YOLOv3:train darknet model on Colab

67 views Asked by At

I want to use the mask dataset on Kaggle and YOLOv3 to train a real-time mask recognition. But I have difficulty training darknet model on Colab.

The code below was what I have inputted into Colab, but it kept giving me the wrong messages. %cd /content/drive/MyDrive/Colab Notebooks/yolo_test ! '/content/drive/MyDrive/Colab Notebooks/yolo_test/'darknet detector train '/content/cfg/obj.data' '/content/darknet/cfg/yolov3-tiny_obj.cfg' '/content/drive/MyDrive/Colab Notebooks/yolo_test/darknet53.conv.74' -dont_show CODES

And what's in my yolo_test folder:

  1. darknet(downloaded from https://github.com/AlexeyAB/darknet.git and completely upload it to Drive)
  2. darknet53.conv.74 (downloaded from https://pjreddie.com/darknet/media/files/darknet53.conv.74 and completely upload it to Drive)

As for obj.data, I get it after I create configuration data. Thanks for everyone's help!

I have seen similar questions on StackOverflow, using the code in their response but it still can't solve my problem. I expect someone can give me some advice or suggestion, telling how to solve the problem.

0

There are 0 answers