Issue with Preprocessing

[vamsi-krishna-mishran]

I downloaded coco dataset 2017 and its annotation file and tried to follow the train.md file
in Preprocessing I provide the path to my train2017 folder (unzipped) and instances_train2017.json (unzipped)
in the end its giving me errors like list indices must be integers or slices, not list. I felt there is an issue with the way i pass the data.

[viyer-research]

When downloading the pre-processed smaller size tracking dataset I get this error. https://imgtracking.blob.core.windows.net/pysiamtracking/data/got10k_train_images.zip