Refactor ONNX export and tutorial

[zhiqwang]

We introduce a new tool to export ONNX models:

from yolort.models import yolov5n6
from yolort.runtime.ort_helper import export_onnx

onnx_path = "yolov5n6.onnx"
score_thresh = 0.35

model = yolov5n6(pretrained=True, score_thresh=score_thresh)
model = model.eval()

export_onnx(model=model, onnx_path=onnx_path)

[codecov]

Codecov Report

Merging #319 (8d704bd) into main (7121d65) will increase coverage by 0.12%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main     #319      +/-   ##
==========================================
+ Coverage   94.98%   95.10%   +0.12%     
==========================================
  Files          11       11              
  Lines         738      736       -2     
==========================================
- Hits          701      700       -1     
+ Misses         37       36       -1     

Flag	Coverage Δ
unittests	95.10% <100.00%> (+0.12%)	⬆️

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

Impacted Files	Coverage Δ
test/test_runtime_ort.py	95.91% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 7121d65...8d704bd. Read the comment docs.