voidful/dpr-ctx_encoder-bert-base-multilingual

DRCD* and MLQA are converted using script from haystack squad_to_dpr.py . Dense Passage Retriever is called DensePassageRetriever . The tutorial is based on DPR's multilingual model for training and devolution . I use the script from the haystack repo to convert the training model to Dense Passage Retrivers . The training model was created using the DPR repo and Dense Retriev model . The model is a multilingual question-based model with a number of questions to be answered by train and dev . The question-answer model is used for dev and train training purposes only to answer questions .,