モデル概要

高性能なテキスト埋め込みモデルで、テキストを高品質なベクトル表現に変換でき、様々な自然言語処理タスクに適用可能

モデル特徴

マルチタスク高性能

分類、クラスタリング、検索など様々なNLPタスクで優れた性能を発揮

汎用埋め込み能力

高品質なテキストベクトル表現を生成でき、様々な下流タスクに適用可能

意味理解

テキストの意味情報を正確に捉え、意味的類似度タスクで優れた性能を発揮

モデル能力

テキスト分類

テキストクラスタリング

情報検索

意味的類似度計算

テキスト再ランキング

使用事例

電子商取引

製品レビュー分類

Amazon製品レビューを分類

AmazonPolarityデータセットで97.31%の精度を達成

反事実的レビュー検出

Amazon上の反事実的レビューを識別

AmazonCounterfactualデータセットで92.72%の精度を達成

金融

銀行顧客問い合わせ分類

銀行顧客問い合わせを分類

Banking77データセットで90.02%の精度を達成

学術研究

学術論文クラスタリング

arXivとbiorxiv論文をクラスタリング

arXivClusteringP2Pでv_measure54.02を達成

tags:

mteb
sentence-transformers
transformers model-index:
name: Salesforce/SFR-Embedding-2_R results:
- task: type: Classification dataset: type: mteb/amazon_counterfactual name: MTEB AmazonCounterfactualClassification (en) config: en split: test revision: e8379541af4e31359cca9fbcf4b00f2671dba205 metrics:
  - type: accuracy value: 92.71641791044776
  - type: ap value: 69.47931007147756
  - type: f1 value: 88.0252625393374
- task: type: Classification dataset: type: mteb/amazon_polarity name: MTEB AmazonPolarityClassification config: default split: test revision: e2d317d38cd51312af73b3d32a06d1a08b442046 metrics:
  - type: accuracy value: 97.31075
  - type: ap value: 96.26693923450127
  - type: f1 value: 97.31042448894502
- task: type: Classification dataset: type: mteb/amazon_reviews_multi name: MTEB AmazonReviewsClassification (en) config: en split: test revision: 1399c76144fd37290681b995c656ef9b2e06e26d metrics:
  - type: accuracy value: 61.040000000000006
  - type: f1 value: 60.78646832640785
- task: type: Retrieval dataset: type: mteb/arguana name: MTEB ArguAna config: default split: test revision: c22ab2a51041ffd869aaddef7af8d8215647e41a metrics:
  - type: map_at_1 value: 37.767
  - type: map_at_10 value: 53.908
  - type: map_at_100 value: 54.583000000000006
  - type: map_at_1000 value: 54.583999999999996
  - type: map_at_20 value: 54.50899999999999
  - type: map_at_3 value: 49.514
  - type: map_at_5 value: 52.059999999999995
  - type: mrr_at_1 value: 38.26458036984353
  - type: mrr_at_10 value: 54.120408001987066
  - type: mrr_at_100 value: 54.780719904297406
  - type: mrr_at_1000 value: 54.78174226698592
  - type: mrr_at_20 value: 54.706604527160295
  - type: mrr_at_3 value: 49.71550497866294
  - type: mrr_at_5 value: 52.247510668563436
  - type: ndcg_at_1 value: 37.767
  - type: ndcg_at_10 value: 62.339999999999996
  - type: ndcg_at_100 value: 64.89399999999999
  - type: ndcg_at_1000 value: 64.914
  - type: ndcg_at_20 value: 64.402
  - type: ndcg_at_3 value: 53.33
  - type: ndcg_at_5 value: 57.93899999999999
  - type: precision_at_1 value: 37.767
  - type: precision_at_10 value: 8.905000000000001
  - type: precision_at_100 value: 0.9950000000000001
  - type: precision_at_1000 value: 0.1
  - type: precision_at_20 value: 4.8469999999999995
  - type: precision_at_3 value: 21.456
  - type: precision_at_5 value: 15.121
  - type: recall_at_1 value: 37.767
  - type: recall_at_10 value: 89.047
  - type: recall_at_100 value: 99.502
  - type: recall_at_1000 value: 99.644
  - type: recall_at_20 value: 96.942
  - type: recall_at_3 value: 64.36699999999999
  - type: recall_at_5 value: 75.605
- task: type: Clustering dataset: type: mteb/arxiv-clustering-p2p name: MTEB ArxivClusteringP2P config: default split: test revision: a122ad7f3f0291bf49cc6f4d32aa80929df69d5d metrics:
  - type: v_measure value: 54.024325012036314
- task: type: Clustering dataset: type: mteb/arxiv-clustering-s2s name: MTEB ArxivClusteringS2S config: default split: test revision: f910caf1a6075f7329cdf8c1a6135696f37dbd53 metrics:
  - type: v_measure value: 48.817300846601675
- task: type: Reranking dataset: type: mteb/askubuntudupquestions-reranking name: MTEB AskUbuntuDupQuestions config: default split: test revision: 2000358ca161889fa9c082cb41daa8dcfb161a54 metrics:
  - type: map value: 66.71478959728732
  - type: mrr value: 79.07202216066482
- task: type: STS dataset: type: mteb/biosses-sts name: MTEB BIOSSES config: default split: test revision: d3fb88f8f02e40887cd149695127462bbcf29b4a metrics:
  - type: cos_sim_pearson value: 88.79517914982239
  - type: cos_sim_spearman value: 87.60440576436838
  - type: euclidean_pearson value: 87.75596873521118
  - type: euclidean_spearman value: 87.60440576436838
  - type: manhattan_pearson value: 87.74113773865973
  - type: manhattan_spearman value: 87.50560833247899
- task: type: Classification dataset: type: mteb/banking77 name: MTEB Banking77Classification config: default split: test revision: 0fd18e25b25c072e09e0d92ab615fda904d66300 metrics:
  - type: accuracy value: 90.02272727272727
  - type: f1 value: 89.96681880265936
- task: type: Clustering dataset: type: mteb/biorxiv-clustering-p2p name: MTEB BiorxivClusteringP2P config: default split: test revision: 65b79d1d13f80053f67aca9498d9402c2d9f1f40 metrics:
  - type: v_measure value: 50.75930389699286
- task: type: Clustering dataset: type: mteb/biorxiv-clustering-s2s name: MTEB BiorxivClusteringS2S config: default split: test revision: 258694dd0231531bc1fd9de6ceb52a0853c6d908 metrics:
  - type: v_measure value: 46.57286439805565
- task: type: Retrieval dataset: type: mteb/cqadupstack name: MTEB CQADupstackRetrieval config: default split: test revision: 65ac3a16b8e91f9cee4c9828cc7c335575432a2a metrics:
  - type: map_at_1 value: 28.056666666666665
  - type: map_at_10 value: 39.61749999999999
  - type: map_at_100 value: 41.00666666666666
  - type: map_at_1000 value: 41.11358333333334
  - type: map_at_20 value: 40.410250000000005
  - type: map_at_3 value: 35.98591666666667
  - type: map_at_5 value: 38.02
  - type: mrr_at_1 value: 33.73950708467142
  - type: mrr_at_10 value: 44.0987162763402
  - type: mrr_at_100 value: 44.94302678553521
  - type: mrr_at_1000 value: 44.98758207055161
