vaibhavad's picture
Update README.md
93f20cc verified
metadata
library_name: peft
license: mit
language:
  - en
pipeline_tag: sentence-similarity
tags:
  - text-embedding
  - embeddings
  - information-retrieval
  - beir
  - text-classification
  - language-model
  - text-clustering
  - text-semantic-similarity
  - text-evaluation
  - text-reranking
  - feature-extraction
  - sentence-similarity
  - Sentence Similarity
  - natural_questions
  - ms_marco
  - fever
  - hotpot_qa
  - mteb
model-index:
  - name: LLM2Vec-Sheared-LLaMA-unsupervised
    results:
      - task:
          type: Classification
        dataset:
          type: mteb/amazon_counterfactual
          name: MTEB AmazonCounterfactualClassification (en)
          config: en
          split: test
          revision: e8379541af4e31359cca9fbcf4b00f2671dba205
        metrics:
          - type: accuracy
            value: 72.92537313432835
          - type: ap
            value: 36.6875749512053
          - type: f1
            value: 67.36274146169845
      - task:
          type: Classification
        dataset:
          type: mteb/amazon_polarity
          name: MTEB AmazonPolarityClassification
          config: default
          split: test
          revision: e2d317d38cd51312af73b3d32a06d1a08b442046
        metrics:
          - type: accuracy
            value: 74.282675
          - type: ap
            value: 69.15441866642587
          - type: f1
            value: 74.13028166370813
      - task:
          type: Classification
        dataset:
          type: mteb/amazon_reviews_multi
          name: MTEB AmazonReviewsClassification (en)
          config: en
          split: test
          revision: 1399c76144fd37290681b995c656ef9b2e06e26d
        metrics:
          - type: accuracy
            value: 36.136
          - type: f1
            value: 35.840498320506235
      - task:
          type: Retrieval
        dataset:
          type: arguana
          name: MTEB ArguAna
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 21.407999999999998
          - type: map_at_10
            value: 35.474
          - type: map_at_100
            value: 36.653999999999996
          - type: map_at_1000
            value: 36.68
          - type: map_at_3
            value: 30.974
          - type: map_at_5
            value: 33.265
          - type: mrr_at_1
            value: 22.119
          - type: mrr_at_10
            value: 35.714
          - type: mrr_at_100
            value: 36.895
          - type: mrr_at_1000
            value: 36.921
          - type: mrr_at_3
            value: 31.2
          - type: mrr_at_5
            value: 33.518
          - type: ndcg_at_1
            value: 21.407999999999998
          - type: ndcg_at_10
            value: 43.644
          - type: ndcg_at_100
            value: 49.035000000000004
          - type: ndcg_at_1000
            value: 49.685
          - type: ndcg_at_3
            value: 34.174
          - type: ndcg_at_5
            value: 38.288
          - type: precision_at_1
            value: 21.407999999999998
          - type: precision_at_10
            value: 6.999
          - type: precision_at_100
            value: 0.9440000000000001
          - type: precision_at_1000
            value: 0.099
          - type: precision_at_3
            value: 14.485999999999999
          - type: precision_at_5
            value: 10.683
          - type: recall_at_1
            value: 21.407999999999998
          - type: recall_at_10
            value: 69.986
          - type: recall_at_100
            value: 94.381
          - type: recall_at_1000
            value: 99.431
          - type: recall_at_3
            value: 43.457
          - type: recall_at_5
            value: 53.413999999999994
      - task:
          type: Clustering
        dataset:
          type: mteb/arxiv-clustering-p2p
          name: MTEB ArxivClusteringP2P
          config: default
          split: test
          revision: a122ad7f3f0291bf49cc6f4d32aa80929df69d5d
        metrics:
          - type: v_measure
            value: 42.915010245699904
      - task:
          type: Clustering
        dataset:
          type: mteb/arxiv-clustering-s2s
          name: MTEB ArxivClusteringS2S
          config: default
          split: test
          revision: f910caf1a6075f7329cdf8c1a6135696f37dbd53
        metrics:
          - type: v_measure
            value: 35.19568272188972
      - task:
          type: Reranking
        dataset:
          type: mteb/askubuntudupquestions-reranking
          name: MTEB AskUbuntuDupQuestions
          config: default
          split: test
          revision: 2000358ca161889fa9c082cb41daa8dcfb161a54
        metrics:
          - type: map
            value: 52.696972763822615
          - type: mrr
            value: 65.87136701402629
      - task:
          type: STS
        dataset:
          type: mteb/biosses-sts
          name: MTEB BIOSSES
          config: default
          split: test
          revision: d3fb88f8f02e40887cd149695127462bbcf29b4a
        metrics:
          - type: cos_sim_spearman
            value: 75.12038636775851
      - task:
          type: Classification
        dataset:
          type: mteb/banking77
          name: MTEB Banking77Classification
          config: default
          split: test
          revision: 0fd18e25b25c072e09e0d92ab615fda904d66300
        metrics:
          - type: accuracy
            value: 78.99675324675324
          - type: f1
            value: 78.90527329824852
      - task:
          type: Clustering
        dataset:
          type: mteb/biorxiv-clustering-p2p
          name: MTEB BiorxivClusteringP2P
          config: default
          split: test
          revision: 65b79d1d13f80053f67aca9498d9402c2d9f1f40
        metrics:
          - type: v_measure
            value: 35.02170435970243
      - task:
          type: Clustering
        dataset:
          type: mteb/biorxiv-clustering-s2s
          name: MTEB BiorxivClusteringS2S
          config: default
          split: test
          revision: 258694dd0231531bc1fd9de6ceb52a0853c6d908
        metrics:
          - type: v_measure
            value: 27.208216971540782
      - task:
          type: Retrieval
        dataset:
          type: cqadupstack/android
          name: MTEB CQADupstackAndroidRetrieval
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 16.432
          - type: map_at_10
            value: 23.769000000000002
          - type: map_at_100
            value: 25.038
          - type: map_at_1000
            value: 25.208000000000002
          - type: map_at_3
            value: 21.532999999999998
          - type: map_at_5
            value: 22.668
          - type: mrr_at_1
            value: 21.316
          - type: mrr_at_10
            value: 28.89
          - type: mrr_at_100
            value: 29.799999999999997
          - type: mrr_at_1000
            value: 29.887999999999998
          - type: mrr_at_3
            value: 26.705000000000002
          - type: mrr_at_5
            value: 27.864
          - type: ndcg_at_1
            value: 21.316
          - type: ndcg_at_10
            value: 28.656
          - type: ndcg_at_100
            value: 34.405
          - type: ndcg_at_1000
            value: 37.771
          - type: ndcg_at_3
            value: 24.98
          - type: ndcg_at_5
            value: 26.384999999999998
          - type: precision_at_1
            value: 21.316
          - type: precision_at_10
            value: 5.8229999999999995
          - type: precision_at_100
            value: 1.157
          - type: precision_at_1000
            value: 0.181
          - type: precision_at_3
            value: 12.446
          - type: precision_at_5
            value: 8.984
          - type: recall_at_1
            value: 16.432
          - type: recall_at_10
            value: 37.696000000000005
          - type: recall_at_100
            value: 63.198
          - type: recall_at_1000
            value: 86.651
          - type: recall_at_3
            value: 26.651000000000003
          - type: recall_at_5
            value: 30.901
      - task:
          type: Retrieval
        dataset:
          type: cqadupstack/english
          name: MTEB CQADupstackEnglishRetrieval
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 16.106
          - type: map_at_10
            value: 21.770999999999997
          - type: map_at_100
            value: 22.538
          - type: map_at_1000
            value: 22.656000000000002
          - type: map_at_3
            value: 19.918
          - type: map_at_5
            value: 20.957
          - type: mrr_at_1
            value: 21.083
          - type: mrr_at_10
            value: 26.502
          - type: mrr_at_100
            value: 27.161
          - type: mrr_at_1000
            value: 27.234
          - type: mrr_at_3
            value: 24.735
          - type: mrr_at_5
            value: 25.753999999999998
          - type: ndcg_at_1
            value: 21.083
          - type: ndcg_at_10
            value: 25.625999999999998
          - type: ndcg_at_100
            value: 29.152
          - type: ndcg_at_1000
            value: 32.025
          - type: ndcg_at_3
            value: 22.721
          - type: ndcg_at_5
            value: 24.029
          - type: precision_at_1
            value: 21.083
          - type: precision_at_10
            value: 4.8919999999999995
          - type: precision_at_100
            value: 0.844
          - type: precision_at_1000
            value: 0.13699999999999998
          - type: precision_at_3
            value: 11.104
          - type: precision_at_5
            value: 7.987
          - type: recall_at_1
            value: 16.106
          - type: recall_at_10
            value: 32.385999999999996
          - type: recall_at_100
            value: 47.961999999999996
          - type: recall_at_1000
            value: 67.63900000000001
          - type: recall_at_3
            value: 23.568
          - type: recall_at_5
            value: 27.326
      - task:
          type: Retrieval
        dataset:
          type: cqadupstack/gaming
          name: MTEB CQADupstackGamingRetrieval
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 22.517
