a-ware/bart-squadv2
A-ware/bart-squadv2 is a machine learning model.
About a-ware/bart-squadv2
BART-large model finetuned on SQuADv2 dataset for question answering task . BART is a seq2seq model intended for both NLG and NLU tasks . BART can handle sequences with upto 1024 tokens . BART was propsed in the paper BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension. BART-Large achives comparable to ROBERTa on SQAD. BART can be trained with following parameters using simpletransformers wrapper: . BartForQuestionAnswering.from_pretrained('a-ware/bart-squadv2') The model was trained with the following parameters,