  - type: mrr_at_20 value: 44.61156907536121
  - type: mrr_at_3 value: 41.247253732468415
  - type: mrr_at_5 value: 42.84859071071954
  - type: ndcg_at_1 value: 33.739666666666665
  - type: ndcg_at_10 value: 46.10683333333334
  - type: ndcg_at_100 value: 51.49275000000001
  - type: ndcg_at_1000 value: 53.2585
  - type: ndcg_at_20 value: 48.349
  - type: ndcg_at_3 value: 40.12416666666667
  - type: ndcg_at_5 value: 42.94783333333333
  - type: precision_at_1 value: 33.739666666666665
  - type: precision_at_10 value: 8.46025
  - type: precision_at_100 value: 1.3215833333333333
  - type: precision_at_1000 value: 0.16524999999999998
  - type: precision_at_20 value: 4.9935833333333335
  - type: precision_at_3 value: 19.00516666666667
  - type: precision_at_5 value: 13.72141666666667
  - type: recall_at_1 value: 28.056666666666665
  - type: recall_at_10 value: 60.68825000000001
  - type: recall_at_100 value: 83.74433333333334
  - type: recall_at_1000 value: 95.62299999999999
  - type: recall_at_20 value: 68.77641666666668
  - type: recall_at_3 value: 44.06991666666667
  - type: recall_at_5 value: 51.324999999999996
- task: type: Retrieval dataset: type: mteb/climate-fever name: MTEB ClimateFEVER config: default split: test revision: 47f2ac6acb640fc46020b02a5b59fdda04d39380 metrics:
  - type: map_at_1 value: 15.609
  - type: map_at_10 value: 25.584
  - type: map_at_100 value: 27.291999999999998
  - type: map_at_1000 value: 27.471
  - type: map_at_20 value: 26.497
  - type: map_at_3 value: 21.61
  - type: map_at_5 value: 23.76
  - type: mrr_at_1 value: 34.98371335504886
  - type: mrr_at_10 value: 45.73747479447807
  - type: mrr_at_100 value: 46.4973410206458
  - type: mrr_at_1000 value: 46.53372527933685
  - type: mrr_at_20 value: 46.19978503202757
  - type: mrr_at_3 value: 42.85559174809991
  - type: mrr_at_5 value: 44.65038002171556
  - type: ndcg_at_1 value: 34.984
  - type: ndcg_at_10 value: 34.427
  - type: ndcg_at_100 value: 40.908
  - type: ndcg_at_1000 value: 44.118
  - type: ndcg_at_20 value: 36.885
  - type: ndcg_at_3 value: 29.09
  - type: ndcg_at_5 value: 30.979
  - type: precision_at_1 value: 34.984
  - type: precision_at_10 value: 10.476
  - type: precision_at_100 value: 1.748
  - type: precision_at_1000 value: 0.23500000000000001
  - type: precision_at_20 value: 6.313000000000001
  - type: precision_at_3 value: 21.39
  - type: precision_at_5 value: 16.378
  - type: recall_at_1 value: 15.609
  - type: recall_at_10 value: 39.619
  - type: recall_at_100 value: 61.952
  - type: recall_at_1000 value: 79.861
  - type: recall_at_20 value: 46.489000000000004
  - type: recall_at_3 value: 26.134
  - type: recall_at_5 value: 31.955
- task: type: Retrieval dataset: type: mteb/dbpedia name: MTEB DBPedia config: default split: test revision: c0f706b76e590d620bd6618b3ca8efdd34e2d659 metrics:
  - type: map_at_1 value: 10.482
  - type: map_at_10 value: 25.155
  - type: map_at_100 value: 36.606
  - type: map_at_1000 value: 38.617000000000004
  - type: map_at_20 value: 29.676000000000002
  - type: map_at_3 value: 16.881
  - type: map_at_5 value: 20.043
  - type: mrr_at_1 value: 76.0
  - type: mrr_at_10 value: 82.5610119047619
  - type: mrr_at_100 value: 82.74795937825128
  - type: mrr_at_1000 value: 82.75526942226163
  - type: mrr_at_20 value: 82.70580357142858
  - type: mrr_at_3 value: 81.41666666666667
  - type: mrr_at_5 value: 82.26666666666667
  - type: ndcg_at_1 value: 63.625
  - type: ndcg_at_10 value: 51.214000000000006
  - type: ndcg_at_100 value: 56.411
  - type: ndcg_at_1000 value: 63.429
  - type: ndcg_at_20 value: 50.595
  - type: ndcg_at_3 value: 54.989
  - type: ndcg_at_5 value: 52.589
  - type: precision_at_1 value: 76.0
  - type: precision_at_10 value: 41.975
  - type: precision_at_100 value: 13.26
  - type: precision_at_1000 value: 2.493
  - type: precision_at_20 value: 32.15
  - type: precision_at_3 value: 59.0
  - type: precision_at_5 value: 51.24999999999999
  - type: recall_at_1 value: 10.482
  - type: recall_at_10 value: 31.075000000000003
  - type: recall_at_100 value: 63.119
  - type: recall_at_1000 value: 85.32300000000001
  - type: recall_at_20 value: 40.345
  - type: recall_at_3 value: 17.916
  - type: recall_at_5 value: 22.475
- task: type: Classification dataset: type: mteb/emotion name: MTEB EmotionClassification config: default split: test revision: 4f58c6b202a23cf9a4da393831edf4f9183cad37 metrics:
  - type: accuracy value: 93.36500000000001
  - type: f1 value: 89.89541440183861
- task: type: Retrieval dataset: type: mteb/fever name: MTEB FEVER config: default split: test revision: bea83ef9e8fb933d90a2f1d5515737465d613e12 metrics:
  - type: map_at_1 value: 81.948
  - type: map_at_10 value: 89.47500000000001
  - type: map_at_100 value: 89.66199999999999
  - type: map_at_1000 value: 89.671
  - type: map_at_20 value: 89.582
  - type: map_at_3 value: 88.646
  - type: map_at_5 value: 89.19
  - type: mrr_at_1 value: 88.23882388238825
  - type: mrr_at_10 value: 93.2122736083131
  - type: mrr_at_100 value: 93.23908769526588
  - type: mrr_at_1000 value: 93.23932393435209
  - type: mrr_at_20 value: 93.23217832106207
  - type: mrr_at_3 value: 92.98679867986787
  - type: mrr_at_5 value: 93.16906690669056
  - type: ndcg_at_1 value: 88.239
  - type: ndcg_at_10 value: 92.155
  - type: ndcg_at_100 value: 92.735
  - type: ndcg_at_1000 value: 92.866
  - type: ndcg_at_20 value: 92.39699999999999
  - type: ndcg_at_3 value: 91.188
  - type: ndcg_at_5 value: 91.754
  - type: precision_at_1 value: 88.239
  - type: precision_at_10 value: 10.903
  - type: precision_at_100 value: 1.147
  - type: precision_at_1000 value: 0.117
  - type: precision_at_20 value: 5.5440000000000005
  - type: precision_at_3 value: 34.598
  - type: precision_at_5 value: 21.302
  - type: recall_at_1 value: 81.948
  - type: recall_at_10 value: 96.518