          - type: map_at_10
            value: 29.593999999999998
          - type: map_at_100
            value: 30.695
          - type: map_at_1000
            value: 30.803000000000004
          - type: map_at_3
            value: 27.592
          - type: map_at_5
            value: 28.768
          - type: mrr_at_1
            value: 26.27
          - type: mrr_at_10
            value: 33.076
          - type: mrr_at_100
            value: 33.998
          - type: mrr_at_1000
            value: 34.073
          - type: mrr_at_3
            value: 31.223
          - type: mrr_at_5
            value: 32.257000000000005
          - type: ndcg_at_1
            value: 26.27
          - type: ndcg_at_10
            value: 33.726
          - type: ndcg_at_100
            value: 39.079
          - type: ndcg_at_1000
            value: 41.762
          - type: ndcg_at_3
            value: 30.064
          - type: ndcg_at_5
            value: 31.858999999999998
          - type: precision_at_1
            value: 26.27
          - type: precision_at_10
            value: 5.448
          - type: precision_at_100
            value: 0.898
          - type: precision_at_1000
            value: 0.121
          - type: precision_at_3
            value: 13.417000000000002
          - type: precision_at_5
            value: 9.317
          - type: recall_at_1
            value: 22.517
          - type: recall_at_10
            value: 42.814
          - type: recall_at_100
            value: 67.037
          - type: recall_at_1000
            value: 86.89099999999999
          - type: recall_at_3
            value: 33.041
          - type: recall_at_5
            value: 37.389
      - task:
          type: Retrieval
        dataset:
          type: cqadupstack/gis
          name: MTEB CQADupstackGisRetrieval
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 7.681
          - type: map_at_10
            value: 10.655000000000001
          - type: map_at_100
            value: 11.274000000000001
          - type: map_at_1000
            value: 11.381
          - type: map_at_3
            value: 9.793000000000001
          - type: map_at_5
            value: 10.202
          - type: mrr_at_1
            value: 8.248999999999999
          - type: mrr_at_10
            value: 11.453000000000001
          - type: mrr_at_100
            value: 12.074
          - type: mrr_at_1000
            value: 12.174
          - type: mrr_at_3
            value: 10.452
          - type: mrr_at_5
            value: 10.989
          - type: ndcg_at_1
            value: 8.248999999999999
          - type: ndcg_at_10
            value: 12.467
          - type: ndcg_at_100
            value: 15.942
          - type: ndcg_at_1000
            value: 19.378999999999998
          - type: ndcg_at_3
            value: 10.631
          - type: ndcg_at_5
            value: 11.411
          - type: precision_at_1
            value: 8.248999999999999
          - type: precision_at_10
            value: 1.966
          - type: precision_at_100
            value: 0.40099999999999997
          - type: precision_at_1000
            value: 0.075
          - type: precision_at_3
            value: 4.444
          - type: precision_at_5
            value: 3.186
          - type: recall_at_1
            value: 7.681
          - type: recall_at_10
            value: 17.302
          - type: recall_at_100
            value: 34.014
          - type: recall_at_1000
            value: 61.207
          - type: recall_at_3
            value: 12.389
          - type: recall_at_5
            value: 14.158999999999999
      - task:
          type: Retrieval
        dataset:
          type: cqadupstack/mathematica
          name: MTEB CQADupstackMathematicaRetrieval
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 3.868
          - type: map_at_10
            value: 6.281000000000001
          - type: map_at_100
            value: 6.903
          - type: map_at_1000
            value: 7.038
          - type: map_at_3
            value: 5.234
          - type: map_at_5
            value: 5.685
          - type: mrr_at_1
            value: 5.1
          - type: mrr_at_10
            value: 8.148
          - type: mrr_at_100
            value: 8.846
          - type: mrr_at_1000
            value: 8.963000000000001
          - type: mrr_at_3
            value: 6.944
          - type: mrr_at_5
            value: 7.498
          - type: ndcg_at_1
            value: 5.1
          - type: ndcg_at_10
            value: 8.405999999999999
          - type: ndcg_at_100
            value: 12.014
          - type: ndcg_at_1000
            value: 15.956999999999999
          - type: ndcg_at_3
            value: 6.22
          - type: ndcg_at_5
            value: 6.962
          - type: precision_at_1
            value: 5.1
          - type: precision_at_10
            value: 1.8159999999999998
          - type: precision_at_100
            value: 0.437
          - type: precision_at_1000
            value: 0.09
          - type: precision_at_3
            value: 3.1510000000000002
          - type: precision_at_5
            value: 2.463
          - type: recall_at_1
            value: 3.868
          - type: recall_at_10
            value: 13.319
          - type: recall_at_100
            value: 29.985
          - type: recall_at_1000
            value: 59.245999999999995
          - type: recall_at_3
            value: 7.0809999999999995
          - type: recall_at_5
            value: 8.914
      - task:
          type: Retrieval
        dataset:
          type: cqadupstack/physics
          name: MTEB CQADupstackPhysicsRetrieval
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 13.091
          - type: map_at_10
            value: 18.701999999999998
          - type: map_at_100
            value: 19.897000000000002
          - type: map_at_1000
            value: 20.044
          - type: map_at_3
            value: 17.041999999999998
          - type: map_at_5
            value: 17.943
          - type: mrr_at_1
            value: 16.939
          - type: mrr_at_10
            value: 23.038
          - type: mrr_at_100
            value: 24.029
          - type: mrr_at_1000
            value: 24.12
          - type: mrr_at_3
            value: 21.221999999999998
          - type: mrr_at_5
            value: 22.198999999999998
          - type: ndcg_at_1
            value: 16.939
          - type: ndcg_at_10
            value: 22.566
          - type: ndcg_at_100
            value: 28.364
          - type: ndcg_at_1000
            value: 31.646
          - type: ndcg_at_3
            value: 19.646
          - type: ndcg_at_5
            value: 20.915
          - type: precision_at_1
            value: 16.939
          - type: precision_at_10
            value: 4.340999999999999
          - type: precision_at_100
            value: 0.882
          - type: precision_at_1000
            value: 0.13799999999999998
          - type: precision_at_3
            value: 9.785
          - type: precision_at_5
            value: 6.93
          - type: recall_at_1
            value: 13.091
          - type: recall_at_10
            value: 30.022
          - type: recall_at_100
            value: 55.579
          - type: recall_at_1000
            value: 78.14
          - type: recall_at_3
            value: 21.4
          - type: recall_at_5
            value: 25.020999999999997
      - task:
          type: Retrieval
        dataset:
          type: cqadupstack/programmers
          name: MTEB CQADupstackProgrammersRetrieval
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 11.315999999999999
          - type: map_at_10
            value: 16.191
          - type: map_at_100
            value: 17.116
          - type: map_at_1000
            value: 17.262
          - type: map_at_3
            value: 14.302999999999999
          - type: map_at_5
            value: 15.278
          - type: mrr_at_1
            value: 14.269000000000002
          - type: mrr_at_10
            value: 19.409000000000002
          - type: mrr_at_100
            value: 20.298
          - type: mrr_at_1000
            value: 20.393
          - type: mrr_at_3
            value: 17.504
          - type: mrr_at_5
            value: 18.423000000000002
          - type: ndcg_at_1
            value: 14.269000000000002
          - type: ndcg_at_10
            value: 19.735
          - type: ndcg_at_100
            value: 24.582
          - type: ndcg_at_1000
            value: 28.337
          - type: ndcg_at_3
            value: 16.220000000000002
          - type: ndcg_at_5
            value: 17.644000000000002
          - type: precision_at_1
            value: 14.269000000000002
          - type: precision_at_10
            value: 3.721
          - type: precision_at_100
            value: 0.752
          - type: precision_at_1000
            value: 0.129
          - type: precision_at_3
            value: 7.800999999999999
          - type: precision_at_5
            value: 5.753
          - type: recall_at_1
            value: 11.315999999999999
          - type: recall_at_10
            value: 27.693
          - type: recall_at_100
            value: 49.265
          - type: recall_at_1000
            value: 76.291
          - type: recall_at_3
            value: 17.593
          - type: recall_at_5
            value: 21.368000000000002
      - task:
          type: Retrieval
        dataset:
          type: mteb/cqadupstack
          name: MTEB CQADupstackRetrieval
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 11.131583333333332
          - type: map_at_10
            value: 15.4605
          - type: map_at_100
            value: 16.3075
          - type: map_at_1000
            value: 16.4375
          - type: map_at_3
            value: 13.995833333333332
          - type: map_at_5
            value: 14.783666666666667
          - type: mrr_at_1
            value: 13.805833333333334
          - type: mrr_at_10
            value: 18.405749999999998
          - type: mrr_at_100
            value: 19.17516666666667
          - type: mrr_at_1000
            value: 19.265833333333333
          - type: mrr_at_3
