I’m trying to use any of the pre-trained BERT models to fine-tune with my own dataset.
My dataset is some sort of time-sequential English sentences, somewhat unstructured. – more like machine languages or SW languages
I want to test if the model can find any representation or feature from a dataset so that it can distinguish a new dataset that is similar to the trained data but contains some different contents.
It’s my first time in the NLP domain, so I was looking around, and initially thought I’d use bert-base-uncased since it’s just one of the most basic, but it’s for classification if input1-input2 is in the right sequence…? (Or am I getting a wrong concept of this moel? not sure…)
So any recommendation on which model I need is appreciated