  - type: recall_at_100 value: 98.646
  - type: recall_at_1000 value: 99.399
  - type: recall_at_20 value: 97.262
  - type: recall_at_3 value: 93.89800000000001
  - type: recall_at_5 value: 95.38600000000001
- task: type: Retrieval dataset: type: mteb/fiqa name: MTEB FiQA2018 config: default split: test revision: 27a168819829fe9bcd655c2df245fb19452e8e06 metrics:
  - type: map_at_1 value: 32.033
  - type: map_at_10 value: 53.55
  - type: map_at_100 value: 55.672
  - type: map_at_1000 value: 55.764
  - type: map_at_20 value: 54.87800000000001
  - type: map_at_3 value: 46.761
  - type: map_at_5 value: 50.529
  - type: mrr_at_1 value: 60.95679012345679
  - type: mrr_at_10 value: 68.70835782872815
  - type: mrr_at_100 value: 69.21918402444501
  - type: mrr_at_1000 value: 69.23608783148705
  - type: mrr_at_20 value: 69.07497388036454
  - type: mrr_at_3 value: 66.76954732510285
  - type: mrr_at_5 value: 67.95781893004109
  - type: ndcg_at_1 value: 60.956999999999994
  - type: ndcg_at_10 value: 61.766
  - type: ndcg_at_100 value: 67.652
  - type: ndcg_at_1000 value: 68.94500000000001
  - type: ndcg_at_20 value: 64.48700000000001
  - type: ndcg_at_3 value: 57.25
  - type: ndcg_at_5 value: 58.64
  - type: precision_at_1 value: 60.956999999999994
  - type: precision_at_10 value: 17.083000000000002
  - type: precision_at_100 value: 2.346
  - type: precision_at_1000 value: 0.257
  - type: precision_at_20 value: 9.807
  - type: precision_at_3 value: 38.477
  - type: precision_at_5 value: 27.962999999999997
  - type: recall_at_1 value: 32.033
  - type: recall_at_10 value: 69.44
  - type: recall_at_100 value: 90.17500000000001
  - type: recall_at_1000 value: 97.90100000000001
  - type: recall_at_20 value: 77.629
  - type: recall_at_3 value: 51.664
  - type: recall_at_5 value: 59.565
- task: type: Retrieval dataset: type: mteb/hotpotqa name: MTEB HotpotQA config: default split: test revision: ab518f4d6fcca38d87c25209f94beba119d02014 metrics:
  - type: map_at_1 value: 42.741
  - type: map_at_10 value: 74.811
  - type: map_at_100 value: 75.508
  - type: map_at_1000 value: 75.541
  - type: map_at_20 value: 75.25699999999999
  - type: map_at_3 value: 71.31
  - type: map_at_5 value: 73.69
  - type: mrr_at_1 value: 85.48278190411884
  - type: mrr_at_10 value: 90.20347684425987
  - type: mrr_at_100 value: 90.29734129342121
  - type: mrr_at_1000 value: 90.30017606259217
  - type: mrr_at_20 value: 90.27225310310567
  - type: mrr_at_3 value: 89.67364393427842
  - type: mrr_at_5 value: 90.02408282691847
  - type: ndcg_at_1 value: 85.483
  - type: ndcg_at_10 value: 81.361
  - type: ndcg_at_100 value: 83.588
  - type: ndcg_at_1000 value: 84.19
  - type: ndcg_at_20 value: 82.42699999999999
  - type: ndcg_at_3 value: 76.779
  - type: ndcg_at_5 value: 79.581
  - type: precision_at_1 value: 85.483
  - type: precision_at_10 value: 17.113
  - type: precision_at_100 value: 1.882
  - type: precision_at_1000 value: 0.196
  - type: precision_at_20 value: 8.899
  - type: precision_at_3 value: 50.397999999999996
  - type: precision_at_5 value: 32.443
  - type: recall_at_1 value: 42.741
  - type: recall_at_10 value: 85.564
  - type: recall_at_100 value: 94.07799999999999
  - type: recall_at_1000 value: 97.995
  - type: recall_at_20 value: 88.98700000000001
  - type: recall_at_3 value: 75.598
  - type: recall_at_5 value: 81.107
- task: type: Classification dataset: type: mteb/imdb name: MTEB ImdbClassification config: default split: test revision: 3d86128a09e091d6018b6d26cad27f2739fc2db7 metrics:
  - type: accuracy value: 96.80320000000002
  - type: ap value: 94.98856145360044
  - type: f1 value: 96.80287885839178
- task: type: Retrieval dataset: type: mteb/msmarco name: MTEB MSMARCO config: default split: dev revision: c5a29a104738b98a9e76336939199e264163d4a0 metrics:
  - type: map_at_1 value: 22.539
  - type: map_at_10 value: 35.109
  - type: map_at_100 value: 36.287000000000006
  - type: map_at_1000 value: 36.335
  - type: map_at_20 value: 35.838
  - type: map_at_3 value: 31.11
  - type: map_at_5 value: 33.455
  - type: mrr_at_1 value: 23.15186246418338
  - type: mrr_at_10 value: 35.70532018920268
  - type: mrr_at_100 value: 36.815167506137584
  - type: mrr_at_1000 value: 36.85695349443505
  - type: mrr_at_20 value: 36.39500867880642
  - type: mrr_at_3 value: 31.81232091690535
  - type: mrr_at_5 value: 34.096704871060155
  - type: ndcg_at_1 value: 23.152
  - type: ndcg_at_10 value: 42.181999999999995
  - type: ndcg_at_100 value: 47.847
  - type: ndcg_at_1000 value: 48.988
  - type: ndcg_at_20 value: 44.767
  - type: ndcg_at_3 value: 34.088
  - type: ndcg_at_5 value: 38.257999999999996
  - type: precision_at_1 value: 23.152
  - type: precision_at_10 value: 6.678000000000001
  - type: precision_at_100 value: 0.9530000000000001
  - type: precision_at_1000 value: 0.105
  - type: precision_at_20 value: 3.881
  - type: precision_at_3 value: 14.518
  - type: precision_at_5 value: 10.831
  - type: recall_at_1 value: 22.539
  - type: recall_at_10 value: 63.965
  - type: recall_at_100 value: 90.129
  - type: recall_at_1000 value: 98.721
  - type: recall_at_20 value: 74.00999999999999
  - type: recall_at_3 value: 42.004999999999995
  - type: recall_at_5 value: 52.028
- task: type: Classification dataset: type: mteb/mtop_domain name: MTEB MTOPDomainClassification (en) config: en split: test revision: d80d48c1eb48d3562165c59d59d0034df9fff0bf metrics:
  - type: accuracy value: 98.5750113999088
  - type: f1 value: 98.41576079230245
- task: type: Classification dataset: type: mteb/mtop_intent name: MTEB MTOPIntentClassification (en) config: en split: test revision: ae001d0e6b1228650b7bd1c2c65fb50ad11a8aba metrics:
  - type: accuracy value: 91.29502963976289
  - type: f1 value: 74.84400169335184