            value: 16.892416666666666
          - type: mrr_at_5
            value: 17.71058333333333
          - type: ndcg_at_1
            value: 13.805833333333334
          - type: ndcg_at_10
            value: 18.500666666666664
          - type: ndcg_at_100
            value: 22.78191666666667
          - type: ndcg_at_1000
            value: 26.095583333333334
          - type: ndcg_at_3
            value: 15.846916666666663
          - type: ndcg_at_5
            value: 17.004250000000003
          - type: precision_at_1
            value: 13.805833333333334
          - type: precision_at_10
            value: 3.4233333333333325
          - type: precision_at_100
            value: 0.6828333333333333
          - type: precision_at_1000
            value: 0.11641666666666667
          - type: precision_at_3
            value: 7.511749999999999
          - type: precision_at_5
            value: 5.440916666666666
          - type: recall_at_1
            value: 11.131583333333332
          - type: recall_at_10
            value: 24.794166666666666
          - type: recall_at_100
            value: 44.356
          - type: recall_at_1000
            value: 68.71899999999998
          - type: recall_at_3
            value: 17.145583333333335
          - type: recall_at_5
            value: 20.229083333333335
      - task:
          type: Retrieval
        dataset:
          type: cqadupstack/stats
          name: MTEB CQADupstackStatsRetrieval
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 7.5520000000000005
          - type: map_at_10
            value: 10.355
          - type: map_at_100
            value: 10.875
          - type: map_at_1000
            value: 10.972999999999999
          - type: map_at_3
            value: 9.341000000000001
          - type: map_at_5
            value: 9.969
          - type: mrr_at_1
            value: 9.049
          - type: mrr_at_10
            value: 12.002
          - type: mrr_at_100
            value: 12.55
          - type: mrr_at_1000
            value: 12.635
          - type: mrr_at_3
            value: 11.12
          - type: mrr_at_5
            value: 11.626
          - type: ndcg_at_1
            value: 9.049
          - type: ndcg_at_10
            value: 12.241
          - type: ndcg_at_100
            value: 15.231
          - type: ndcg_at_1000
            value: 18.265
          - type: ndcg_at_3
            value: 10.424999999999999
          - type: ndcg_at_5
            value: 11.360000000000001
          - type: precision_at_1
            value: 9.049
          - type: precision_at_10
            value: 2.147
          - type: precision_at_100
            value: 0.411
          - type: precision_at_1000
            value: 0.073
          - type: precision_at_3
            value: 4.755
          - type: precision_at_5
            value: 3.558
          - type: recall_at_1
            value: 7.5520000000000005
          - type: recall_at_10
            value: 16.448999999999998
          - type: recall_at_100
            value: 30.505
          - type: recall_at_1000
            value: 54.435
          - type: recall_at_3
            value: 11.366
          - type: recall_at_5
            value: 13.758999999999999
      - task:
          type: Retrieval
        dataset:
          type: cqadupstack/tex
          name: MTEB CQADupstackTexRetrieval
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 5.954000000000001
          - type: map_at_10
            value: 8.229000000000001
          - type: map_at_100
            value: 8.694
          - type: map_at_1000
            value: 8.788
          - type: map_at_3
            value: 7.5
          - type: map_at_5
            value: 7.856000000000001
          - type: mrr_at_1
            value: 7.983
          - type: mrr_at_10
            value: 10.833
          - type: mrr_at_100
            value: 11.324
          - type: mrr_at_1000
            value: 11.404
          - type: mrr_at_3
            value: 9.911
          - type: mrr_at_5
            value: 10.401
          - type: ndcg_at_1
            value: 7.983
          - type: ndcg_at_10
            value: 10.126
          - type: ndcg_at_100
            value: 12.702
          - type: ndcg_at_1000
            value: 15.581999999999999
          - type: ndcg_at_3
            value: 8.779
          - type: ndcg_at_5
            value: 9.279
          - type: precision_at_1
            value: 7.983
          - type: precision_at_10
            value: 1.955
          - type: precision_at_100
            value: 0.392
          - type: precision_at_1000
            value: 0.076
          - type: precision_at_3
            value: 4.382
          - type: precision_at_5
            value: 3.09
          - type: recall_at_1
            value: 5.954000000000001
          - type: recall_at_10
            value: 13.472000000000001
          - type: recall_at_100
            value: 25.407999999999998
          - type: recall_at_1000
            value: 47.028
          - type: recall_at_3
            value: 9.367
          - type: recall_at_5
            value: 10.867
      - task:
          type: Retrieval
        dataset:
          type: cqadupstack/unix
          name: MTEB CQADupstackUnixRetrieval
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 8.894
          - type: map_at_10
            value: 12.758
          - type: map_at_100
            value: 13.639999999999999
          - type: map_at_1000
            value: 13.76
          - type: map_at_3
            value: 11.447000000000001
          - type: map_at_5
            value: 12.205
          - type: mrr_at_1
            value: 10.914
          - type: mrr_at_10
            value: 15.739
          - type: mrr_at_100
            value: 16.589000000000002
          - type: mrr_at_1000
            value: 16.679
          - type: mrr_at_3
            value: 14.179
          - type: mrr_at_5
            value: 15.162999999999998
          - type: ndcg_at_1
            value: 10.914
          - type: ndcg_at_10
            value: 15.629000000000001
          - type: ndcg_at_100
            value: 20.261000000000003
          - type: ndcg_at_1000
            value: 23.781
          - type: ndcg_at_3
            value: 13.102
          - type: ndcg_at_5
            value: 14.338000000000001
          - type: precision_at_1
            value: 10.914
          - type: precision_at_10
            value: 2.91
          - type: precision_at_100
            value: 0.601
          - type: precision_at_1000
            value: 0.10200000000000001
          - type: precision_at_3
            value: 6.311999999999999
          - type: precision_at_5
            value: 4.683
          - type: recall_at_1
            value: 8.894
          - type: recall_at_10
            value: 21.45
          - type: recall_at_100
            value: 42.617
          - type: recall_at_1000
            value: 69.233
          - type: recall_at_3
            value: 14.52
          - type: recall_at_5
            value: 17.681
      - task:
          type: Retrieval
        dataset:
          type: cqadupstack/webmasters
          name: MTEB CQADupstackWebmastersRetrieval
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 12.158
          - type: map_at_10
            value: 16.332
          - type: map_at_100
            value: 17.458000000000002
          - type: map_at_1000
            value: 17.687
          - type: map_at_3
            value: 14.529
          - type: map_at_5
            value: 15.515
          - type: mrr_at_1
            value: 15.809999999999999
          - type: mrr_at_10
            value: 19.917
          - type: mrr_at_100
            value: 20.875
          - type: mrr_at_1000
            value: 20.985
          - type: mrr_at_3
            value: 18.116
          - type: mrr_at_5
            value: 19.025
          - type: ndcg_at_1
            value: 15.809999999999999
          - type: ndcg_at_10
            value: 19.869999999999997
          - type: ndcg_at_100
            value: 24.907
          - type: ndcg_at_1000
            value: 29.076999999999998
          - type: ndcg_at_3
            value: 16.899
          - type: ndcg_at_5
            value: 18.23
          - type: precision_at_1
            value: 15.809999999999999
          - type: precision_at_10
            value: 3.972
          - type: precision_at_100
            value: 0.9860000000000001
          - type: precision_at_1000
            value: 0.203
          - type: precision_at_3
            value: 8.169
          - type: precision_at_5
            value: 6.087
          - type: recall_at_1
            value: 12.158
          - type: recall_at_10
            value: 26.338
          - type: recall_at_100
            value: 49.845
          - type: recall_at_1000
            value: 78.82000000000001
          - type: recall_at_3
            value: 16.997
          - type: recall_at_5
            value: 20.848
      - task:
          type: Retrieval
        dataset:
          type: cqadupstack/wordpress
          name: MTEB CQADupstackWordpressRetrieval
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 8.01
          - type: map_at_10
            value: 10.889
          - type: map_at_100
            value: 11.562
          - type: map_at_1000
            value: 11.65
          - type: map_at_3
            value: 9.718
          - type: map_at_5
            value: 10.358
          - type: mrr_at_1
            value: 8.688
          - type: mrr_at_10
            value: 11.862
          - type: mrr_at_100
            value: 12.558
          - type: mrr_at_1000
            value: 12.642000000000001
          - type: mrr_at_3
            value: 10.598
          - type: mrr_at_5
            value: 11.328000000000001
          - type: ndcg_at_1
            value: 8.688
          - type: ndcg_at_10
            value: 12.959999999999999
          - type: ndcg_at_100
            value: 16.744
          - type: ndcg_at_1000
            value: 19.564999999999998
          - type: ndcg_at_3
            value: 10.476
          - type: ndcg_at_5
            value: 11.639
          - type: precision_at_1
            value: 8.688
          - type: precision_at_10
