regression.dl22-passage.unicoil-0shot.cached.yaml Maven / Gradle / Ivy

Go to download
---
corpus: msmarco-v2-passage-unicoil-0shot
corpus_path: collections/msmarco/msmarco_v2_passage_unicoil_0shot/

download_url: https://rgw.cs.uwaterloo.ca/JIMMYLIN-bucket0/data/msmarco_v2_passage_unicoil_0shot.tar
download_checksum: 1949a00bfd5e1f1a230a04bbc1f01539
download_corpus: msmarco_v2_passage_unicoil_0shot

index_path: indexes/lucene-inverted.msmarco-v2-passage.unicoil-0shot/
collection_class: JsonVectorCollection
generator_class: DefaultLuceneDocumentGenerator
index_threads: 24
index_options: -impact -pretokenized -storeRaw
index_stats:
  documents: 138364198
  documents (non-empty): 138364198
  total terms: 775253560148

metrics:
  - metric: MAP@100
    command: bin/trec_eval
    params: -c -M 100 -m map -l 2
    separator: "\t"
    parse_index: 2
    metric_precision: 4
    can_combine: false
  - metric: MRR@100
    command: bin/trec_eval
    params: -c -M 100 -m recip_rank -l 2
    separator: "\t"
    parse_index: 2
    metric_precision: 4
    can_combine: false
  - metric: nDCG@10
    command: bin/trec_eval
    params: -c -m ndcg_cut.10
    separator: "\t"
    parse_index: 2
    metric_precision: 4
    can_combine: false
  - metric: R@100
    command: bin/trec_eval
    params: -c -m recall.100 -l 2
    separator: "\t"
    parse_index: 2
    metric_precision: 4
    can_combine: false
  - metric: R@1000
    command: bin/trec_eval
    params: -c -m recall.1000 -l 2
    separator: "\t"
    parse_index: 2
    metric_precision: 4
    can_combine: false

topic_reader: TsvInt
topics:
  - name: "[DL22 (Passage)](https://microsoft.github.io/msmarco/TREC-Deep-Learning)"
    id: dl22
    path: topics.dl22.unicoil.0shot.tsv.gz
    qrel: qrels.dl22-passage.txt

models:
  - name: unicoil-0shot-cached
    display: uniCOIL (with doc2query-T5) zero-shot
    params: -impact -pretokenized
    results:
      MAP@100:
        - 0.1050
      MRR@100:
        - 0.5831
      nDCG@10:
        - 0.4614
      R@100:
        - 0.2716
      R@1000:
        - 0.5253
  - name: unicoil-0shot-cached+rm3
    display: +RM3
    params: -impact -pretokenized -rm3 -collection JsonVectorCollection
    results:
      MAP@100:
        - 0.1170
      MRR@100:
        - 0.5634
      nDCG@10:
        - 0.4608
      R@100:
        - 0.2751
      R@1000:
        - 0.5372
  - name: unicoil-0shot-cached+rocchio
    display: +Rocchio
    params: -impact -pretokenized -rocchio -collection JsonVectorCollection
    results:
      MAP@100:
        - 0.1225
      MRR@100:
        - 0.5577
      nDCG@10:
        - 0.4886
      R@100:
        - 0.2860
      R@1000:
        - 0.5559