Model trained on TriviaQA, filtered of all questions used in ViQuAE, as described in https://paullerner.github.io/ViQuAE/index.html#reading-comprehension but after fixing the loss (mean-pooling instead of max-pooling, see https://github.com/facebookresearch/DPR/issues/244)

Downloads last month
32
Inference Providers NEW
This model is not currently available via any of the supported third-party Inference Providers, and HF Inference API was unable to determine this model’s pipeline type.

Dataset used to train PaulLerner/multi_passage_bert_triviaqa_without_viquae_mean_pool_loss