            value: 2.089
          - type: precision_at_100
            value: 0.43299999999999994
          - type: precision_at_1000
            value: 0.07200000000000001
          - type: precision_at_3
            value: 4.375
          - type: precision_at_5
            value: 3.253
          - type: recall_at_1
            value: 8.01
          - type: recall_at_10
            value: 18.589
          - type: recall_at_100
            value: 36.857
          - type: recall_at_1000
            value: 59.047000000000004
          - type: recall_at_3
            value: 11.774
          - type: recall_at_5
            value: 14.516000000000002
      - task:
          type: Retrieval
        dataset:
          type: climate-fever
          name: MTEB ClimateFEVER
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 6.4719999999999995
          - type: map_at_10
            value: 12.322
          - type: map_at_100
            value: 14.122000000000002
          - type: map_at_1000
            value: 14.35
          - type: map_at_3
            value: 9.667
          - type: map_at_5
            value: 10.931000000000001
          - type: mrr_at_1
            value: 15.179
          - type: mrr_at_10
            value: 24.864
          - type: mrr_at_100
            value: 26.144000000000002
          - type: mrr_at_1000
            value: 26.198
          - type: mrr_at_3
            value: 20.999000000000002
          - type: mrr_at_5
            value: 23.097
          - type: ndcg_at_1
            value: 15.179
          - type: ndcg_at_10
            value: 18.951999999999998
          - type: ndcg_at_100
            value: 26.924
          - type: ndcg_at_1000
            value: 30.991999999999997
          - type: ndcg_at_3
            value: 13.778000000000002
          - type: ndcg_at_5
            value: 15.549
          - type: precision_at_1
            value: 15.179
          - type: precision_at_10
            value: 6.625
          - type: precision_at_100
            value: 1.516
          - type: precision_at_1000
            value: 0.22599999999999998
          - type: precision_at_3
            value: 10.51
          - type: precision_at_5
            value: 8.847
          - type: recall_at_1
            value: 6.4719999999999995
          - type: recall_at_10
            value: 25.191999999999997
          - type: recall_at_100
            value: 53.315
          - type: recall_at_1000
            value: 76.163
          - type: recall_at_3
            value: 12.834999999999999
          - type: recall_at_5
            value: 17.388
      - task:
          type: Retrieval
        dataset:
          type: dbpedia-entity
          name: MTEB DBPedia
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 1.947
          - type: map_at_10
            value: 4.858
          - type: map_at_100
            value: 7.185999999999999
          - type: map_at_1000
            value: 7.931000000000001
          - type: map_at_3
            value: 3.2939999999999996
          - type: map_at_5
            value: 3.914
          - type: mrr_at_1
            value: 23.25
          - type: mrr_at_10
            value: 33.035
          - type: mrr_at_100
            value: 33.721000000000004
          - type: mrr_at_1000
            value: 33.789
          - type: mrr_at_3
            value: 29.75
          - type: mrr_at_5
            value: 31.738
          - type: ndcg_at_1
            value: 15.625
          - type: ndcg_at_10
            value: 13.211999999999998
          - type: ndcg_at_100
            value: 16.422
          - type: ndcg_at_1000
            value: 23.058999999999997
          - type: ndcg_at_3
            value: 14.573
          - type: ndcg_at_5
            value: 13.733999999999998
          - type: precision_at_1
            value: 23.25
          - type: precision_at_10
            value: 12.45
          - type: precision_at_100
            value: 4.192
          - type: precision_at_1000
            value: 1.083
          - type: precision_at_3
            value: 18.667
          - type: precision_at_5
            value: 15.950000000000001
          - type: recall_at_1
            value: 1.947
          - type: recall_at_10
            value: 9.317
          - type: recall_at_100
            value: 23.066
          - type: recall_at_1000
            value: 45.704
          - type: recall_at_3
            value: 4.12
          - type: recall_at_5
            value: 5.591
      - task:
          type: Classification
        dataset:
          type: mteb/emotion
          name: MTEB EmotionClassification
          config: default
          split: test
          revision: 4f58c6b202a23cf9a4da393831edf4f9183cad37
        metrics:
          - type: accuracy
            value: 42.855
          - type: f1
            value: 39.029787102377576
      - task:
          type: Retrieval
        dataset:
          type: fever
          name: MTEB FEVER
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 8.461
          - type: map_at_10
            value: 13.655999999999999
          - type: map_at_100
            value: 14.499
          - type: map_at_1000
            value: 14.585999999999999
          - type: map_at_3
            value: 11.848
          - type: map_at_5
            value: 12.842999999999998
          - type: mrr_at_1
            value: 9.136
          - type: mrr_at_10
            value: 14.587
          - type: mrr_at_100
            value: 15.436
          - type: mrr_at_1000
            value: 15.518
          - type: mrr_at_3
            value: 12.690999999999999
          - type: mrr_at_5
            value: 13.747000000000002
          - type: ndcg_at_1
            value: 9.136
          - type: ndcg_at_10
            value: 16.958000000000002
          - type: ndcg_at_100
            value: 21.43
          - type: ndcg_at_1000
            value: 24.031
          - type: ndcg_at_3
            value: 13.191
          - type: ndcg_at_5
            value: 14.987
          - type: precision_at_1
            value: 9.136
          - type: precision_at_10
            value: 2.897
          - type: precision_at_100
            value: 0.532
          - type: precision_at_1000
            value: 0.077
          - type: precision_at_3
            value: 5.8709999999999996
          - type: precision_at_5
            value: 4.47
          - type: recall_at_1
            value: 8.461
          - type: recall_at_10
            value: 26.509
          - type: recall_at_100
            value: 47.776
          - type: recall_at_1000
            value: 68.26299999999999
          - type: recall_at_3
            value: 16.203
          - type: recall_at_5
            value: 20.505000000000003
      - task:
          type: Retrieval
        dataset:
          type: fiqa
          name: MTEB FiQA2018
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 7.396
          - type: map_at_10
            value: 12.393
          - type: map_at_100
            value: 13.857
          - type: map_at_1000
            value: 14.086000000000002
          - type: map_at_3
            value: 10.545
          - type: map_at_5
            value: 11.505
          - type: mrr_at_1
            value: 15.432000000000002
          - type: mrr_at_10
            value: 21.615000000000002
          - type: mrr_at_100
            value: 22.833000000000002
          - type: mrr_at_1000
            value: 22.931
          - type: mrr_at_3
            value: 19.522000000000002
          - type: mrr_at_5
            value: 20.663999999999998
          - type: ndcg_at_1
            value: 15.432000000000002
          - type: ndcg_at_10
            value: 16.986
          - type: ndcg_at_100
            value: 23.880000000000003
          - type: ndcg_at_1000
            value: 28.762999999999998
          - type: ndcg_at_3
            value: 14.482999999999999
          - type: ndcg_at_5
            value: 15.334999999999999
          - type: precision_at_1
            value: 15.432000000000002
          - type: precision_at_10
            value: 4.984999999999999
          - type: precision_at_100
            value: 1.167
          - type: precision_at_1000
            value: 0.2
          - type: precision_at_3
            value: 9.825000000000001
          - type: precision_at_5
            value: 7.469
          - type: recall_at_1
            value: 7.396
          - type: recall_at_10
            value: 21.389
          - type: recall_at_100
            value: 48.107
          - type: recall_at_1000
            value: 78.366
          - type: recall_at_3
            value: 13.181000000000001
          - type: recall_at_5
            value: 16.611
      - task:
          type: Retrieval
        dataset:
          type: hotpotqa
          name: MTEB HotpotQA
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 11.884
          - type: map_at_10
            value: 17.09
          - type: map_at_100
            value: 17.96
          - type: map_at_1000
            value: 18.081
          - type: map_at_3
            value: 15.296000000000001
          - type: map_at_5
            value: 16.289
          - type: mrr_at_1
            value: 23.768
          - type: mrr_at_10
            value: 29.991
          - type: mrr_at_100
            value: 30.862000000000002
          - type: mrr_at_1000
            value: 30.935000000000002
          - type: mrr_at_3
            value: 27.986
          - type: mrr_at_5
            value: 29.078
          - type: ndcg_at_1
            value: 23.768
          - type: ndcg_at_10
            value: 22.634999999999998
          - type: ndcg_at_100
            value: 27.059
          - type: ndcg_at_1000
            value: 30.145
          - type: ndcg_at_3
            value: 19.058
          - type: ndcg_at_5
            value: 20.762
          - type: precision_at_1
            value: 23.768
          - type: precision_at_10
            value: 5.2490000000000006
          - type: precision_at_100
            value: 0.8829999999999999
          - type: precision_at_1000
            value: 0.13
          - type: precision_at_3
            value: 12.091000000000001
          - type: precision_at_5
            value: 8.605
