.49	940.28	4.49%	:high_brightness:
bert_base_cased_fp16	64	1,153.82	1,151.15	0.23%	:white_check_mark:
bert_large_uncased_fp16	32	355.68	355.17	0.14%	:white_check_mark:
bert_large_fp16	1	212.25	211.59	0.31%	:white_check_mark:
distilgpt2_fp16	16	2,152.28	2,157.04	-0.22%	:white_check_mark:
yolov5s	1	540.85	528.24	2.39%	:white_check_mark:
tinyllama	1	43.44	43.34	0.24%	:white_check_mark:
vicuna-fastchat	1	170.84	171.71	-0.50%	:white_check_mark:
whisper-tiny-encoder	1	407.54	406.85	0.17%	:white_check_mark:
whisper-tiny-decoder	1	429.89	433.17	-0.76%	:white_check_mark:

Check results before merge :high_brightness:

[migraphx-bot]

     :white_check_mark: bert-mrpc-onnx: PASSED: MIGraphX meets tolerance

     :white_check_mark: bert-mrpc-tf: PASSED: MIGraphX meets tolerance

     :white_check_mark: pytorch-examples-wlang-gru: PASSED: MIGraphX meets tolerance

     :white_check_mark: pytorch-examples-wlang-lstm: PASSED: MIGraphX meets tolerance

     :white_check_mark: torchvision-resnet50_1: PASSED: MIGraphX meets tolerance

     :white_check_mark: cadene-dpn92_1: PASSED: MIGraphX meets tolerance

     :white_check_mark: cadene-resnext101_1: PASSED: MIGraphX meets tolerance

     :white_check_mark: dlrm-criteoterabyte: PASSED: MIGraphX meets tolerance

     :white_check_mark: agentmodel: PASSED: MIGraphX meets tolerance

     :white_check_mark: unet: PASSED: MIGraphX meets tolerance

     :white_check_mark: resnet50v1: PASSED: MIGraphX meets tolerance

     :white_check_mark: bert_base_cased_fp16: PASSED: MIGraphX meets tolerance

:red_circle:bert_large_uncased_fp16: FAILED: MIGraphX is not within tolerance - check verbose output

     :white_check_mark: bert_large: PASSED: MIGraphX meets tolerance

     :white_check_mark: yolov5s: PASSED: MIGraphX meets tolerance

     :white_check_mark: tinyllama: PASSED: MIGraphX meets tolerance

     :white_check_mark: vicuna-fastchat: PASSED: MIGraphX meets tolerance

     :white_check_mark: whisper-tiny-encoder: PASSED: MIGraphX meets tolerance

     :white_check_mark: whisper-tiny-decoder: PASSED: MIGraphX meets tolerance

     :white_check_mark: distilgpt2_fp16: PASSED: MIGraphX meets tolerance

[jayhawk-commits]

/AzurePipelines run

[azure-pipelines[bot]]

Azure Pipelines successfully started running 1 pipeline(s).