- task: type: Classification dataset: type: mteb/amazon_massive_intent name: MTEB MassiveIntentClassification (en) config: en split: test revision: 31efe3c427b0bae9c22cbb560b8f15491cc6bed7 metrics:
  - type: accuracy value: 85.96839273705447
  - type: f1 value: 82.43129186593926
- task: type: Classification dataset: type: mteb/amazon_massive_scenario name: MTEB MassiveScenarioClassification (en) config: en split: test revision: 7d571f92784cd94a019292a1f45445077d0ef634 metrics:
  - type: accuracy value: 90.60860793544047
  - type: f1 value: 89.79415994859477
- task: type: Clustering dataset: type: mteb/medrxiv-clustering-p2p name: MTEB MedrxivClusteringP2P config: default split: test revision: e7a26af6f3ae46b30dde8737f02c07b1505bcc73 metrics:
  - type: v_measure value: 46.661892807041355
- task: type: Clustering dataset: type: mteb/medrxiv-clustering-s2s name: MTEB MedrxivClusteringS2S config: default split: test revision: 35191c8c0dca72d8ff3efcd72aa802307d469663 metrics:
  - type: v_measure value: 44.17598473858937
- task: type: Reranking dataset: type: mteb/mind_small name: MTEB MindSmallReranking config: default split: test revision: 59042f120c80e8afa9cdbb224f67076cec0fc9a7 metrics:
  - type: map value: 31.260919294024603
  - type: mrr value: 32.37049108835034
- task: type: Retrieval dataset: type: mteb/nfcorpus name: MTEB NFCorpus config: default split: test revision: ec0fa4fe99da2ff19ca1214b7966684033a58814 metrics:
  - type: map_at_1 value: 6.672000000000001
  - type: map_at_10 value: 15.972
  - type: map_at_100 value: 20.94
  - type: map_at_1000 value: 22.877
  - type: map_at_20 value: 17.986
  - type: map_at_3 value: 11.161
  - type: map_at_5 value: 13.293
  - type: mrr_at_1 value: 53.56037151702786
  - type: mrr_at_10 value: 61.915696103002595
  - type: mrr_at_100 value: 62.4130902631107
  - type: mrr_at_1000 value: 62.45228087711845
  - type: mrr_at_20 value: 62.1983715004112
  - type: mrr_at_3 value: 60.31991744066049
  - type: mrr_at_5 value: 61.27966976264191
  - type: ndcg_at_1 value: 50.929
  - type: ndcg_at_10 value: 41.336
  - type: ndcg_at_100 value: 38.586999999999996
  - type: ndcg_at_1000 value: 48.155
  - type: ndcg_at_20 value: 38.888
  - type: ndcg_at_3 value: 47.0
  - type: ndcg_at_5 value: 44.335
  - type: precision_at_1 value: 53.251000000000005
  - type: precision_at_10 value: 31.146
  - type: precision_at_100 value: 10.040000000000001
  - type: precision_at_1000 value: 2.432
  - type: precision_at_20 value: 23.421
  - type: precision_at_3 value: 45.098
  - type: precision_at_5 value: 39.071
  - type: recall_at_1 value: 6.672000000000001
  - type: recall_at_10 value: 20.764
  - type: recall_at_100 value: 40.759
  - type: recall_at_1000 value: 75.015
  - type: recall_at_20 value: 25.548
  - type: recall_at_3 value: 12.328
  - type: recall_at_5 value: 15.601999999999999
- task: type: Retrieval dataset: type: mteb/nq name: MTEB NQ config: default split: test revision: b774495ed302d8c44a3a7ea25c90dbce03968f31 metrics:
  - type: map_at_1 value: 50.944
  - type: map_at_10 value: 67.565
  - type: map_at_100 value: 68.10300000000001
  - type: map_at_1000 value: 68.109
  - type: map_at_20 value: 67.973
  - type: map_at_3 value: 64.176
  - type: map_at_5 value: 66.39699999999999
  - type: mrr_at_1 value: 57.01042873696408
  - type: mrr_at_10 value: 69.76629605105849
  - type: mrr_at_100 value: 70.09927347130204
  - type: mrr_at_1000 value: 70.10309675839956
  - type: mrr_at_20 value: 70.02288627712392
  - type: mrr_at_3 value: 67.46813441483191
  - type: mrr_at_5 value: 68.93105446118189
  - type: ndcg_at_1 value: 57.010000000000005
  - type: ndcg_at_10 value: 73.956
  - type: ndcg_at_100 value: 75.90299999999999
  - type: ndcg_at_1000 value: 76.03999999999999
  - type: ndcg_at_20 value: 75.17
  - type: ndcg_at_3 value: 68.13900000000001
  - type: ndcg_at_5 value: 71.532
  - type: precision_at_1 value: 57.010000000000005
  - type: precision_at_10 value: 10.91
  - type: precision_at_100 value: 1.2
  - type: precision_at_1000 value: 0.121
  - type: precision_at_20 value: 5.753
  - type: precision_at_3 value: 29.828
  - type: precision_at_5 value: 19.971
  - type: recall_at_1 value: 50.944
  - type: recall_at_10 value: 90.754
  - type: recall_at_100 value: 98.699
  - type: recall_at_1000 value: 99.701
  - type: recall_at_20 value: 95.148
  - type: recall_at_3 value: 76.224
  - type: recall_at_5 value: 83.872
- task: type: Retrieval dataset: type: mteb/quora name: MTEB QuoraRetrieval config: default split: test revision: e4e08e0b7dbe3c8700f0daef558ff32256715259 metrics:
  - type: map_at_1 value: 71.856
  - type: map_at_10 value: 86.077
  - type: map_at_100 value: 86.696
  - type: map_at_1000 value: 86.708
  - type: map_at_20 value: 86.493
  - type: map_at_3 value: 83.176
  - type: map_at_5 value: 85.008
  - type: mrr_at_1 value: 82.74000000000001
  - type: mrr_at_10 value: 88.68947222222207
  - type: mrr_at_100 value: 88.78196949571182
  - type: mrr_at_1000 value: 88.78223256200576
  - type: mrr_at_20 value: 88.76455636228219
  - type: mrr_at_3 value: 87.85833333333316
  - type: mrr_at_5 value: 88.43933333333311
  - type: ndcg_at_1 value: 82.74000000000001
  - type: ndcg_at_10 value: 89.583
  - type: ndcg_at_100 value: 90.652
  - type: ndcg_at_1000 value: 90.711
  - type: ndcg_at_20 value: 90.203
  - type: ndcg_at_3 value: 86.967
  - type: ndcg_at_5 value: 88.43299999999999
  - type: precision_at_1 value: 82.74000000000001
  - type: precision_at_10 value: 13.617
  - type: precision_at_100 value: 1.542
  - type: precision_at_1000 value: 0.157
  - type: precision_at_20 value: 7.217999999999999
  - type: precision_at_3 value: 38.163000000000004
  - type: precision_at_5 value: 25.05
  - type: recall_at_1 value: 71.856
  - type: recall_at_10 value: 96.244
  - type: recall_at_100 value: 99.773
  - type: recall_at_1000 value: 99.99900000000001