          - type: recall_at_1
            value: 11.884
          - type: recall_at_10
            value: 26.246000000000002
          - type: recall_at_100
            value: 44.153
          - type: recall_at_1000
            value: 64.889
          - type: recall_at_3
            value: 18.136
          - type: recall_at_5
            value: 21.512
      - task:
          type: Classification
        dataset:
          type: mteb/imdb
          name: MTEB ImdbClassification
          config: default
          split: test
          revision: 3d86128a09e091d6018b6d26cad27f2739fc2db7
        metrics:
          - type: accuracy
            value: 71.9232
          - type: ap
            value: 66.56619827391917
          - type: f1
            value: 71.60536244284128
      - task:
          type: Retrieval
        dataset:
          type: msmarco
          name: MTEB MSMARCO
          config: default
          split: dev
          revision: None
        metrics:
          - type: map_at_1
            value: 3.037
          - type: map_at_10
            value: 5.414
          - type: map_at_100
            value: 6.072
          - type: map_at_1000
            value: 6.172
          - type: map_at_3
            value: 4.437
          - type: map_at_5
            value: 4.939
          - type: mrr_at_1
            value: 3.123
          - type: mrr_at_10
            value: 5.572
          - type: mrr_at_100
            value: 6.235
          - type: mrr_at_1000
            value: 6.334
          - type: mrr_at_3
            value: 4.563
          - type: mrr_at_5
            value: 5.09
          - type: ndcg_at_1
            value: 3.123
          - type: ndcg_at_10
            value: 7.027
          - type: ndcg_at_100
            value: 10.776
          - type: ndcg_at_1000
            value: 13.904
          - type: ndcg_at_3
            value: 4.95
          - type: ndcg_at_5
            value: 5.865
          - type: precision_at_1
            value: 3.123
          - type: precision_at_10
            value: 1.252
          - type: precision_at_100
            value: 0.32299999999999995
          - type: precision_at_1000
            value: 0.059000000000000004
          - type: precision_at_3
            value: 2.168
          - type: precision_at_5
            value: 1.7680000000000002
          - type: recall_at_1
            value: 3.037
          - type: recall_at_10
            value: 12.11
          - type: recall_at_100
            value: 30.714999999999996
          - type: recall_at_1000
            value: 56.006
          - type: recall_at_3
            value: 6.3229999999999995
          - type: recall_at_5
            value: 8.518
      - task:
          type: Classification
        dataset:
          type: mteb/mtop_domain
          name: MTEB MTOPDomainClassification (en)
          config: en
          split: test
          revision: d80d48c1eb48d3562165c59d59d0034df9fff0bf
        metrics:
          - type: accuracy
            value: 91.24259005927954
          - type: f1
            value: 90.7594022786747
      - task:
          type: Classification
        dataset:
          type: mteb/mtop_intent
          name: MTEB MTOPIntentClassification (en)
          config: en
          split: test
          revision: ae001d0e6b1228650b7bd1c2c65fb50ad11a8aba
        metrics:
          - type: accuracy
            value: 74.08344733242134
          - type: f1
            value: 52.377556461789055
      - task:
          type: Classification
        dataset:
          type: mteb/amazon_massive_intent
          name: MTEB MassiveIntentClassification (en)
          config: en
          split: test
          revision: 31efe3c427b0bae9c22cbb560b8f15491cc6bed7
        metrics:
          - type: accuracy
            value: 69.99327505043712
          - type: f1
            value: 66.15141376479805
      - task:
          type: Classification
        dataset:
          type: mteb/amazon_massive_scenario
          name: MTEB MassiveScenarioClassification (en)
          config: en
          split: test
          revision: 7d571f92784cd94a019292a1f45445077d0ef634
        metrics:
          - type: accuracy
            value: 75.1546738399462
          - type: f1
            value: 74.83013584700711
      - task:
          type: Clustering
        dataset:
          type: mteb/medrxiv-clustering-p2p
          name: MTEB MedrxivClusteringP2P
          config: default
          split: test
          revision: e7a26af6f3ae46b30dde8737f02c07b1505bcc73
        metrics:
          - type: v_measure
            value: 30.146364191412356
      - task:
          type: Clustering
        dataset:
          type: mteb/medrxiv-clustering-s2s
          name: MTEB MedrxivClusteringS2S
          config: default
          split: test
          revision: 35191c8c0dca72d8ff3efcd72aa802307d469663
        metrics:
          - type: v_measure
            value: 26.96347584990607
      - task:
          type: Reranking
        dataset:
          type: mteb/mind_small
          name: MTEB MindSmallReranking
          config: default
          split: test
          revision: 3bdac13927fdc888b903db93b2ffdbd90b295a69
        metrics:
          - type: map
            value: 29.520993847103533
          - type: mrr
            value: 30.402007095845374
      - task:
          type: Retrieval
        dataset:
          type: nfcorpus
          name: MTEB NFCorpus
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 1.72
          - type: map_at_10
            value: 4.041
          - type: map_at_100
            value: 5.356000000000001
          - type: map_at_1000
            value: 6.413
          - type: map_at_3
            value: 2.9770000000000003
          - type: map_at_5
            value: 3.3689999999999998
          - type: mrr_at_1
            value: 21.981
          - type: mrr_at_10
            value: 30.286
          - type: mrr_at_100
            value: 31.272
          - type: mrr_at_1000
            value: 31.347
          - type: mrr_at_3
            value: 27.193
          - type: mrr_at_5
            value: 28.694999999999997
          - type: ndcg_at_1
            value: 19.814
          - type: ndcg_at_10
            value: 15.732
          - type: ndcg_at_100
            value: 16.033
          - type: ndcg_at_1000
            value: 25.865
          - type: ndcg_at_3
            value: 17.944
          - type: ndcg_at_5
            value: 16.634
          - type: precision_at_1
            value: 21.981
          - type: precision_at_10
            value: 12.786
          - type: precision_at_100
            value: 4.83
          - type: precision_at_1000
            value: 1.765
          - type: precision_at_3
            value: 17.75
          - type: precision_at_5
            value: 15.232000000000001
          - type: recall_at_1
            value: 1.72
          - type: recall_at_10
            value: 7.436
          - type: recall_at_100
            value: 20.275000000000002
          - type: recall_at_1000
            value: 54.19500000000001
          - type: recall_at_3
            value: 3.787
          - type: recall_at_5
            value: 4.829
      - task:
          type: Retrieval
        dataset:
          type: nq
          name: MTEB NQ
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 7.964
          - type: map_at_10
            value: 14.025000000000002
          - type: map_at_100
            value: 15.222
          - type: map_at_1000
            value: 15.32
          - type: map_at_3
            value: 11.886
          - type: map_at_5
            value: 13.056999999999999
          - type: mrr_at_1
            value: 9.183
          - type: mrr_at_10
            value: 15.651000000000002
          - type: mrr_at_100
            value: 16.753999999999998
          - type: mrr_at_1000
            value: 16.833000000000002
          - type: mrr_at_3
            value: 13.437
          - type: mrr_at_5
            value: 14.69
          - type: ndcg_at_1
            value: 9.183
          - type: ndcg_at_10
            value: 17.96
          - type: ndcg_at_100
            value: 23.823
          - type: ndcg_at_1000
            value: 26.461000000000002
          - type: ndcg_at_3
            value: 13.536999999999999
          - type: ndcg_at_5
            value: 15.642
          - type: precision_at_1
            value: 9.183
          - type: precision_at_10
            value: 3.366
          - type: precision_at_100
            value: 0.67
          - type: precision_at_1000
            value: 0.092
          - type: precision_at_3
            value: 6.547
          - type: precision_at_5
            value: 5.098
          - type: recall_at_1
            value: 7.964
          - type: recall_at_10
            value: 28.599000000000004
          - type: recall_at_100
            value: 55.381
          - type: recall_at_1000
            value: 75.63
          - type: recall_at_3
            value: 16.77
          - type: recall_at_5
            value: 21.671000000000003
      - task:
          type: Retrieval
        dataset:
          type: quora
          name: MTEB QuoraRetrieval
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 59.846999999999994
          - type: map_at_10
            value: 73.18599999999999
          - type: map_at_100
            value: 74.055
          - type: map_at_1000
            value: 74.09
          - type: map_at_3
            value: 69.95700000000001
          - type: map_at_5
            value: 71.925
          - type: mrr_at_1
            value: 69
          - type: mrr_at_10
            value: 77.23299999999999
          - type: mrr_at_100
            value: 77.52
          - type: mrr_at_1000
            value: 77.526
          - type: mrr_at_3
            value: 75.59
          - type: mrr_at_5
            value: 76.63799999999999
          - type: ndcg_at_1
            value: 69.02000000000001
          - type: ndcg_at_10
            value: 78.226
          - type: ndcg_at_100
            value: 80.60199999999999
          - type: ndcg_at_1000
            value: 80.971
          - type: ndcg_at_3
            value: 74.124
          - type: ndcg_at_5
            value: 76.265
