checkpoint-4000/README.md

metadata

tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:124788
  - loss:GISTEmbedLoss
base_model: BAAI/bge-m3
widget:
  - source_sentence: 其他机械、设备和有形货物租赁服务代表
    sentences:
      - 其他机械和设备租赁服务工作人员
      - 电子和电信设备及零部件物流经理
      - 工业主厨
  - source_sentence: 公交车司机
    sentences:
      - 表演灯光设计师
      - 乙烯基地板安装工
      - 国际巴士司机
  - source_sentence: online communication manager
    sentences:
      - trades union official
      - social media manager
      - budget manager
  - source_sentence: Projektmanagerin
    sentences:
      - Projektmanager/Projektmanagerin
      - Category-Manager
      - Infanterist
  - source_sentence: Volksvertreter
    sentences:
      - Parlamentarier
      - Oberbürgermeister
      - Konsul
pipeline_tag: sentence-similarity
library_name: sentence-transformers
metrics:
  - cosine_accuracy@1
  - cosine_accuracy@20
  - cosine_accuracy@50
  - cosine_accuracy@100
  - cosine_accuracy@150
  - cosine_accuracy@200
  - cosine_precision@1
  - cosine_precision@20
  - cosine_precision@50
  - cosine_precision@100
  - cosine_precision@150
  - cosine_precision@200
  - cosine_recall@1
  - cosine_recall@20
  - cosine_recall@50
  - cosine_recall@100
  - cosine_recall@150
  - cosine_recall@200
  - cosine_ndcg@1
  - cosine_ndcg@20
  - cosine_ndcg@50
  - cosine_ndcg@100
  - cosine_ndcg@150
  - cosine_ndcg@200
  - cosine_mrr@1
  - cosine_mrr@20
  - cosine_mrr@50
  - cosine_mrr@100
  - cosine_mrr@150
  - cosine_mrr@200
  - cosine_map@1
  - cosine_map@20
  - cosine_map@50
  - cosine_map@100
  - cosine_map@150
  - cosine_map@200
  - cosine_map@500
model-index:
  - name: SentenceTransformer based on BAAI/bge-m3
    results:
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: full en
          type: full_en
        metrics:
          - type: cosine_accuracy@1
            value: 0.6476190476190476
            name: Cosine Accuracy@1
          - type: cosine_accuracy@20
            value: 0.9904761904761905
            name: Cosine Accuracy@20
          - type: cosine_accuracy@50
            value: 0.9904761904761905
            name: Cosine Accuracy@50
          - type: cosine_accuracy@100
            value: 0.9904761904761905
            name: Cosine Accuracy@100
          - type: cosine_accuracy@150
            value: 0.9904761904761905
            name: Cosine Accuracy@150
          - type: cosine_accuracy@200
            value: 0.9904761904761905
            name: Cosine Accuracy@200
          - type: cosine_precision@1
            value: 0.6476190476190476
            name: Cosine Precision@1
          - type: cosine_precision@20
            value: 0.499047619047619
            name: Cosine Precision@20
          - type: cosine_precision@50
            value: 0.30266666666666664
            name: Cosine Precision@50
          - type: cosine_precision@100
            value: 0.18447619047619046
            name: Cosine Precision@100
          - type: cosine_precision@150
            value: 0.13155555555555554
            name: Cosine Precision@150
          - type: cosine_precision@200
            value: 0.10171428571428573
            name: Cosine Precision@200
          - type: cosine_recall@1
            value: 0.06690172806447445
            name: Cosine Recall@1
          - type: cosine_recall@20
            value: 0.5288155255988508
            name: Cosine Recall@20
          - type: cosine_recall@50
            value: 0.7128731386766649
            name: Cosine Recall@50
          - type: cosine_recall@100
            value: 0.821589853989195
            name: Cosine Recall@100
          - type: cosine_recall@150
            value: 0.8669290529739844
            name: Cosine Recall@150
          - type: cosine_recall@200
            value: 0.8881772271562451
            name: Cosine Recall@200
          - type: cosine_ndcg@1
            value: 0.6476190476190476
            name: Cosine Ndcg@1
          - type: cosine_ndcg@20
            value: 0.6737021289484512
            name: Cosine Ndcg@20
          - type: cosine_ndcg@50
            value: 0.6897381539459008
            name: Cosine Ndcg@50
          - type: cosine_ndcg@100
            value: 0.7455379155828873
            name: Cosine Ndcg@100
          - type: cosine_ndcg@150
            value: 0.7657730626526685
            name: Cosine Ndcg@150
          - type: cosine_ndcg@200
            value: 0.7746920852324353
            name: Cosine Ndcg@200
          - type: cosine_mrr@1
            value: 0.6476190476190476
            name: Cosine Mrr@1
          - type: cosine_mrr@20
            value: 0.7969444444444443
            name: Cosine Mrr@20
          - type: cosine_mrr@50
            value: 0.7969444444444443
            name: Cosine Mrr@50
          - type: cosine_mrr@100
            value: 0.7969444444444443
            name: Cosine Mrr@100
          - type: cosine_mrr@150
            value: 0.7969444444444443
            name: Cosine Mrr@150
          - type: cosine_mrr@200
            value: 0.7969444444444443
