Closed celadaris closed 1 year ago
the results above appear when trying to export yolov7 without including --end2end
for example:
python export.py --weights "C:\Users\imnot\Downloads\yolov7-main\yolov7-main\runs\train\yolov7-custom2\weights\best.pt" --grid --simplify --topk-all 100 --img-size 640 640 --max-wh 640
but if i include --end2end
like so:
python export.py --weights "C:\Users\imnot\Downloads\yolov7-main\yolov7-main\runs\train\yolov7-custom2\weights\best.pt" --grid --end2end --simplify --topk-all 100 --img-size 640 640 --max-wh 640
then I get other problems, when using ONNX-YOLOv7-Object-Detection i get no inferences on the object (pokeball)
and in yolov7net i get the following error:
so im really confused on if im supposed to use --end2end
when exporting to .onnx because with or without i run into problems.
while im still a little confused on the inconsistency, after downloading a pre-made dataset from roboflow and getting it to work, i think my problems have to do with either my webcam quality, my sample size, or both.
ive had great results on both yolov7net and ONNX-YOLOv7-Object-Detection using models from here but...
i trained a custom model and if i use it in ONNX-YOLOv7-Object-Detection i get results, however if i use Yolov7net i got no results. Im still very new at training custom models but when im seeing results on one library and not on another it makes me question on whether im doing something wrong or if theres something wrong with yolov7net...
Note:
results from using ONNX-YOLOv7-Object-Detection:
results from using Yolov7net:
code from yolov7net:
code from ONNX-YOLOv7-Object-Detection: