[ALBERT] --random_next_sentence option in create_pretraining_data.py

[tomohideshibata]

Thanks for releasing ALBERT codes.

The default for --random_next_sentence option in create_pretraining_data.py is "True", but to enable the sentence order prediction (SOP) task, this option is set to be "False", right? If so, it is better to set the default to be "False".

The explanation for --random_next_sentence option in the source code seems strange. This explanation is for "False".

Thanks.

[lanzhzh]

Hi Hi,

We have updated the default settings to the ones we used in the paper. Thanks for mentioning this.