            name: Cosine Mrr@200
          - type: cosine_map@1
            value: 0.6476190476190476
            name: Cosine Map@1
          - type: cosine_map@20
            value: 0.5299368408688423
            name: Cosine Map@20
          - type: cosine_map@50
            value: 0.5170402457535271
            name: Cosine Map@50
          - type: cosine_map@100
            value: 0.549577105065989
            name: Cosine Map@100
          - type: cosine_map@150
            value: 0.5580348324082148
            name: Cosine Map@150
          - type: cosine_map@200
            value: 0.5609705433942662
            name: Cosine Map@200
          - type: cosine_map@500
            value: 0.5664835460503455
            name: Cosine Map@500
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: full es
          type: full_es
        metrics:
          - type: cosine_accuracy@1
            value: 0.12432432432432433
            name: Cosine Accuracy@1
          - type: cosine_accuracy@20
            value: 1
            name: Cosine Accuracy@20
          - type: cosine_accuracy@50
            value: 1
            name: Cosine Accuracy@50
          - type: cosine_accuracy@100
            value: 1
            name: Cosine Accuracy@100
          - type: cosine_accuracy@150
            value: 1
            name: Cosine Accuracy@150
          - type: cosine_accuracy@200
            value: 1
            name: Cosine Accuracy@200
          - type: cosine_precision@1
            value: 0.12432432432432433
            name: Cosine Precision@1
          - type: cosine_precision@20
            value: 0.5718918918918918
            name: Cosine Precision@20
          - type: cosine_precision@50
            value: 0.38832432432432434
            name: Cosine Precision@50
          - type: cosine_precision@100
            value: 0.25135135135135134
            name: Cosine Precision@100
          - type: cosine_precision@150
            value: 0.1886486486486487
            name: Cosine Precision@150
          - type: cosine_precision@200
            value: 0.15083783783783786
            name: Cosine Precision@200
          - type: cosine_recall@1
            value: 0.0036542148230633313
            name: Cosine Recall@1
          - type: cosine_recall@20
            value: 0.3813088657975513
            name: Cosine Recall@20
          - type: cosine_recall@50
            value: 0.5589819018381946
            name: Cosine Recall@50
          - type: cosine_recall@100
            value: 0.6712879484837694
            name: Cosine Recall@100
          - type: cosine_recall@150
            value: 0.7296378671854172
            name: Cosine Recall@150
          - type: cosine_recall@200
            value: 0.7646529145750729
            name: Cosine Recall@200
          - type: cosine_ndcg@1
            value: 0.12432432432432433
            name: Cosine Ndcg@1
          - type: cosine_ndcg@20
            value: 0.6162786673767947
            name: Cosine Ndcg@20
          - type: cosine_ndcg@50
            value: 0.5875500387824142
            name: Cosine Ndcg@50
          - type: cosine_ndcg@100
            value: 0.6146487956773306
            name: Cosine Ndcg@100
          - type: cosine_ndcg@150
            value: 0.6449661586574366
            name: Cosine Ndcg@150
          - type: cosine_ndcg@200
            value: 0.6628313427507618
            name: Cosine Ndcg@200
          - type: cosine_mrr@1
            value: 0.12432432432432433
            name: Cosine Mrr@1
          - type: cosine_mrr@20
            value: 0.5585585585585586
            name: Cosine Mrr@20
          - type: cosine_mrr@50
            value: 0.5585585585585586
            name: Cosine Mrr@50
          - type: cosine_mrr@100
            value: 0.5585585585585586
            name: Cosine Mrr@100
          - type: cosine_mrr@150
            value: 0.5585585585585586
            name: Cosine Mrr@150
          - type: cosine_mrr@200
            value: 0.5585585585585586
            name: Cosine Mrr@200
          - type: cosine_map@1
            value: 0.12432432432432433
            name: Cosine Map@1
          - type: cosine_map@20
            value: 0.4830935685993706
            name: Cosine Map@20
          - type: cosine_map@50
            value: 0.4268637780839156
            name: Cosine Map@50
          - type: cosine_map@100
            value: 0.43032040469750343
            name: Cosine Map@100
          - type: cosine_map@150
            value: 0.4449589410699155
            name: Cosine Map@150
          - type: cosine_map@200
            value: 0.4523102942291434
            name: Cosine Map@200
          - type: cosine_map@500
            value: 0.4643631946508736
            name: Cosine Map@500
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: full de
          type: full_de
        metrics:
          - type: cosine_accuracy@1
            value: 0.2955665024630542
            name: Cosine Accuracy@1
          - type: cosine_accuracy@20