  - type: recall_at_20 value: 98.221
  - type: recall_at_3 value: 88.715
  - type: recall_at_5 value: 92.88499999999999
- task: type: Clustering dataset: type: mteb/reddit-clustering name: MTEB RedditClustering config: default split: test revision: 24640382cdbf8abc73003fb0fa6d111a705499eb metrics:
  - type: v_measure value: 62.91969510127886
- task: type: Clustering dataset: type: mteb/reddit-clustering-p2p name: MTEB RedditClusteringP2P config: default split: test revision: 385e3cb46b4cfa89021f56c4380204149d0efe33 metrics:
  - type: v_measure value: 72.74201090913765
- task: type: Retrieval dataset: type: mteb/scidocs name: MTEB SCIDOCS config: default split: test revision: f8c2fcf00f625baaa80f62ec5bd9e1fff3b8ae88 metrics:
  - type: map_at_1 value: 5.8229999999999995
  - type: map_at_10 value: 15.152
  - type: map_at_100 value: 17.936
  - type: map_at_1000 value: 18.292
  - type: map_at_20 value: 16.526
  - type: map_at_3 value: 10.294
  - type: map_at_5 value: 12.794
  - type: mrr_at_1 value: 28.599999999999998
  - type: mrr_at_10 value: 40.68206349206347
  - type: mrr_at_100 value: 41.673752995361795
  - type: mrr_at_1000 value: 41.71500072915374
  - type: mrr_at_20 value: 41.28552805166964
  - type: mrr_at_3 value: 36.84999999999998
  - type: mrr_at_5 value: 39.19999999999995
  - type: ndcg_at_1 value: 28.599999999999998
  - type: ndcg_at_10 value: 24.866
  - type: ndcg_at_100 value: 34.597
  - type: ndcg_at_1000 value: 39.994
  - type: ndcg_at_20 value: 28.309
  - type: ndcg_at_3 value: 22.749
  - type: ndcg_at_5 value: 20.502000000000002
  - type: precision_at_1 value: 28.599999999999998
  - type: precision_at_10 value: 13.089999999999998
  - type: precision_at_100 value: 2.7119999999999997
  - type: precision_at_1000 value: 0.39899999999999997
  - type: precision_at_20 value: 8.53
  - type: precision_at_3 value: 21.099999999999998
  - type: precision_at_5 value: 18.22
  - type: recall_at_1 value: 5.8229999999999995
  - type: recall_at_10 value: 26.522000000000002
  - type: recall_at_100 value: 55.003
  - type: recall_at_1000 value: 80.977
  - type: recall_at_20 value: 34.618
  - type: recall_at_3 value: 12.848
  - type: recall_at_5 value: 18.477
- task: type: STS dataset: type: mteb/sickr-sts name: MTEB SICK-R config: default split: test revision: 20a6d6f312dd54037fe07a32d58e5e168867909d metrics:
  - type: cos_sim_pearson value: 80.72562067620224
  - type: cos_sim_spearman value: 77.00710192931953
  - type: euclidean_pearson value: 78.65843289108192
  - type: euclidean_spearman value: 77.00710077709005
  - type: manhattan_pearson value: 78.48859522905846
  - type: manhattan_spearman value: 76.8213740840866
- task: type: STS dataset: type: mteb/sts12-sts name: MTEB STS12 config: default split: test revision: a0d554a64d88156834ff5ae9920b964011b16384 metrics:
  - type: cos_sim_pearson value: 81.15015325911659
  - type: cos_sim_spearman value: 75.67268325741222
  - type: euclidean_pearson value: 75.54004763633206
  - type: euclidean_spearman value: 75.67262179635058
  - type: manhattan_pearson value: 75.80681616893116
  - type: manhattan_spearman value: 75.93721016401406
- task: type: STS dataset: type: mteb/sts13-sts name: MTEB STS13 config: default split: test revision: 7e90230a92c190f1bf69ae9002b8cea547a64cca metrics:
  - type: cos_sim_pearson value: 81.71651874476737
  - type: cos_sim_spearman value: 82.39667472464997
  - type: euclidean_pearson value: 82.28256504757712
  - type: euclidean_spearman value: 82.39663674872656
  - type: manhattan_pearson value: 82.3192873176068
  - type: manhattan_spearman value: 82.41915252757059
- task: type: STS dataset: type: mteb/sts14-sts name: MTEB STS14 config: default split: test revision: 6031580fec1f6af667f0bd2da0a551cf4f0b2375 metrics:
  - type: cos_sim_pearson value: 81.222967367593
  - type: cos_sim_spearman value: 79.92685877403252
  - type: euclidean_pearson value: 79.95053542861498
  - type: euclidean_spearman value: 79.9268858850991
  - type: manhattan_pearson value: 79.90485851323321
  - type: manhattan_spearman value: 79.93878025669312
- task: type: STS dataset: type: mteb/sts15-sts name: MTEB STS15 config: default split: test revision: ae752c7c21bf194d8b67fd573edf7ae58183cbe3 metrics:
  - type: cos_sim_pearson value: 85.27539130156643
  - type: cos_sim_spearman value: 85.81645767911826
  - type: euclidean_pearson value: 85.5488615685444
  - type: euclidean_spearman value: 85.81647022566916
  - type: manhattan_pearson value: 85.6358149547879
  - type: manhattan_spearman value: 85.96347118567043
- task: type: STS dataset: type: mteb/sts16-sts name: MTEB STS16 config: default split: test revision: 4d8694f8f0e0100860b497b999b3dbed754a0513 metrics:
  - type: cos_sim_pearson value: 83.43727336154858
  - type: cos_sim_spearman value: 84.50468882202796
  - type: euclidean_pearson value: 83.23576727105372
  - type: euclidean_spearman value: 84.50468882202796
  - type: manhattan_pearson value: 83.28843314503176
  - type: manhattan_spearman value: 84.60383766214322
- task: type: STS dataset: type: mteb/sts17-crosslingual-sts name: MTEB STS17 (en-en) config: en-en split: test revision: faeb762787bd10488a50c8b5be4a3b82e411949c metrics:
  - type: cos_sim_pearson value: 88.86589365166874
  - type: cos_sim_spearman value: 88.93117996163835
  - type: euclidean_pearson value: 89.12271565981082
  - type: euclidean_spearman value: 88.93117996163835
  - type: manhattan_pearson value: 88.94419759325545
  - type: manhattan_spearman value: 88.63073561731899
- task: type: STS dataset: type: mteb/sts22-crosslingual-sts name: MTEB STS22 (en) config: en split: test revision: 6d1ba47164174a496b7fa5d3569dae26a6813b80 metrics:
  - type: cos_sim_pearson value: 67.96578378422929
  - type: cos_sim_spearman value: 67.10257461424345
  - type: euclidean_pearson value: 67.51317866195149
