aj8907

Reverting back to version 1.10 fixed this problem for me. `pip install gpytorch==1.10` Ran [this](https://docs.gpytorch.ai/en/latest/examples/02_Scalable_Exact_GPs/KeOps_GP_Regression.html) example on Google Colab V100 High Ram.

I had a similar error with the corrupted double-linked list (Colab A100). Running this code fixed the problem for me: ``` from threading import RLock from contextlib import contextmanager import...