Add cast cases used by OPT-1.25M

[newling]

• add casting for boolean and int64 (it took some time to find the magic numbers in ONNX and torch for the types , so I added it in a comment @vivekkhandelwal1 please confirm)
• clang-formatting : I tried clang-format-12, clang-format-13, and clang-format-14, but not of them would preserve the original formatting of this file.

This resolves the cast issue in #2786, but compilation proceeds only to

 failed to legalize operation 'torch.operator' that was explicitly marked illegal
    %241 = torch.operator "onnx.LayerNormalization"(%240, %8, %9) {torch.onnx.axis = -1 : si64, torch.onnx.epsilon = 9.99999974E-6 : f32} : (!torch.vtensor<[1,4,768],f32>, !torch.vtensor<[768],f32>, !torch.vtensor<[768],f32>) -> !torch.vtensor<[1,4,768],f32>

• I will add a lit test in a follow-up PR.