  - type: euclidean_spearman value: 67.10257461424345
  - type: manhattan_pearson value: 67.74940912013754
  - type: manhattan_spearman value: 67.46694183937207
- task: type: STS dataset: type: mteb/stsbenchmark-sts name: MTEB STSBenchmark config: default split: test revision: b0fddb56ed78048fa8b90373c8a3cfc37b684831 metrics:
  - type: cos_sim_pearson value: 83.55433725920493
  - type: cos_sim_spearman value: 83.60373857254014
  - type: euclidean_pearson value: 83.08086082334839
  - type: euclidean_spearman value: 83.6036864776559
  - type: manhattan_pearson value: 83.2232267589246
  - type: manhattan_spearman value: 83.78923946962664
- task: type: Reranking dataset: type: mteb/scidocs-reranking name: MTEB SciDocsRR config: default split: test revision: d3c5e1fc0b855ab6097bf1cda04dd73947d7caab metrics:
  - type: map value: 87.28566757174322
  - type: mrr value: 96.63827639317836
- task: type: Retrieval dataset: type: mteb/scifact name: MTEB SciFact config: default split: test revision: 0228b52cf27578f30900b9e5271d331663a030d7 metrics:
  - type: map_at_1 value: 70.661
  - type: map_at_10 value: 82.051
  - type: map_at_100 value: 82.162
  - type: map_at_1000 value: 82.167
  - type: map_at_20 value: 82.122
  - type: map_at_3 value: 79.919
  - type: map_at_5 value: 81.368
  - type: mrr_at_1 value: 74.33333333333333
  - type: mrr_at_10 value: 82.98452380952381
  - type: mrr_at_100 value: 83.09512420633841
  - type: mrr_at_1000 value: 83.10026279387446
  - type: mrr_at_20 value: 83.05460927960928
  - type: mrr_at_3 value: 81.8888888888889
  - type: mrr_at_5 value: 82.65555555555557
  - type: ndcg_at_1 value: 74.333
  - type: ndcg_at_10 value: 85.914
  - type: ndcg_at_100 value: 86.473
  - type: ndcg_at_1000 value: 86.602
  - type: ndcg_at_20 value: 86.169
  - type: ndcg_at_3 value: 83.047
  - type: ndcg_at_5 value: 84.72
  - type: precision_at_1 value: 74.333
  - type: precision_at_10 value: 10.933
  - type: precision_at_100 value: 1.1199999999999999
  - type: precision_at_1000 value: 0.11299999999999999
  - type: precision_at_20 value: 5.5169999999999995
  - type: precision_at_3 value: 32.444
  - type: precision_at_5 value: 20.8
  - type: recall_at_1 value: 70.661
  - type: recall_at_10 value: 96.333
  - type: recall_at_100 value: 99.0
  - type: recall_at_1000 value: 100.0
  - type: recall_at_20 value: 97.333
  - type: recall_at_3 value: 88.64999999999999
  - type: recall_at_5 value: 93.089
- task: type: PairClassification dataset: type: mteb/sprintduplicatequestions-pairclassification name: MTEB SprintDuplicateQuestions config: default split: test revision: d66bd1f72af766a5cc4b0ca5e00c162f89e8cc46 metrics:
  - type: cos_sim_accuracy value: 99.89108910891089
  - type: cos_sim_ap value: 97.61815451002174
  - type: cos_sim_f1 value: 94.51097804391219
  - type: cos_sim_precision value: 94.32270916334662
  - type: cos_sim_recall value: 94.69999999999999
  - type: dot_accuracy value: 99.89108910891089
  - type: dot_ap value: 97.61815451002175
  - type: dot_f1 value: 94.51097804391219
  - type: dot_precision value: 94.32270916334662
  - type: dot_recall value: 94.69999999999999
  - type: euclidean_accuracy value: 99.89108910891089
  - type: euclidean_ap value: 97.61815534251431
  - type: euclidean_f1 value: 94.51097804391219
  - type: euclidean_precision value: 94.32270916334662
  - type: euclidean_recall value: 94.69999999999999
  - type: manhattan_accuracy value: 99.8940594059406
  - type: manhattan_ap value: 97.66124472227202
  - type: manhattan_f1 value: 94.65267366316841
  - type: manhattan_precision value: 94.60539460539461
  - type: manhattan_recall value: 94.69999999999999
  - type: max_accuracy value: 99.8940594059406
  - type: max_ap value: 97.66124472227202
  - type: max_f1 value: 94.65267366316841
- task: type: Clustering dataset: type: mteb/stackexchange-clustering name: MTEB StackExchangeClustering config: default split: test revision: 6cbc1f7b2bc0622f2e39d2c77fa502909748c259 metrics:
  - type: v_measure value: 76.482776391195
- task: type: Clustering dataset: type: mteb/stackexchange-clustering-p2p name: MTEB StackExchangeClusteringP2P config: default split: test revision: 815ca46b2622cec33ccafc3735d572c266efdb44 metrics:
  - type: v_measure value: 48.29023235124473
- task: type: Reranking dataset: type: mteb/stackoverflowdupquestions-reranking name: MTEB StackOverflowDupQuestions config: default split: test revision: e185fbe320c72810689fc5848eb6114e1ef5ec69 metrics:
  - type: map value: 55.3190739691685
  - type: mrr value: 56.40441972243442
- task: type: Summarization dataset: type: mteb/summeval name: MTEB SummEval config: default split: test revision: cda12ad7615edc362dbf25a00fdd61d3b1eaf93c metrics:
  - type: cos_sim_pearson value: 31.98570594378664
  - type: cos_sim_spearman value: 30.712965330802174
  - type: dot_pearson value: 31.98570540209124
  - type: dot_spearman value: 30.712965330802174
- task: type: Retrieval dataset: type: mteb/trec-covid name: MTEB TRECCOVID config: default split: test revision: bb9466bac8153a0349341eb1b22e06409e78ef4e metrics:
  - type: map_at_1 value: 0.25
  - type: map_at_10 value: 2.2640000000000002
  - type: map_at_100 value: 14.447
  - type: map_at_1000 value: 35.452
  - type: map_at_20 value: 4.163
  - type: map_at_3 value: 0.715
  - type: map_at_5 value: 1.1780000000000002
  - type: mrr_at_1 value: 94.0
  - type: mrr_at_10 value: 96.66666666666667
  - type: mrr_at_100 value: 96.66666666666667
  - type: mrr_at_1000 value: 96.66666666666667
  - type: mrr_at_20 value: 96.66666666666667
  - type: mrr_at_3 value: 96.66666666666667
  - type: mrr_at_5 value: 96.66666666666667
  - type: ndcg_at_1 value: 92.0
  - type: ndcg_at_10 value: 87.26899999999999
  - type: ndcg_at_100 value: 68.586
  - type: ndcg_at_1000 value: 61.056999999999995
  - type: ndcg_at_20 value: 83.452
