YoloV7

YoloV7 training on custom dataset

1. Clone the repo

 ! git clone  https://github.com/sivaji123256/YoloV7_Custom_Training.git

2. Change directory to yolov7 parent folder.

* In the parent folder go to data folder and access custom.yaml file and add the required path to the train dataset and test dateset and change other parametrs as   
  required
* Run the following command to install required packages 
  !pip install -r requirements.txt
  !sudo apt install -y zip htop screen libgl1-mesa-glx
* In the parent folder add required yolov7 pretrained weights.pt files

3. Training on custom dataset

* Use the following command to start training on custom dataset. 
  !python train.py --weights yolov7.pt --data "data/custom.yaml" --workers 4 --batch-size 4 --img 416 --cfg cfg/training/yolov7.yaml --name yolov7 --hyp       
  data/hyp.scratch.p5.yaml --epochs 5     

4. For Evaluation

* Run the following command 
  !python tools/eval.py --data dataset.yaml  --weights runs/train/exp/weights/best_ckpt.pt --device 0

5. For inference run the following command

  !python detect.py --weights runs/train/yolov7/weights/best.pt --source "path to your testing image"