ycdhqzhiai / yolov5_tensorrt

yolov5 pth convert tensorrt and inference

yolov5_tensorRT

1.Export tensorrt model

git clone https://github.com/ycdhqzhiai/yolov5_tensorRT
cd yolov5_tensorRT
pip install -r requirements.txt

复制export.py到原始yolov5仓库models目录下
python models/export.py 生成yolov5s.onnx,拷贝到weights目录下

2.run demo

python main.py

3.tensorrt int8 convert

python  convert_int8.py --onnx_file  weights/yolov5s.onnx  --data_path $your coustom datasets

yolov5_tensorrt's Issues

ValueError: too many values to unpack (expected 4)

Hi, thank you for sharing the code. I have got this error. any idea ? I tried to make sure that versions doesn't mis-match but still got this.

(1, 3, 80, 80, 6)
Traceback (most recent call last):
  File "main_yolov5.py", line 116, in <module>
    dets = yolonet.detect(srcimg)
  File "main_yolov5.py", line 89, in detect
    bs, _, ny, nx = outs[i].shape  # x(bs,255,20,20) to x(bs,3,20,20,85)
ValueError: too many values to unpack (expected 4)

Nice project! Could you please tell me which version of yolov5 are you choosed?

Nice project! Could you please tell me which version of yolov5 are you choosed? V5.0 or v4.0?
Thank you!