Run on Google Colab, but out of System RAM

[Holy-Egg]

Because Colab cannot accommodate all the mentioned KITTI data sets, So I only downloaded the 2011_09_29 data set.
After everything is set up start running train.py.
During the operation, Colab's system RAM began to increase until it exceeded 12.7GB, and then it automatically stopped running. How can I solve the problem of loading training data?

[R0nSam]

How did you only download 2011_09_29?