Edit model card

GPT2 QA - Persian

It is a new approach to using GPT2 in other downstream NLP tasks like QA. The model was trained on PersianQA and evaluated on PersianQA and PersiNLU (Reading Comprehension).

Dataset

Evaluation

The following table summarizes the scores obtained by the model.

Dataset F1 Score (%) Exact Match (%) Total (#)
ParsNLU 46.95 20.39 564
PersianQA 45.93 23.19 651

Demo

Streamlit GPT2 QA - Persian

How to use

TODO (will be filled shortly)...

Downloads last month
31

Dataset used to train m3hrdadfi/gpt2-persian-qa

Space using m3hrdadfi/gpt2-persian-qa 1