Problem with reading the dev.tsv in find_motif.py about dnabert

[mitchgill16]

Hi I think I had this issue. If I remember correctly my dev.tsv file's first line was 'sequence \t label'. I assume yours would be as well as finetuning in previous steps reads in the dev.tsv with pandas but automatically assigns a header to the first line. My solution was to change line 86
from
pd.read_csv(os.path.join(args.data_dir,"dev.tsv"),sep='\t',header=None)
to
pd.read_csv(os.path.join(args.data_dir,"dev.tsv"),sep='\t')

and comment out line 87
#dev.columns = ['sequence','label']

Hope this helps :)

[jerryji1993]

Hi @merouone and @mitchgill16,

Thanks for reporting this issue and sorry about the delay in my response. The previous dev data we used has no headers while we later switched to one that has headers. Hence causing this bug. We have recently updated the test data and fixed this in the latest version of the code. Please try it out and see if the bugs are fixed. Thank you!

Best,
Jerry