            value: 0.9753694581280788
            name: Cosine Accuracy@20
          - type: cosine_accuracy@50
            value: 0.9852216748768473
            name: Cosine Accuracy@50
          - type: cosine_accuracy@100
            value: 0.9852216748768473
            name: Cosine Accuracy@100
          - type: cosine_accuracy@150
            value: 0.9901477832512315
            name: Cosine Accuracy@150
          - type: cosine_accuracy@200
            value: 0.9901477832512315
            name: Cosine Accuracy@200
          - type: cosine_precision@1
            value: 0.2955665024630542
            name: Cosine Precision@1
          - type: cosine_precision@20
            value: 0.5399014778325123
            name: Cosine Precision@20
          - type: cosine_precision@50
            value: 0.3829556650246305
            name: Cosine Precision@50
          - type: cosine_precision@100
            value: 0.25098522167487686
            name: Cosine Precision@100
          - type: cosine_precision@150
            value: 0.18742200328407224
            name: Cosine Precision@150
          - type: cosine_precision@200
            value: 0.14911330049261085
            name: Cosine Precision@200
          - type: cosine_recall@1
            value: 0.01108543831680986
            name: Cosine Recall@1
          - type: cosine_recall@20
            value: 0.33926725064737134
            name: Cosine Recall@20
          - type: cosine_recall@50
            value: 0.5319613376214742
            name: Cosine Recall@50
          - type: cosine_recall@100
            value: 0.6497082600959269
            name: Cosine Recall@100
          - type: cosine_recall@150
            value: 0.7094703332321319
            name: Cosine Recall@150
          - type: cosine_recall@200
            value: 0.7445597670438818
            name: Cosine Recall@200
          - type: cosine_ndcg@1
            value: 0.2955665024630542
            name: Cosine Ndcg@1
          - type: cosine_ndcg@20
            value: 0.5621043185251402
            name: Cosine Ndcg@20
          - type: cosine_ndcg@50
            value: 0.5505636839954736
            name: Cosine Ndcg@50
          - type: cosine_ndcg@100
            value: 0.5784375922614946
            name: Cosine Ndcg@100
          - type: cosine_ndcg@150
            value: 0.6091764880384499
            name: Cosine Ndcg@150
          - type: cosine_ndcg@200
            value: 0.6263384735475871
            name: Cosine Ndcg@200
          - type: cosine_mrr@1
            value: 0.2955665024630542
            name: Cosine Mrr@1
          - type: cosine_mrr@20
            value: 0.5127296895769795
            name: Cosine Mrr@20
          - type: cosine_mrr@50
            value: 0.5130763416477695
            name: Cosine Mrr@50
          - type: cosine_mrr@100
            value: 0.5130763416477695
            name: Cosine Mrr@100
          - type: cosine_mrr@150
            value: 0.5131188080992728
            name: Cosine Mrr@150
          - type: cosine_mrr@200
            value: 0.5131188080992728
            name: Cosine Mrr@200
          - type: cosine_map@1
            value: 0.2955665024630542
            name: Cosine Map@1
          - type: cosine_map@20
            value: 0.42085554479107096
            name: Cosine Map@20
          - type: cosine_map@50
            value: 0.3779379416896035
            name: Cosine Map@50
          - type: cosine_map@100
            value: 0.38163165810143573
            name: Cosine Map@100
          - type: cosine_map@150
            value: 0.3961646378244818
            name: Cosine Map@150
          - type: cosine_map@200
            value: 0.40295816570523324
            name: Cosine Map@200
          - type: cosine_map@500
            value: 0.4167002568710484
            name: Cosine Map@500
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: full zh
          type: full_zh
        metrics:
          - type: cosine_accuracy@1
            value: 0.6407766990291263
            name: Cosine Accuracy@1
          - type: cosine_accuracy@20
            value: 0.9902912621359223
            name: Cosine Accuracy@20
          - type: cosine_accuracy@50
            value: 0.9902912621359223
            name: Cosine Accuracy@50
          - type: cosine_accuracy@100
            value: 0.9902912621359223
            name: Cosine Accuracy@100
          - type: cosine_accuracy@150
            value: 0.9902912621359223
            name: Cosine Accuracy@150
          - type: cosine_accuracy@200
            value: 0.9902912621359223
            name: Cosine Accuracy@200
          - type: cosine_precision@1
            value: 0.6407766990291263
            name: Cosine Precision@1
          - type: cosine_precision@20
            value: 0.46504854368932047
            name: Cosine Precision@20
          - type: cosine_precision@50
            value: 0.27611650485436895
            name: Cosine Precision@50
          - type: cosine_precision@100
