a-ware/xlmroberta-squadv2

This is xlm-roberta-large model finetuned on SQuADv2 dataset for question answering task . XLM-Roberta was propsed in the paper **XLM-R: State-of-the-art cross-lingual understanding through self-supervision . This model was trained with following parameters using simpletransformers wrapper: . The model was . trained with the following parameters: . 'train_batch_size': 8, 'num_train_epochs': 2, 'Gradual_accumulation_steps': . 'Gradually_accommod' and 'no_cache': True, 'use_cached,