Fine-tuning process

[smruti241]

Hi @Zhihan1996 , thanks for providing the code for finetuning DNABERT2. But there is no mention of how to generate the dev.tsv, test.csv and train.csv from our own dataset and how to provide the label 1 and 0 to the sequences. can you please let me know how to do that?

[TheRainInSpain]

I generated the csv files using the python csv library by writing each sequence and label into one row. But when I ran the code, error happened (described in #42 ). I wonder whether this is the right way to generate the csv files.

[smruti241]

but how did you label the sequence, based on which parameter? or randomly assigning 1 or 0? because randomly assigning wont make any sense.

[smruti241]

42

[smruti241]

I have covid data and DNABERT-2 has already covid data for finetuning. I just want to know how did you know each sequence should be given 1,2,3,4 , etc upto 9 as label? Please let me know @Zhihan1996