Fine-Tuning Qwen for Fake News Detection

#1
by Ask951169 - opened

Hey, I came across your FakeNews model. Using a small model to classify misinformation and explain its decision is an interesting approach. What was the biggest challenge in preparing the training data?

Hi, thanks for having a look at my model.

Biggest challenge was primarily finding a large enough dataset containing labelled annotations of factual and real context. Most open source datasets had only a couple 100 or 1000 rows but didn’t contain a source or factual annotation.

Sign up or log in to comment