Error while running code in 5_pytorch_retinanet/Train Resnet34

[venkatchintha]

Hello,

getting error while running below code

gtf.Dataset([train_root, train_img_dir, train_anno_file], batch_size=batch_size);

100%|██████████| 49281/49281 [00:00<00:00, 68809.42it/s]

---

AttributeError Traceback (most recent call last)

in ()
----> 1 gtf.Dataset([train_root, train_img_dir, train_anno_file], batch_size=batch_size);

2 frames

/content/Monk_Object_Detection/2_pytorch_finetune/lib/data_loader_class.py in get_labels(self)
35 label_list = [];
36 for i in range(len(self.train_list)):
---> 37 label = self.train_list[self.columns[1]][i];
38 tmp = label.split(" ");
39 for j in range(len(tmp)//5):

AttributeError: 'CustomDatasetMultiObject' object has no attribute 'columns'

https://colab.research.google.com/drive/1xMfTHyQvwjNGZvQIu6PJ4HxChKE3D_rz#scrollTo=jEEUINedgmDq

[abhi-kumar]

The notebook is not accessible

[abhi-kumar]

The sample notebook runs without any errors please check your json annotation file if it was created in proper format or not.

[abhi-kumar]

Is this error resolved now?

[venkatchintha]

Yes Abhishek, error now resolved. Thank you

[abhi-kumar]

Was the error related to corrupted json file or were code changes required?

[venkatchintha]

Sorry Abhishek, I was looking at wrong notebook. Still I am getting error. Annos looking good. I used same Annos for different model also.

TRAIN-Pytorch-faster-rcnn_mobilenet-v2

This is the model I am using