模型:
BeIR/sparta-msmarco-distilbert-base-v1
重新实现了 SPARTA: Efficient Open-Domain Question Answering via Sparse Transformer Matching Retrieval 。这是我们用于 BEIR: A Heterogenous Benchmark for Zero-shot Evaluation of Information Retrieval Models 的重新实现。
还请查看我们的BEIR存储库: https://github.com/UKPLab/beir
请查看 https://github.com/nreimers/beir-sparta 以获取此SPARTA模型的训练和推断代码