          - type: precision_at_1
            value: 69.02000000000001
          - type: precision_at_10
            value: 12.102
          - type: precision_at_100
            value: 1.468
          - type: precision_at_1000
            value: 0.155
          - type: precision_at_3
            value: 32.5
          - type: precision_at_5
            value: 21.7
          - type: recall_at_1
            value: 59.846999999999994
          - type: recall_at_10
            value: 88.485
          - type: recall_at_100
            value: 97.425
          - type: recall_at_1000
            value: 99.523
          - type: recall_at_3
            value: 77.051
          - type: recall_at_5
            value: 82.762
      - task:
          type: Clustering
        dataset:
          type: mteb/reddit-clustering
          name: MTEB RedditClustering
          config: default
          split: test
          revision: 24640382cdbf8abc73003fb0fa6d111a705499eb
        metrics:
          - type: v_measure
            value: 38.67296729610079
      - task:
          type: Clustering
        dataset:
          type: mteb/reddit-clustering-p2p
          name: MTEB RedditClusteringP2P
          config: default
          split: test
          revision: 282350215ef01743dc01b456c7f5241fa8937f16
        metrics:
          - type: v_measure
            value: 53.42017351823769
      - task:
          type: Retrieval
        dataset:
          type: scidocs
          name: MTEB SCIDOCS
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 0.893
          - type: map_at_10
            value: 2.804
          - type: map_at_100
            value: 3.6740000000000004
          - type: map_at_1000
            value: 3.94
          - type: map_at_3
            value: 1.926
          - type: map_at_5
            value: 2.363
          - type: mrr_at_1
            value: 4.3
          - type: mrr_at_10
            value: 9.520000000000001
          - type: mrr_at_100
            value: 10.692
          - type: mrr_at_1000
            value: 10.841000000000001
          - type: mrr_at_3
            value: 7.6
          - type: mrr_at_5
            value: 8.63
          - type: ndcg_at_1
            value: 4.3
          - type: ndcg_at_10
            value: 5.531
          - type: ndcg_at_100
            value: 10.512
          - type: ndcg_at_1000
            value: 16.683
          - type: ndcg_at_3
            value: 4.632
          - type: ndcg_at_5
            value: 4.3229999999999995
          - type: precision_at_1
            value: 4.3
          - type: precision_at_10
            value: 3.16
          - type: precision_at_100
            value: 1.065
          - type: precision_at_1000
            value: 0.256
          - type: precision_at_3
            value: 4.667000000000001
          - type: precision_at_5
            value: 4.1000000000000005
          - type: recall_at_1
            value: 0.893
          - type: recall_at_10
            value: 6.428000000000001
          - type: recall_at_100
            value: 21.662
          - type: recall_at_1000
            value: 52.162
          - type: recall_at_3
            value: 2.868
          - type: recall_at_5
            value: 4.188
      - task:
          type: STS
        dataset:
          type: mteb/sickr-sts
          name: MTEB SICK-R
          config: default
          split: test
          revision: a6ea5a8cab320b040a23452cc28066d9beae2cee
        metrics:
          - type: cos_sim_spearman
            value: 69.34396953516386
      - task:
          type: STS
        dataset:
          type: mteb/sts12-sts
          name: MTEB STS12
          config: default
          split: test
          revision: a0d554a64d88156834ff5ae9920b964011b16384
        metrics:
          - type: cos_sim_spearman
            value: 60.094374065360746
      - task:
          type: STS
        dataset:
          type: mteb/sts13-sts
          name: MTEB STS13
          config: default
          split: test
          revision: 7e90230a92c190f1bf69ae9002b8cea547a64cca
        metrics:
          - type: cos_sim_spearman
            value: 72.51503781013379
      - task:
          type: STS
        dataset:
          type: mteb/sts14-sts
          name: MTEB STS14
          config: default
          split: test
          revision: 6031580fec1f6af667f0bd2da0a551cf4f0b2375
        metrics:
          - type: cos_sim_spearman
            value: 66.6954698644186
      - task:
          type: STS
        dataset:
          type: mteb/sts15-sts
          name: MTEB STS15
          config: default
          split: test
          revision: ae752c7c21bf194d8b67fd573edf7ae58183cbe3
        metrics:
          - type: cos_sim_spearman
            value: 77.69462578028768
      - task:
          type: STS
        dataset:
          type: mteb/sts16-sts
          name: MTEB STS16
          config: default
          split: test
          revision: 4d8694f8f0e0100860b497b999b3dbed754a0513
        metrics:
          - type: cos_sim_spearman
            value: 75.9397626457859
      - task:
          type: STS
        dataset:
          type: mteb/sts17-crosslingual-sts
          name: MTEB STS17 (en-en)
          config: en-en
          split: test
          revision: af5e6fb845001ecf41f4c1e033ce921939a2a68d
        metrics:
          - type: cos_sim_spearman
            value: 81.67242768943406
      - task:
          type: STS
        dataset:
          type: mteb/sts22-crosslingual-sts
          name: MTEB STS22 (en)
          config: en
          split: test
          revision: 6d1ba47164174a496b7fa5d3569dae26a6813b80
        metrics:
          - type: cos_sim_spearman
            value: 63.7027324700292
      - task:
          type: STS
        dataset:
          type: mteb/stsbenchmark-sts
          name: MTEB STSBenchmark
          config: default
          split: test
          revision: b0fddb56ed78048fa8b90373c8a3cfc37b684831
        metrics:
          - type: cos_sim_spearman
            value: 73.36074244064153
      - task:
          type: Reranking
        dataset:
          type: mteb/scidocs-reranking
          name: MTEB SciDocsRR
          config: default
          split: test
          revision: d3c5e1fc0b855ab6097bf1cda04dd73947d7caab
        metrics:
          - type: map
            value: 67.75984402370518
          - type: mrr
            value: 86.9951798383171
      - task:
          type: Retrieval
        dataset:
          type: scifact
          name: MTEB SciFact
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 24.583
          - type: map_at_10
            value: 33.125
          - type: map_at_100
            value: 34.14
          - type: map_at_1000
            value: 34.22
          - type: map_at_3
            value: 29.616
          - type: map_at_5
            value: 31.896
          - type: mrr_at_1
            value: 26.333000000000002
          - type: mrr_at_10
            value: 34.437
          - type: mrr_at_100
            value: 35.363
          - type: mrr_at_1000
            value: 35.433
          - type: mrr_at_3
            value: 31.333
          - type: mrr_at_5
            value: 33.267
          - type: ndcg_at_1
            value: 26.333000000000002
          - type: ndcg_at_10
            value: 38.311
          - type: ndcg_at_100
            value: 43.923
          - type: ndcg_at_1000
            value: 45.923
          - type: ndcg_at_3
            value: 31.596000000000004
          - type: ndcg_at_5
            value: 35.448
          - type: precision_at_1
            value: 26.333000000000002
          - type: precision_at_10
            value: 5.933
          - type: precision_at_100
            value: 0.91
          - type: precision_at_1000
            value: 0.109
          - type: precision_at_3
            value: 13
          - type: precision_at_5
            value: 9.933
          - type: recall_at_1
            value: 24.583
          - type: recall_at_10
            value: 53.417
          - type: recall_at_100
            value: 80.989
          - type: recall_at_1000
            value: 96.322
          - type: recall_at_3
            value: 35.611
          - type: recall_at_5
            value: 44.833
      - task:
          type: PairClassification
        dataset:
          type: mteb/sprintduplicatequestions-pairclassification
          name: MTEB SprintDuplicateQuestions
          config: default
          split: test
          revision: d66bd1f72af766a5cc4b0ca5e00c162f89e8cc46
        metrics:
          - type: cos_sim_accuracy
            value: 99.48514851485149
          - type: cos_sim_ap
            value: 77.36426466374054
          - type: cos_sim_f1
            value: 72.0702116675271
          - type: cos_sim_precision
            value: 74.49306296691569
          - type: cos_sim_recall
            value: 69.8
          - type: dot_accuracy
            value: 99.15049504950495
          - type: dot_ap
            value: 46.792474140260715
          - type: dot_f1
            value: 48.76476906552094
          - type: dot_precision
            value: 52.66821345707656
          - type: dot_recall
            value: 45.4
          - type: euclidean_accuracy
            value: 99.46534653465346
          - type: euclidean_ap
            value: 74.1978837990589
          - type: euclidean_f1
            value: 69.47256259989345
          - type: euclidean_precision
            value: 74.34435575826683
          - type: euclidean_recall
            value: 65.2
          - type: manhattan_accuracy
            value: 99.47128712871287
          - type: manhattan_ap
            value: 75.31910551743364
          - type: manhattan_f1
            value: 70.1582105837425
          - type: manhattan_precision
            value: 77.19087635054022
          - type: manhattan_recall
            value: 64.3
          - type: max_accuracy
            value: 99.48514851485149
          - type: max_ap
            value: 77.36426466374054
          - type: max_f1
            value: 72.0702116675271
      - task:
          type: Clustering
        dataset:
          type: mteb/stackexchange-clustering
          name: MTEB StackExchangeClustering
          config: default
          split: test