            value: 0.17097087378640777
            name: Cosine Precision@100
          - type: cosine_precision@150
            value: 0.12291262135922332
            name: Cosine Precision@150
          - type: cosine_precision@200
            value: 0.0969417475728155
            name: Cosine Precision@200
          - type: cosine_recall@1
            value: 0.05744396078263393
            name: Cosine Recall@1
          - type: cosine_recall@20
            value: 0.4978573021507442
            name: Cosine Recall@20
          - type: cosine_recall@50
            value: 0.6611813069264482
            name: Cosine Recall@50
          - type: cosine_recall@100
            value: 0.7796553453979224
            name: Cosine Recall@100
          - type: cosine_recall@150
            value: 0.8271677009796732
            name: Cosine Recall@150
          - type: cosine_recall@200
            value: 0.8637730394316714
            name: Cosine Recall@200
          - type: cosine_ndcg@1
            value: 0.6407766990291263
            name: Cosine Ndcg@1
          - type: cosine_ndcg@20
            value: 0.6374339653798218
            name: Cosine Ndcg@20
          - type: cosine_ndcg@50
            value: 0.6458466090741598
            name: Cosine Ndcg@50
          - type: cosine_ndcg@100
            value: 0.7026844413104963
            name: Cosine Ndcg@100
          - type: cosine_ndcg@150
            value: 0.7238302410564206
            name: Cosine Ndcg@150
          - type: cosine_ndcg@200
            value: 0.7383757321568225
            name: Cosine Ndcg@200
          - type: cosine_mrr@1
            value: 0.6407766990291263
            name: Cosine Mrr@1
          - type: cosine_mrr@20
            value: 0.7983818770226538
            name: Cosine Mrr@20
          - type: cosine_mrr@50
            value: 0.7983818770226538
            name: Cosine Mrr@50
          - type: cosine_mrr@100
            value: 0.7983818770226538
            name: Cosine Mrr@100
          - type: cosine_mrr@150
            value: 0.7983818770226538
            name: Cosine Mrr@150
          - type: cosine_mrr@200
            value: 0.7983818770226538
            name: Cosine Mrr@200
          - type: cosine_map@1
            value: 0.6407766990291263
            name: Cosine Map@1
          - type: cosine_map@20
            value: 0.4902515378001179
            name: Cosine Map@20
          - type: cosine_map@50
            value: 0.46828607843970593
            name: Cosine Map@50
          - type: cosine_map@100
            value: 0.49742002930709256
            name: Cosine Map@100
          - type: cosine_map@150
            value: 0.5055517135202557
            name: Cosine Map@150
          - type: cosine_map@200
            value: 0.5100267276205871
            name: Cosine Map@200
          - type: cosine_map@500
            value: 0.5152273086702759
            name: Cosine Map@500
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: mix es
          type: mix_es
        metrics:
          - type: cosine_accuracy@1
            value: 0.7358294331773271
            name: Cosine Accuracy@1
          - type: cosine_accuracy@20
            value: 0.9625585023400937
            name: Cosine Accuracy@20
          - type: cosine_accuracy@50
            value: 0.9802392095683827
            name: Cosine Accuracy@50
          - type: cosine_accuracy@100
            value: 0.9927197087883516
            name: Cosine Accuracy@100
          - type: cosine_accuracy@150
            value: 0.9947997919916797
            name: Cosine Accuracy@150
          - type: cosine_accuracy@200
            value: 0.9958398335933437
            name: Cosine Accuracy@200
          - type: cosine_precision@1
            value: 0.7358294331773271
            name: Cosine Precision@1
          - type: cosine_precision@20
            value: 0.12438897555902236
            name: Cosine Precision@20
          - type: cosine_precision@50
            value: 0.05158606344253771
            name: Cosine Precision@50
          - type: cosine_precision@100
            value: 0.026224648985959446
            name: Cosine Precision@100
          - type: cosine_precision@150
            value: 0.017628705148205928
            name: Cosine Precision@150
          - type: cosine_precision@200
            value: 0.013268330733229333
            name: Cosine Precision@200
          - type: cosine_recall@1
            value: 0.28403164698016486
            name: Cosine Recall@1
          - type: cosine_recall@20
            value: 0.9190414283237995
            name: Cosine Recall@20
          - type: cosine_recall@50
            value: 0.952244756456925
            name: Cosine Recall@50
          - type: cosine_recall@100
            value: 0.9685820766163981
            name: Cosine Recall@100
          - type: cosine_recall@150
            value: 0.9762870514820593
            name: Cosine Recall@150
          - type: cosine_recall@200
            value: 0.9801872074882996
