methods/searchain/ColBERT/index.py

from colbert.infra import Run, RunConfig, ColBERTConfig
from colbert import Indexer

dataset = "manyqa_text"

if __name__=='__main__':
    with Run().context(RunConfig(nranks=1, experiment=dataset)):

        config = ColBERTConfig(
            nbits=2,
            root="/ColBERT-main/experiments",
        )
        indexer = Indexer(checkpoint="colbert-ir/colbertv2.0", config=config)
        indexer.index(name=f"{dataset}.nbits=2", collection=f"/data1/liuyaoyang/Papers/icml2025/multi_rag/RAG/Search-in-the-Chain/data/{dataset}.tsv")