          revision: 6cbc1f7b2bc0622f2e39d2c77fa502909748c259
        metrics:
          - type: v_measure
            value: 59.353792480720436
      - task:
          type: Clustering
        dataset:
          type: mteb/stackexchange-clustering-p2p
          name: MTEB StackExchangeClusteringP2P
          config: default
          split: test
          revision: 815ca46b2622cec33ccafc3735d572c266efdb44
        metrics:
          - type: v_measure
            value: 31.474896484744836
      - task:
          type: Reranking
        dataset:
          type: mteb/stackoverflowdupquestions-reranking
          name: MTEB StackOverflowDupQuestions
          config: default
          split: test
          revision: e185fbe320c72810689fc5848eb6114e1ef5ec69
        metrics:
          - type: map
            value: 40.82378653430986
          - type: mrr
            value: 41.13905600118835
      - task:
          type: Summarization
        dataset:
          type: mteb/summeval
          name: MTEB SummEval
          config: default
          split: test
          revision: cda12ad7615edc362dbf25a00fdd61d3b1eaf93c
        metrics:
          - type: cos_sim_pearson
            value: 31.08154836998798
          - type: cos_sim_spearman
            value: 31.232033308845907
          - type: dot_pearson
            value: 23.767593496465828
          - type: dot_spearman
            value: 25.6201612766572
      - task:
          type: Retrieval
        dataset:
          type: trec-covid
          name: MTEB TRECCOVID
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 0.186
          - type: map_at_10
            value: 1.1809999999999998
          - type: map_at_100
            value: 5.21
          - type: map_at_1000
            value: 12.447999999999999
          - type: map_at_3
            value: 0.44200000000000006
          - type: map_at_5
            value: 0.673
          - type: mrr_at_1
            value: 72
          - type: mrr_at_10
            value: 80.01899999999999
          - type: mrr_at_100
            value: 80.42099999999999
          - type: mrr_at_1000
            value: 80.42099999999999
          - type: mrr_at_3
            value: 78
          - type: mrr_at_5
            value: 79.4
          - type: ndcg_at_1
            value: 66
          - type: ndcg_at_10
            value: 56.041
          - type: ndcg_at_100
            value: 37.987
          - type: ndcg_at_1000
            value: 34.198
          - type: ndcg_at_3
            value: 60.23500000000001
          - type: ndcg_at_5
            value: 58.025999999999996
          - type: precision_at_1
            value: 72
          - type: precision_at_10
            value: 60.4
          - type: precision_at_100
            value: 38.940000000000005
          - type: precision_at_1000
            value: 16.106
          - type: precision_at_3
            value: 63.333
          - type: precision_at_5
            value: 61.6
          - type: recall_at_1
            value: 0.186
          - type: recall_at_10
            value: 1.458
          - type: recall_at_100
            value: 8.455
          - type: recall_at_1000
            value: 33.141999999999996
          - type: recall_at_3
            value: 0.461
          - type: recall_at_5
            value: 0.756
      - task:
          type: Retrieval
        dataset:
          type: webis-touche2020
          name: MTEB Touche2020
          config: default
          split: test
          revision: None
        metrics:
          - type: map_at_1
            value: 2.2849999999999997
          - type: map_at_10
            value: 6.909
          - type: map_at_100
            value: 11.231
          - type: map_at_1000
            value: 12.472
          - type: map_at_3
            value: 3.53
          - type: map_at_5
            value: 4.675
          - type: mrr_at_1
            value: 26.531
          - type: mrr_at_10
            value: 40.73
          - type: mrr_at_100
            value: 41.637
          - type: mrr_at_1000
            value: 41.647
          - type: mrr_at_3
            value: 34.354
          - type: mrr_at_5
            value: 38.741
          - type: ndcg_at_1
            value: 24.490000000000002
          - type: ndcg_at_10
            value: 19.17
          - type: ndcg_at_100
            value: 29.946
          - type: ndcg_at_1000
            value: 40.842
          - type: ndcg_at_3
            value: 19.088
          - type: ndcg_at_5
            value: 19.445999999999998
          - type: precision_at_1
            value: 26.531
          - type: precision_at_10
            value: 17.959
          - type: precision_at_100
            value: 6.468999999999999
          - type: precision_at_1000
            value: 1.351
          - type: precision_at_3
            value: 19.048000000000002
          - type: precision_at_5
            value: 19.592000000000002
          - type: recall_at_1
            value: 2.2849999999999997
          - type: recall_at_10
            value: 12.973
          - type: recall_at_100
            value: 40.239999999999995
          - type: recall_at_1000
            value: 73.247
          - type: recall_at_3
            value: 4.407
          - type: recall_at_5
            value: 6.908
      - task:
          type: Classification
        dataset:
          type: mteb/toxic_conversations_50k
          name: MTEB ToxicConversationsClassification
          config: default
          split: test
          revision: d7c0de2777da35d6aae2200a62c6e0e5af397c4c
        metrics:
          - type: accuracy
            value: 68.405
          - type: ap
            value: 13.9913678628558
          - type: f1
            value: 53.209691917560285
      - task:
          type: Classification
        dataset:
          type: mteb/tweet_sentiment_extraction
          name: MTEB TweetSentimentExtractionClassification
          config: default
          split: test
          revision: d604517c81ca91fe16a244d1248fc021f9ecee7a
        metrics:
          - type: accuracy
            value: 56.080928126768534
          - type: f1
            value: 56.36329965117965
      - task:
          type: Clustering
        dataset:
          type: mteb/twentynewsgroups-clustering
          name: MTEB TwentyNewsgroupsClustering
          config: default
          split: test
          revision: 6125ec4e24fa026cec8a478383ee943acfbd5449
        metrics:
          - type: v_measure
            value: 31.540976715818065
      - task:
          type: PairClassification
        dataset:
          type: mteb/twittersemeval2015-pairclassification
          name: MTEB TwitterSemEval2015
          config: default
          split: test
          revision: 70970daeab8776df92f5ea462b6173c0b46fd2d1
        metrics:
          - type: cos_sim_accuracy
            value: 82.90516778923526
          - type: cos_sim_ap
            value: 61.5394989621502
          - type: cos_sim_f1
            value: 58.02297689685646
          - type: cos_sim_precision
            value: 55.62817719680465
          - type: cos_sim_recall
            value: 60.633245382585756
          - type: dot_accuracy
            value: 78.95928950348691
          - type: dot_ap
            value: 48.61088896690895
          - type: dot_f1
            value: 51.0104674059488
          - type: dot_precision
            value: 42.00375490698071
          - type: dot_recall
            value: 64.93403693931398
          - type: euclidean_accuracy
            value: 82.476008821601
          - type: euclidean_ap
            value: 59.59406971314053
          - type: euclidean_f1
            value: 56.424962447084525
          - type: euclidean_precision
            value: 58.47721483158789
          - type: euclidean_recall
            value: 54.51187335092348
          - type: manhattan_accuracy
            value: 82.66078559933241
          - type: manhattan_ap
            value: 60.414321716856925
          - type: manhattan_f1
            value: 56.88221089348002
          - type: manhattan_precision
            value: 57.86026200873362
          - type: manhattan_recall
            value: 55.93667546174142
          - type: max_accuracy
            value: 82.90516778923526
          - type: max_ap
            value: 61.5394989621502
          - type: max_f1
            value: 58.02297689685646
      - task:
          type: PairClassification
        dataset:
          type: mteb/twitterurlcorpus-pairclassification
          name: MTEB TwitterURLCorpus
          config: default
          split: test
          revision: 8b6510b0b1fa4e4c4f879467980e9be563ec1cdf
        metrics:
          - type: cos_sim_accuracy
            value: 85.71622618077386
          - type: cos_sim_ap
            value: 77.72774861009667
          - type: cos_sim_f1
            value: 71.40275165062152
          - type: cos_sim_precision
            value: 68.53359767754726
          - type: cos_sim_recall
            value: 74.52263627964275
          - type: dot_accuracy
            value: 83.97174680793262
          - type: dot_ap
            value: 72.89480417427734
          - type: dot_f1
            value: 68.57803792366198
          - type: dot_precision
            value: 62.94151708164447
          - type: dot_recall
            value: 75.32337542346782
          - type: euclidean_accuracy
            value: 84.88570652384834
          - type: euclidean_ap
            value: 75.78371710915128
          - type: euclidean_f1
            value: 69.44268877569989
          - type: euclidean_precision
            value: 67.1435761018046
          - type: euclidean_recall
            value: 71.90483523252233
          - type: manhattan_accuracy
            value: 85.6114409904141
          - type: manhattan_ap
            value: 77.38579436755944
          - type: manhattan_f1
            value: 70.8608538430316
          - type: manhattan_precision
            value: 68.03656203500319
          - type: manhattan_recall
            value: 73.92978133661842
          - type: max_accuracy
            value: 85.71622618077386
          - type: max_ap
            value: 77.72774861009667
          - type: max_f1
            value: 71.40275165062152