            name: Cosine Recall@200
          - type: cosine_ndcg@1
            value: 0.7358294331773271
            name: Cosine Ndcg@1
          - type: cosine_ndcg@20
            value: 0.8089516774866639
            name: Cosine Ndcg@20
          - type: cosine_ndcg@50
            value: 0.8181299102768375
            name: Cosine Ndcg@50
          - type: cosine_ndcg@100
            value: 0.8217009899252086
            name: Cosine Ndcg@100
          - type: cosine_ndcg@150
            value: 0.8232345422421572
            name: Cosine Ndcg@150
          - type: cosine_ndcg@200
            value: 0.8239096085290897
            name: Cosine Ndcg@200
          - type: cosine_mrr@1
            value: 0.7358294331773271
            name: Cosine Mrr@1
          - type: cosine_mrr@20
            value: 0.8035232306901704
            name: Cosine Mrr@20
          - type: cosine_mrr@50
            value: 0.8041564269676074
            name: Cosine Mrr@50
          - type: cosine_mrr@100
            value: 0.8043491602665708
            name: Cosine Mrr@100
          - type: cosine_mrr@150
            value: 0.8043649132860833
            name: Cosine Mrr@150
          - type: cosine_mrr@200
            value: 0.8043707455995762
            name: Cosine Mrr@200
          - type: cosine_map@1
            value: 0.7358294331773271
            name: Cosine Map@1
          - type: cosine_map@20
            value: 0.7407296211762635
            name: Cosine Map@20
          - type: cosine_map@50
            value: 0.7433011890905112
            name: Cosine Map@50
          - type: cosine_map@100
            value: 0.7437599072934008
            name: Cosine Map@100
          - type: cosine_map@150
            value: 0.7439220951644092
            name: Cosine Map@150
          - type: cosine_map@200
            value: 0.7439677461223776
            name: Cosine Map@200
          - type: cosine_map@500
            value: 0.7440630263326289
            name: Cosine Map@500
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: mix de
          type: mix_de
        metrics:
          - type: cosine_accuracy@1
            value: 0.6947477899115965
            name: Cosine Accuracy@1
          - type: cosine_accuracy@20
            value: 0.967758710348414
            name: Cosine Accuracy@20
          - type: cosine_accuracy@50
            value: 0.984399375975039
            name: Cosine Accuracy@50
          - type: cosine_accuracy@100
            value: 0.9901196047841914
            name: Cosine Accuracy@100
          - type: cosine_accuracy@150
            value: 0.9932397295891836
            name: Cosine Accuracy@150
          - type: cosine_accuracy@200
            value: 0.9932397295891836
            name: Cosine Accuracy@200
          - type: cosine_precision@1
            value: 0.6947477899115965
            name: Cosine Precision@1
          - type: cosine_precision@20
            value: 0.12769110764430577
            name: Cosine Precision@20
          - type: cosine_precision@50
            value: 0.05316692667706709
            name: Cosine Precision@50
          - type: cosine_precision@100
            value: 0.026978679147165893
            name: Cosine Precision@100
          - type: cosine_precision@150
            value: 0.018082856647599233
            name: Cosine Precision@150
          - type: cosine_precision@200
            value: 0.013595943837753513
            name: Cosine Precision@200
          - type: cosine_recall@1
            value: 0.26064309239036226
            name: Cosine Recall@1
          - type: cosine_recall@20
            value: 0.9266163979892529
            name: Cosine Recall@20
          - type: cosine_recall@50
            value: 0.9632518634078697
            name: Cosine Recall@50
          - type: cosine_recall@100
            value: 0.9771190847633905
            name: Cosine Recall@100
          - type: cosine_recall@150
            value: 0.982232622638239
            name: Cosine Recall@150
          - type: cosine_recall@200
            value: 0.984659386375455
            name: Cosine Recall@200
          - type: cosine_ndcg@1
            value: 0.6947477899115965
            name: Cosine Ndcg@1
          - type: cosine_ndcg@20
            value: 0.7916550876560119
            name: Cosine Ndcg@20
          - type: cosine_ndcg@50
            value: 0.8018356667177752
            name: Cosine Ndcg@50
          - type: cosine_ndcg@100
            value: 0.8049830038156018
            name: Cosine Ndcg@100
          - type: cosine_ndcg@150
            value: 0.8060041518104935
            name: Cosine Ndcg@150
          - type: cosine_ndcg@200
            value: 0.8064526867706615
            name: Cosine Ndcg@200
          - type: cosine_mrr@1
            value: 0.6947477899115965
            name: Cosine Mrr@1
          - type: cosine_mrr@20
            value: 0.775106319970792
            name: Cosine Mrr@20
          - type: cosine_mrr@50