  - type: ndcg_at_3 value: 90.11200000000001
  - type: ndcg_at_5 value: 89.103
  - type: precision_at_1 value: 94.0
  - type: precision_at_10 value: 91.2
  - type: precision_at_100 value: 70.12
  - type: precision_at_1000 value: 26.773999999999997
  - type: precision_at_20 value: 87.3
  - type: precision_at_3 value: 92.667
  - type: precision_at_5 value: 92.4
  - type: recall_at_1 value: 0.25
  - type: recall_at_10 value: 2.3970000000000002
  - type: recall_at_100 value: 17.233999999999998
  - type: recall_at_1000 value: 57.879000000000005
  - type: recall_at_20 value: 4.508
  - type: recall_at_3 value: 0.734
  - type: recall_at_5 value: 1.2269999999999999
- task: type: Retrieval dataset: type: mteb/touche2020 name: MTEB Touche2020 config: default split: test revision: a34f9a33db75fa0cbb21bb5cfc3dae8dc8bec93f metrics:
  - type: map_at_1 value: 2.806
  - type: map_at_10 value: 11.369
  - type: map_at_100 value: 17.791
  - type: map_at_1000 value: 19.363
  - type: map_at_20 value: 14.038999999999998
  - type: map_at_3 value: 5.817
  - type: map_at_5 value: 8.331
  - type: mrr_at_1 value: 36.734693877551024
  - type: mrr_at_10 value: 53.355199222546155
  - type: mrr_at_100 value: 53.648197984932665
  - type: mrr_at_1000 value: 53.648197984932665
  - type: mrr_at_20 value: 53.500971817298336
  - type: mrr_at_3 value: 48.63945578231292
  - type: mrr_at_5 value: 51.29251700680272
  - type: ndcg_at_1 value: 35.714
  - type: ndcg_at_10 value: 28.18
  - type: ndcg_at_100 value: 39.22
  - type: ndcg_at_1000 value: 50.807
  - type: ndcg_at_20 value: 28.979
  - type: ndcg_at_3 value: 31.114000000000004
  - type: ndcg_at_5 value: 29.687
  - type: precision_at_1 value: 36.735
  - type: precision_at_10 value: 24.898
  - type: precision_at_100 value: 7.918
  - type: precision_at_1000 value: 1.5779999999999998
  - type: precision_at_20 value: 18.878
  - type: precision_at_3 value: 31.293
  - type: precision_at_5 value: 29.387999999999998
  - type: recall_at_1 value: 2.806
  - type: recall_at_10 value: 17.776
  - type: recall_at_100 value: 49.41
  - type: recall_at_1000 value: 84.97200000000001
  - type: recall_at_20 value: 26.589000000000002
  - type: recall_at_3 value: 6.866999999999999
  - type: recall_at_5 value: 10.964
- task: type: Classification dataset: type: mteb/toxic_conversations_50k name: MTEB ToxicConversationsClassification config: default split: test revision: edfaf9da55d3dd50d43143d90c1ac476895ae6de metrics:
  - type: accuracy value: 91.1376953125
  - type: ap value: 40.51219896084815
  - type: f1 value: 77.5195445434559
- task: type: Classification dataset: type: mteb/tweet_sentiment_extraction name: MTEB TweetSentimentExtractionClassification config: default split: test revision: d604517c81ca91fe16a244d1248fc021f9ecee7a metrics:
  - type: accuracy value: 79.69722693831352
  - type: f1 value: 80.02969178591319
- task: type: Clustering dataset: type: mteb/twentynewsgroups-clustering name: MTEB TwentyNewsgroupsClustering config: default split: test revision: 6125ec4e24fa026cec8a478383ee943acfbd5449 metrics:
  - type: v_measure value: 66.42427742893598
- task: type: PairClassification dataset: type: mteb/twittersemeval2015-pairclassification name: MTEB TwitterSemEval2015 config: default split: test revision: 70970daeab8776df92f5ea462b6173c0b46fd2d1 metrics:
  - type: cos_sim_accuracy value: 87.81069321094355
  - type: cos_sim_ap value: 78.57014017906349
  - type: cos_sim_f1 value: 72.38883143743536
  - type: cos_sim_precision value: 70.95793208312215
  - type: cos_sim_recall value: 73.87862796833772
  - type: dot_accuracy value: 87.81069321094355
  - type: dot_ap value: 78.5701399541226
  - type: dot_f1 value: 72.38883143743536
  - type: dot_precision value: 70.95793208312215
  - type: dot_recall value: 73.87862796833772
  - type: euclidean_accuracy value: 87.81069321094355
  - type: euclidean_ap value: 78.57015336777854
  - type: euclidean_f1 value: 72.38883143743536
  - type: euclidean_precision value: 70.95793208312215
  - type: euclidean_recall value: 73.87862796833772
  - type: manhattan_accuracy value: 87.57227156225785
  - type: manhattan_ap value: 78.19109731614216
  - type: manhattan_f1 value: 71.87819856704198
  - type: manhattan_precision value: 69.77148534525584
  - type: manhattan_recall value: 74.1160949868074
  - type: max_accuracy value: 87.81069321094355
  - type: max_ap value: 78.57015336777854
  - type: max_f1 value: 72.38883143743536
- task: type: PairClassification dataset: type: mteb/twitterurlcorpus-pairclassification name: MTEB TwitterURLCorpus config: default split: test revision: 8b6510b0b1fa4e4c4f879467980e9be563ec1cdf metrics:
  - type: cos_sim_accuracy value: 89.95032405790352
  - type: cos_sim_ap value: 88.03104739249996
  - type: cos_sim_f1 value: 80.34377190070451
  - type: cos_sim_precision value: 77.11534376548892
  - type: cos_sim_recall value: 83.85432707114259
  - type: dot_accuracy value: 89.95032405790352
  - type: dot_ap value: 88.03105328515932
  - type: dot_f1 value: 80.34377190070451
  - type: dot_precision value: 77.11534376548892
  - type: dot_recall value: 83.85432707114259
  - type: euclidean_accuracy value: 89.95032405790352
  - type: euclidean_ap value: 88.03105084564575
  - type: euclidean_f1 value: 80.34377190070451
  - type: euclidean_precision value: 77.11534376548892
  - type: euclidean_recall value: 83.85432707114259
  - type: manhattan_accuracy value: 89.88046726433035
  - type: manhattan_ap value: 88.01484191858279
  - type: manhattan_f1 value: 80.34005593993817
  - type: manhattan_precision value: 76.95290468133108
  - type: manhattan_recall value: 84.03911302740991
  - type: max_accuracy value: 89.95032405790352
  - type: max_ap value: 88.03105328515932
  - type: max_f1 value: 80.34377190070451 language:
en license: cc-by-nc-4.0

