BearlyWorkingYT/OPT-125M-Kaggle-Creepypasta

This is the model trained for this video:

Due to hardware limitations, I trained this model with only a batch size of 2. (I know this isn't ideal).

The quality of the model may be affected.

After training was complete, the best model according to a hold-out set was used.

This model was trained using a filtered version of this dataset:

This dataset had a lot of blank entries and missing text.

Please subscribe to my YouTube Channel for bad quality videos and poorly trained models.