            value: 0.7756762344136855
            name: Cosine Mrr@50
          - type: cosine_mrr@100
            value: 0.7757636235577245
            name: Cosine Mrr@100
          - type: cosine_mrr@150
            value: 0.7757917238264626
            name: Cosine Mrr@150
          - type: cosine_mrr@200
            value: 0.7757917238264626
            name: Cosine Mrr@200
          - type: cosine_map@1
            value: 0.6947477899115965
            name: Cosine Map@1
          - type: cosine_map@20
            value: 0.7123386461179687
            name: Cosine Map@20
          - type: cosine_map@50
            value: 0.7151736057555711
            name: Cosine Map@50
          - type: cosine_map@100
            value: 0.7156740227134941
            name: Cosine Map@100
          - type: cosine_map@150
            value: 0.7157705885677804
            name: Cosine Map@150
          - type: cosine_map@200
            value: 0.7158097678043102
            name: Cosine Map@200
          - type: cosine_map@500
            value: 0.7158747359338941
            name: Cosine Map@500
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: mix zh
          type: mix_zh
        metrics:
          - type: cosine_accuracy@1
            value: 0.1814872594903796
            name: Cosine Accuracy@1
          - type: cosine_accuracy@20
            value: 1
            name: Cosine Accuracy@20
          - type: cosine_accuracy@50
            value: 1
            name: Cosine Accuracy@50
          - type: cosine_accuracy@100
            value: 1
            name: Cosine Accuracy@100
          - type: cosine_accuracy@150
            value: 1
            name: Cosine Accuracy@150
          - type: cosine_accuracy@200
            value: 1
            name: Cosine Accuracy@200
          - type: cosine_precision@1
            value: 0.1814872594903796
            name: Cosine Precision@1
          - type: cosine_precision@20
            value: 0.15439417576703063
            name: Cosine Precision@20
          - type: cosine_precision@50
            value: 0.0617576703068123
            name: Cosine Precision@50
          - type: cosine_precision@100
            value: 0.03087883515340615
            name: Cosine Precision@100
          - type: cosine_precision@150
            value: 0.020585890102270757
            name: Cosine Precision@150
          - type: cosine_precision@200
            value: 0.015439417576703075
            name: Cosine Precision@200
          - type: cosine_recall@1
            value: 0.058722729861575416
            name: Cosine Recall@1
          - type: cosine_recall@20
            value: 1
            name: Cosine Recall@20
          - type: cosine_recall@50
            value: 1
            name: Cosine Recall@50
          - type: cosine_recall@100
            value: 1
            name: Cosine Recall@100
          - type: cosine_recall@150
            value: 1
            name: Cosine Recall@150
          - type: cosine_recall@200
            value: 1
            name: Cosine Recall@200
          - type: cosine_ndcg@1
            value: 0.1814872594903796
            name: Cosine Ndcg@1
          - type: cosine_ndcg@20
            value: 0.5447038314336347
            name: Cosine Ndcg@20
          - type: cosine_ndcg@50
            value: 0.5447038314336347
            name: Cosine Ndcg@50
          - type: cosine_ndcg@100
            value: 0.5447038314336347
            name: Cosine Ndcg@100
          - type: cosine_ndcg@150
            value: 0.5447038314336347
            name: Cosine Ndcg@150
          - type: cosine_ndcg@200
            value: 0.5447038314336347
            name: Cosine Ndcg@200
          - type: cosine_mrr@1
            value: 0.1814872594903796
            name: Cosine Mrr@1
          - type: cosine_mrr@20
            value: 0.40366659543726713
            name: Cosine Mrr@20
          - type: cosine_mrr@50
            value: 0.40366659543726713
            name: Cosine Mrr@50
          - type: cosine_mrr@100
            value: 0.40366659543726713
            name: Cosine Mrr@100
          - type: cosine_mrr@150
            value: 0.40366659543726713
            name: Cosine Mrr@150
          - type: cosine_mrr@200
            value: 0.40366659543726713
            name: Cosine Mrr@200
          - type: cosine_map@1
            value: 0.1814872594903796
            name: Cosine Map@1
          - type: cosine_map@20
            value: 0.32665499722442
            name: Cosine Map@20
          - type: cosine_map@50
            value: 0.32665499722442
            name: Cosine Map@50
          - type: cosine_map@100
            value: 0.32665499722442
            name: Cosine Map@100
          - type: cosine_map@150
            value: 0.32665499722442
            name: Cosine Map@150
          - type: cosine_map@200
            value: 0.32665499722442
            name: Cosine Map@200
          - type: cosine_map@500
            value: 0.32665499722442
            name: Cosine Map@500