Salesforce/SFR-Embedding-2_R

SFR-Embedding by Salesforce Research.

The model is for research purposes only.

More technical details will be updated later. Meanwhile, please refer to our previous work SFR-Embedding for details.

Ethical Considerations

This release is for research purposes only in support of an academic paper. Our models, datasets, and code are not specifically designed or evaluated for all downstream purposes. We strongly recommend users evaluate and address potential concerns related to accuracy, safety, and fairness before deploying this model. We encourage users to consider the common limitations of AI, comply with applicable laws, and leverage best practices when selecting use cases, particularly for high-risk scenarios where errors or misuse could significantly impact people’s lives, rights, or safety. For further guidance on use cases, refer to our AUP and AI AUP.

SFR-Embedding Team (∗indicates equal contributors, † indicates co-leaders).

Rui Meng*
Ye Liu*
Tong Niu
Shafiq Rayhan Joty
Caiming Xiong †
Yingbo Zhou †
Semih Yavuz †

Citation

@misc{SFR-embedding-2,
  title={SFR-Embedding-2: Advanced Text Embedding with Multi-stage Training},
  author={Rui Meng*, Ye Liu*, Shafiq Rayhan Joty, Caiming Xiong, Yingbo Zhou, Semih Yavuz},
  year={2024},
  url={https://huggingface.co/Salesforce/SFR-Embedding-2_R}
}

How to run

Transformers

The models can be used as follows:

import torch
import torch.nn.functional as F
from torch import Tensor
from transformers import AutoTokenizer, AutoModel

def last_token_pool(last_hidden_states: Tensor,
                 attention_mask: Tensor) -> Tensor:
    left_padding = (attention_mask[:, -1].sum() == attention_mask.shape[0])
    if left_padding:
        return last_hidden_states[:, -1]
    else:
        sequence_lengths = attention_mask.sum(dim=1) - 1
        batch_size = last_hidden_states.shape[0]
        return last_hidden_states[torch.arange(batch_size, device=last_hidden_states.device), sequence_lengths]

def get_detailed_instruct(task_description: str, query: str) -> str:
    return f'Instruct: {task_description}\nQuery: {query}'

# Each query must come with a one-sentence instruction that describes the task
task = 'Given a web search query, retrieve relevant passages that answer the query'
queries = [
    get_detailed_instruct(task, 'How to bake a chocolate cake'),
    get_detailed_instruct(task, 'Symptoms of the flu')
]
# No need to add instruction for retrieval documents
passages = [
    "To bake a delicious chocolate cake, you'll need the following ingredients: all-purpose flour, sugar, cocoa powder, baking powder, baking soda, salt, eggs, milk, vegetable oil, and vanilla extract. Start by preheating your oven to 350°F (175°C). In a mixing bowl, combine the dry ingredients (flour, sugar, cocoa powder, baking powder, baking soda, and salt). In a separate bowl, whisk together the wet ingredients (eggs, milk, vegetable oil, and vanilla extract). Gradually add the wet mixture to the dry ingredients, stirring until well combined. Pour the batter into a greased cake pan and bake for 30-35 minutes. Let it cool before frosting with your favorite chocolate frosting. Enjoy your homemade chocolate cake!",
    "The flu, or influenza, is an illness caused by influenza viruses. Common symptoms of the flu include a high fever, chills, cough, sore throat, runny or stuffy nose, body aches, headache, fatigue, and sometimes nausea and vomiting. These symptoms can come on suddenly and are usually more severe than the common cold. It's important to get plenty of rest, stay hydrated, and consult a healthcare professional if you suspect you have the flu. In some cases, antiviral medications can help alleviate symptoms and reduce the duration of the illness."
]

# load model and tokenizer
tokenizer = AutoTokenizer.from_pretrained('Salesforce/SFR-Embedding-2_R')
model = AutoModel.from_pretrained('Salesforce/SFR-Embedding-2_R')

# get the embeddings
max_length = 4096
input_texts = queries + passages
batch_dict = tokenizer(input_texts, max_length=max_length, padding=True, truncation=True, return_tensors="pt")
outputs = model(**batch_dict)
embeddings = last_token_pool(outputs.last_hidden_state, batch_dict['attention_mask'])

# normalize embeddings
embeddings = F.normalize(embeddings, p=2, dim=1)
scores = (embeddings[:2] @ embeddings[2:].T) * 100
print(scores.tolist())
# [[40.132083892822266, 25.032529830932617], [15.006855010986328, 39.93733215332031]]

Sentence Transformers

from sentence_transformers import SentenceTransformer

model = SentenceTransformer("Salesforce/SFR-Embedding-2_R")

def get_detailed_instruct(task_description: str, query: str) -> str:
    return f'Instruct: {task_description}\nQuery: {query}'

# Each query must come with a one-sentence instruction that describes the task
task = 'Given a web search query, retrieve relevant passages that answer the query'
queries = [
    get_detailed_instruct(task, 'How to bake a chocolate cake'),
    get_detailed_instruct(task, 'Symptoms of the flu')
]
# No need to add instruction for retrieval documents
passages = [
    "To bake a delicious chocolate cake, you'll need the following ingredients: all-purpose flour, sugar, cocoa powder, baking powder, baking soda, salt, eggs, milk, vegetable oil, and vanilla extract. Start by preheating your oven to 350°F (175°C). In a mixing bowl, combine the dry ingredients (flour, sugar, cocoa powder, baking powder, baking soda, and salt). In a separate bowl, whisk together the wet ingredients (eggs, milk, vegetable oil, and vanilla extract). Gradually add the wet mixture to the dry ingredients, stirring until well combined. Pour the batter into a greased cake pan and bake for 30-35 minutes. Let it cool before frosting with your favorite chocolate frosting. Enjoy your homemade chocolate cake!",
    "The flu, or influenza, is an illness caused by influenza viruses. Common symptoms of the flu include a high fever, chills, cough, sore throat, runny or stuffy nose, body aches, headache, fatigue, and sometimes nausea and vomiting. These symptoms can come on suddenly and are usually more severe than the common cold. It's important to get plenty of rest, stay hydrated, and consult a healthcare professional if you suspect you have the flu. In some cases, antiviral medications can help alleviate symptoms and reduce the duration of the illness."
]

embeddings = model.encode(queries + passages)
scores = model.similarity(embeddings[:2], embeddings[2:]) * 100
print(scores.tolist())
# [[40.13203811645508, 25.032546997070312], [15.00684642791748, 39.937339782714844]]