feat(src/layer): add vision_transformer benchmark

[tpoisonooo]

支持 vision_transformer。

• onnx2ncnn.cpp 增加了转换 GELU 和 Crop
• MultiHeadAttenction opr 加了句注释，冷门 opr 不说来源真的扛不住...
• benchmark 增加了转好的 ViT.param 。完整的模型已经测过数值对的

速度测试 on 8 核 i5 CPU，330MB 大模型。

vision_transformer min = 3236.20 max = 3541.62 avg = 3403.61

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 93.35%. Comparing base (bf1109c) to head (6b5b942).
Report is 789 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #3730      +/-   ##
==========================================
+ Coverage   93.33%   93.35%   +0.01%     
==========================================
  Files         672      674       +2     
  Lines      171577   171980     +403     
==========================================
+ Hits       160146   160556     +410     
+ Misses      11431    11424       -7     

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

[lgtm-com]

This pull request introduces 1 alert when merging eb7d4b2 into f14388a - view on LGTM.com

new alerts:

• 1 for Multiplication result converted to larger type

[tpoisonooo]

@nihui test_squeezenet 的错，是要我重造模型？ 我也没改涉及 sqz 的东西。

[tpoisonooo]

@nihui 我没有要改的了。

[tpoisonooo]

@nihui 好人给 review 一下呗。

[nihui]

InnerProduct     MatMul_28                1 1 onnx::MatMul_180 onnx::Transpose_182 0=3072 1=0 2=2359296
Permute          Transpose_29             1 1 onnx::Transpose_182 onnx::Reshape_183 0=1
Reshape          Reshape_31               1 1 onnx::Reshape_183 onnx::Add_185 0=1 1=-1 2=3072
BinaryOp         Add_32                   2 1 onnx::Add_185 onnx::Add_653 onnx::Reshape_188 0=0
Reshape          Reshape_34               1 1 onnx::Reshape_188 onnx::Transpose_190 0=-1 1=3072
Permute          Transpose_35             1 1 onnx::Transpose_190 mmdeploy::Gelu_191 0=1

这套东西，其实就是一个 nn.Linear ，被onnx拆散了 😆

[tpoisonooo]

@nihui 改好啦，扔掉了 gather； ViT.param 也小了很多。

[nihui]

Thanks for your contribution !