SentenceTransformer based on BAAI/bge-m3

This is a sentence-transformers model finetuned from BAAI/bge-m3 on the full_en, full_de, full_es, full_zh and mix datasets. It maps sentences & paragraphs to a 1024-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.

Model Details

Model Description

• Model Type: Sentence Transformer
• Base model: BAAI/bge-m3
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 1024 dimensions
• Similarity Function: Cosine Similarity
• Training Datasets:
  • full_en
  • full_de
  • full_es
  • full_zh
  • mix

Model Sources

• Documentation: Sentence Transformers Documentation
• Repository: Sentence Transformers on GitHub
• Hugging Face: Sentence Transformers on Hugging Face

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: XLMRobertaModel 
  (1): Pooling({'word_embedding_dimension': 1024, 'pooling_mode_cls_token': True, 'pooling_mode_mean_tokens': False, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
  (2): Normalize()
)

Usage

Direct Usage (Sentence Transformers)

First install the Sentence Transformers library:

pip install -U sentence-transformers

Then you can load this model and run inference.

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("sentence_transformers_model_id")
# Run inference
sentences = [
    'Volksvertreter',
    'Parlamentarier',
    'Oberbürgermeister',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 1024]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]

Evaluation

Metrics

Information Retrieval

• Datasets: full_en, full_es, full_de, full_zh, mix_es, mix_de and mix_zh
• Evaluated with InformationRetrievalEvaluator

Metric	full_en	full_es	full_de	full_zh	mix_es	mix_de	mix_zh
cosine_accuracy@1	0.6476	0.1243	0.2956	0.6408	0.7358	0.6947	0.1815
cosine_accuracy@20	0.9905	1.0	0.9754	0.9903	0.9626	0.9678	1.0
cosine_accuracy@50	0.9905	1.0	0.9852	0.9903	0.9802	0.9844	1.0
cosine_accuracy@100	0.9905	1.0	0.9852	0.9903	0.9927	0.9901	1.0
cosine_accuracy@150	0.9905	1.0	0.9901	0.9903	0.9948	0.9932	1.0
cosine_accuracy@200	0.9905	1.0	0.9901	0.9903	0.9958	0.9932	1.0
cosine_precision@1	0.6476	0.1243	0.2956	0.6408	0.7358	0.6947	0.1815
cosine_precision@20	0.499	0.5719	0.5399	0.465	0.1244	0.1277	0.1544
cosine_precision@50	0.3027	0.3883	0.383	0.2761	0.0516	0.0532	0.0618
cosine_precision@100	0.1845	0.2514	0.251	0.171	0.0262	0.027	0.0309
cosine_precision@150	0.1316	0.1886	0.1874	0.1229	0.0176	0.0181	0.0206
cosine_precision@200	0.1017	0.1508	0.1491	0.0969	0.0133	0.0136	0.0154
cosine_recall@1	0.0669	0.0037	0.0111	0.0574	0.284	0.2606	0.0587
cosine_recall@20	0.5288	0.3813	0.3393	0.4979	0.919	0.9266	1.0
cosine_recall@50	0.7129	0.559	0.532	0.6612	0.9522	0.9633	1.0
cosine_recall@100	0.8216	0.6713	0.6497	0.7797	0.9686	0.9771	1.0
cosine_recall@150	0.8669	0.7296	0.7095	0.8272	0.9763	0.9822	1.0
cosine_recall@200	0.8882	0.7647	0.7446	0.8638	0.9802	0.9847	1.0
cosine_ndcg@1	0.6476	0.1243	0.2956	0.6408	0.7358	0.6947	0.1815
cosine_ndcg@20	0.6737	0.6163	0.5621	0.6374	0.809	0.7917	0.5447
cosine_ndcg@50	0.6897	0.5876	0.5506	0.6458	0.8181	0.8018	0.5447
cosine_ndcg@100	0.7455	0.6146	0.5784	0.7027	0.8217	0.805	0.5447
cosine_ndcg@150	0.7658	0.645	0.6092	0.7238	0.8232	0.806	0.5447
cosine_ndcg@200	0.7747	0.6628	0.6263	0.7384	0.8239	0.8065	0.5447
cosine_mrr@1	0.6476	0.1243	0.2956	0.6408	0.7358	0.6947	0.1815
cosine_mrr@20	0.7969	0.5586	0.5127	0.7984	0.8035	0.7751	0.4037
cosine_mrr@50	0.7969	0.5586	0.5131	0.7984	0.8042	0.7757	0.4037
cosine_mrr@100	0.7969	0.5586	0.5131	0.7984	0.8043	0.7758	0.4037
cosine_mrr@150	0.7969	0.5586	0.5131	0.7984	0.8044	0.7758	0.4037
cosine_mrr@200	0.7969	0.5586	0.5131	0.7984	0.8044	0.7758	0.4037
cosine_map@1	0.6476	0.1243	0.2956	0.6408	0.7358	0.6947	0.1815
cosine_map@20	0.5299	0.4831	0.4209	0.4903	0.7407	0.7123	0.3267
cosine_map@50	0.517	0.4269	0.3779	0.4683	0.7433	0.7152	0.3267
cosine_map@100	0.5496	0.4303	0.3816	0.4974	0.7438	0.7157	0.3267
cosine_map@150	0.558	0.445	0.3962	0.5056	0.7439	0.7158	0.3267
cosine_map@200	0.561	0.4523	0.403	0.51	0.744	0.7158	0.3267
cosine_map@500	0.5665	0.4644	0.4167	0.5152	0.7441	0.7159	0.3267