LLM2Vec: Large Language Models Are Secretly Powerful Text Encoders

LLM2Vec is a simple recipe to convert decoder-only LLMs into text encoders. It consists of 3 simple steps: 1) enabling bidirectional attention, 2) masked next token prediction, and 3) unsupervised contrastive learning. The model can be further fine-tuned to achieve state-of-the-art performance.

Installation

pip install llm2vec

Usage

from llm2vec import LLM2Vec

import torch
from transformers import AutoTokenizer, AutoModel, AutoConfig
from peft import PeftModel

# Loading base Mistral model, along with custom code that enables bidirectional connections in decoder-only LLMs. MNTP LoRA weights are merged into the base model.
tokenizer = AutoTokenizer.from_pretrained(
    "McGill-NLP/LLM2Vec-Sheared-LLaMA-mntp"
)
config = AutoConfig.from_pretrained(
    "McGill-NLP/LLM2Vec-Sheared-LLaMA-mntp", trust_remote_code=True
)
model = AutoModel.from_pretrained(
    "McGill-NLP/LLM2Vec-Sheared-LLaMA-mntp",
    trust_remote_code=True,
    config=config,
    torch_dtype=torch.bfloat16,
    device_map="cuda" if torch.cuda.is_available() else "cpu",
)
model = PeftModel.from_pretrained(
    model,
    "McGill-NLP/LLM2Vec-Sheared-LLaMA-mntp",
)
model = model.merge_and_unload()  # This can take several minutes on cpu

# Loading unsupervised SimCSE model. This loads the trained LoRA weights on top of MNTP model. Hence the final weights are -- Base model + MNTP (LoRA) + SimCSE (LoRA).
model = PeftModel.from_pretrained(
    model, "McGill-NLP/LLM2Vec-Sheared-LLaMA-mntp-unsup-simcse"
)

# Wrapper for encoding and pooling operations
l2v = LLM2Vec(model, tokenizer, pooling_mode="mean", max_length=512)

# Encoding queries using instructions
instruction = (
    "Given a web search query, retrieve relevant passages that answer the query:"
)
queries = [
    [instruction, "how much protein should a female eat"],
    [instruction, "summit define"],
]
q_reps = l2v.encode(queries)

# Encoding documents. Instruction are not required for documents
documents = [
    "As a general guideline, the CDC's average requirement of protein for women ages 19 to 70 is 46 grams per day. But, as you can see from this chart, you'll need to increase that if you're expecting or training for a marathon. Check out the chart below to see how much protein you should be eating each day.",
    "Definition of summit for English Language Learners. : 1  the highest point of a mountain : the top of a mountain. : 2  the highest level. : 3  a meeting or series of meetings between the leaders of two or more governments.",
]
d_reps = l2v.encode(documents)

# Compute cosine similarity
q_reps_norm = torch.nn.functional.normalize(q_reps, p=2, dim=1)
d_reps_norm = torch.nn.functional.normalize(d_reps, p=2, dim=1)
cos_sim = torch.mm(q_reps_norm, d_reps_norm.transpose(0, 1))

print(cos_sim)
"""
tensor([[0.5964, 0.1270],
        [0.0698, 0.2394]])
"""

Questions

If you have any question about the code, feel free to email Parishad (parishad.behnamghader@mila.quebec) and Vaibhav (vaibhav.adlakha@mila.quebec).