Training Process gets killed after starting

[Swayatta]

Firstly, thanks for providing a baseline code.
When we try to use the data provided to run the baseline, the script keeps getting killed. What is the problem?

[lisaterumi]

Same problem here...

[bfetahu]

Please check the amount of memory your GPU has. If you do not have a GPU with high memory, please reduce the batch size and alternatively you can shorten the sentence length (by default is 50 tokens) by passing the argument "--max_length 30".

The baseline models were trained on a Nvidia V100 GPU which has 16GB of GPU memory.