Fix YOLOv5 Detect Layer compatibility

[zhiqwang]

Let's use YOLOv5's layout to load the trained checkpoints. Currently this tool load_yolov5_model() is mainly used to load the checkpoints trained by yolov5 with support for versions v3.1, v4.0 (v5.0) and v6.0 (v6.1). In addition it is available for inference with AutoShape attached for versions v6.0 (v6.1).

from torch import Tensor
from yolort.v5 import AutoShape, attempt_download, load_yolov5_model

img_path = "test/assets/zidane.jpg"

model_url = "https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5s.pt"
checkpoint_path = attempt_download(model_url, hash_prefix="8b3b748c")

model = load_yolov5_model(checkpoint_path)
# Attach AutoShape
model = AutoShape(model)

results = model(img_path)

assert isinstance(results.pred, list)
assert len(results.pred) == 1
assert isinstance(results.pred[0], Tensor)
assert results.pred[0].shape == (4, 6)

[codecov]

Codecov Report

Merging #345 (875eefa) into main (719f76f) will increase coverage by 0.37%.
The diff coverage is 53.84%.

❗ Current head 875eefa differs from pull request most recent head d0ebe9e. Consider uploading reports for the commit d0ebe9e to get more accurate results

@@            Coverage Diff             @@
##             main     #345      +/-   ##
==========================================
+ Coverage   95.53%   95.90%   +0.37%     
==========================================
  Files          11       11              
  Lines         783      781       -2     
==========================================
+ Hits          748      749       +1     
+ Misses         35       32       -3     

Flag	Coverage Δ
unittests	95.90% <53.84%> (+0.37%)	⬆️

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

Impacted Files	Coverage Δ
test/test_utils.py	80.37% <0.00%> (+2.19%)	⬆️
test/test_v5.py	100.00% <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 719f76f...d0ebe9e. Read the comment docs.