Learn.fine_tune() cause out of memory error

Part 1 (2020)
1

When running the first model in chapter 1, I got a three kinds of error, each time you run the cell, one of the three following error will happen:

OSError: [Errno 12] Cannot allocate memory

RuntimeError: [enforce fail at CPUAllocator.cpp:64] . DefaultCPUAllocator: can't allocate memory: you tried to allocate 205520896 bytes. Error code 12 (Cannot allocate memory)
RuntimeError: DataLoader worker (pid 1201) is killed by signal: Killed.

I tried with other cell with learn.fine_tune() in chapter 2, which gives the same error.
I am using gradient paperspace, is this the problem of their server? How should I troubleshoot this?

2

Please see this topic thread Run Time Error on lesson 2

1 Like
3

Turns out my notebook is not running with a GPU.
For future reference, if you are using gradient, make sure you start the notebook with a free GPU, sometimes the default will be the free CPU.