LLMeBench
LLMeBench copied to clipboard
Issue with OSACT4SubtaskB dataset loading
File: llmebench/datasets/OSACT4SubtaskB.py Line: 44 If we take the first two as text and label, the labels are ["HS", "NOT_HS", "OFF", "NOT_OFF"]. The "OFF" and "NOT_OFF" labels are for subtaskA.
the line 44 would be:
text, _, label = line.split("\t")