deepstack-trainer
deepstack-trainer copied to clipboard
johnolafenwa
Custom Object Detection Training for DeepStack
The most recent functional Torch version that works with deepstack-trainer is 1.7.1 (without getting in place leaf node operation error that can be found in the resolved issues tab), which...
This is a simple example that shows the issue. If I load Yolo v6 from the hub and then also try to use a model trained with this, it seems...
Addresses https://github.com/johnolafenwa/deepstack-trainer/issues/5
I used Roboflow to import my yolo5 training data, and this doesn't include a classes.txt so I had to read the source to know what was required. We can just...
As its always nice to see. In my case performed with: ``` !python /content/deepstack-trainer/detect.py --weights /content/deepstack-trainer/train-runs/exp/weights/best.pt --img 640 --conf 0.25 --source /content/valid/images/ ``` This batch processes all the images in...
Training is taking surprisingly long, so would be useful to monitor metrics as we go
Good example writeups: - https://github.com/shayanalibhatti/Retail-Store-Item-Detection-using-YOLOv5 - https://blog.roboflow.com/using-computer-vision-to-count-fish-populations/ - https://blog.roboflow.com/computer-vision-american-sign-language/
300 appears overkill if this is correct, although I have only a single class.
Metadata
Owner
Metadata
Custom Object Detection Training for DeepStack