transformer_rankers.datasets.dataset.QueryPosDocNegDocDataset

class transformer_rankers.datasets.dataset.QueryPosDocNegDocDataset(data, tokenizer, data_partition, negative_sampler, task_type, max_seq_len, sample_data, cache_path, generate_inverted_pairs=False)

Bases: torch.utils.data.dataset.Dataset

Dataset for pairwise learning with <Query,PosDoc,NegDoc> triplets. For each batch both the positive and the negative documents are kept in memory.

__init__(data, tokenizer, data_partition, negative_sampler, task_type, max_seq_len, sample_data, cache_path, generate_inverted_pairs=False)

Initialize self. See help(type(self)) for accurate signature.

Methods

__init__(data, tokenizer, data_partition, …)

Initialize self.