[Tracker] All the issue related with ONNX model zoo models

[pdhirajkumarprasad]

We have enabled total 2325 models from ONNX model zoo. This issue is to track all the issues seen in the models.

Passing Summary

TOTAL TESTS = 2325

Stage	# Passing	% of Total	% of Attempted
Setup	2311	99.4%	99.4%
IREE Compilation	2024	87.1%	87.6%
Gold Inference	2024	87.1%	100.0%
IREE Inference Invocation	2015	86.7%	99.6%
Inference Comparison (PASS)	1959	84.3%	97.2%

Fail Summary

TOTAL TESTS = 2325

Stage	# Failed at Stage	% of Total
Setup	14	0.6%
IREE Compilation	287	12.3%
Gold Inference	0	0.0%
IREE Inference Invocation	9	0.4%
Inference Comparison	56	2.4%

CPU

#	issue type	issue no	#model impacted	list of model	assignee	status
1	failed to legalize operation 'torch.operator' that was explicitly marked illegal		62	NonMaxSuppression, MatMulInteger, Scan, Conv, Loop, QLinearConv, DequantizeLinear, AveragePool, QLinearMatMul
2	One or more operations with large vector sizes (16384 bytes) were found		139
3	expected offsets to be non-negative, but got -1	19501	6
4	'vector.transfer_write' op inferred mask type ('vector<1x1x8xi1>') and mask operand type ('vector<1x8x1xi1>') don't match	19383	9
5	failed to legalize operation 'torch.aten.nonzero'	llvm/torch-mlir#3876	8
6	'tensor.reshape' op source and destination tensor should have the same number of elements		3
7	'linalg.generic' op inferred input/output operand #0 has shape's dimension #2 to be greater than or equal to 4, but found 1		5
8	failed to legalize operation 'torch.aten.convolution' that was explicitly marked illegal		3
9	expected type to be 'tensor<?x?x?xf32>' or a rank-reduced version. (size mismatch)		3
10	'tensor.cast' op operand type 'tensor<80x15130xf32>' and result type 'tensor<15130xf32>' are cast incompatible	891	3