Training Details

Training Datasets

full_en

full_en

• Dataset: full_en
• Size: 28,880 training samples
• Columns: anchor and positive
• Approximate statistics based on the first 1000 samples:

  	anchor	positive
  type	string	string
  details	min: 3 tokens mean: 5.68 tokens max: 11 tokens	min: 3 tokens mean: 5.76 tokens max: 12 tokens

• Samples:

  anchor	positive
  air commodore	flight lieutenant
  command and control officer	flight officer
  air commodore	command and control officer

• Loss: GISTEmbedLoss with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 128, 'do_lower_case': False}) with Transformer model: BertModel 
    (1): Pooling({'word_embedding_dimension': 384, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
    (2): Normalize()
  ), 'temperature': 0.01, 'margin_strategy': 'absolute', 'margin': 0.0}
  

full_de

full_de

• Dataset: full_de
• Size: 23,023 training samples
• Columns: anchor and positive
• Approximate statistics based on the first 1000 samples:

  	anchor	positive
  type	string	string
  details	min: 3 tokens mean: 7.99 tokens max: 30 tokens	min: 3 tokens mean: 8.19 tokens max: 30 tokens

• Samples:

  anchor	positive
  Staffelkommandantin	Kommodore
  Luftwaffenoffizierin	Luftwaffenoffizier/Luftwaffenoffizierin
  Staffelkommandantin	Luftwaffenoffizierin

• Loss: GISTEmbedLoss with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 128, 'do_lower_case': False}) with Transformer model: BertModel 
    (1): Pooling({'word_embedding_dimension': 384, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
    (2): Normalize()
  ), 'temperature': 0.01, 'margin_strategy': 'absolute', 'margin': 0.0}
  

full_es

full_es

• Dataset: full_es
• Size: 20,724 training samples
• Columns: anchor and positive
• Approximate statistics based on the first 1000 samples:

  	anchor	positive
  type	string	string
  details	min: 3 tokens mean: 9.13 tokens max: 32 tokens	min: 3 tokens mean: 8.84 tokens max: 32 tokens

• Samples:

  anchor	positive
  jefe de escuadrón	instructor
  comandante de aeronave	instructor de simulador
  instructor	oficial del Ejército del Aire

• Loss: GISTEmbedLoss with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 128, 'do_lower_case': False}) with Transformer model: BertModel 
    (1): Pooling({'word_embedding_dimension': 384, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
    (2): Normalize()
  ), 'temperature': 0.01, 'margin_strategy': 'absolute', 'margin': 0.0}
  

full_zh

full_zh

• Dataset: full_zh
• Size: 30,401 training samples
• Columns: anchor and positive
• Approximate statistics based on the first 1000 samples:

  	anchor	positive
  type	string	string
  details	min: 5 tokens mean: 7.15 tokens max: 14 tokens	min: 5 tokens mean: 7.46 tokens max: 21 tokens

• Samples:

  anchor	positive
  技术总监	技术和运营总监
  技术总监	技术主管
  技术总监	技术艺术总监

• Loss: GISTEmbedLoss with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 128, 'do_lower_case': False}) with Transformer model: BertModel 
    (1): Pooling({'word_embedding_dimension': 384, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
    (2): Normalize()
  ), 'temperature': 0.01, 'margin_strategy': 'absolute', 'margin': 0.0}
  

mix

mix

• Dataset: mix
• Size: 21,760 training samples
• Columns: anchor and positive
• Approximate statistics based on the first 1000 samples:

  	anchor	positive
  type	string	string
  details	min: 2 tokens mean: 6.71 tokens max: 19 tokens	min: 2 tokens mean: 7.69 tokens max: 19 tokens

• Samples:

  anchor	positive
  technical manager	Technischer Direktor für Bühne, Film und Fernsehen
  head of technical	directora técnica
  head of technical department	技术艺术总监

• Loss: GISTEmbedLoss with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 128, 'do_lower_case': False}) with Transformer model: BertModel 
    (1): Pooling({'word_embedding_dimension': 384, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
    (2): Normalize()
  ), 'temperature': 0.01, 'margin_strategy': 'absolute', 'margin': 0.0}
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 64
• per_device_eval_batch_size: 128
• gradient_accumulation_steps: 2
• num_train_epochs: 5
• warmup_ratio: 0.05
• log_on_each_node: False
• fp16: True
• dataloader_num_workers: 4
• ddp_find_unused_parameters: True
• batch_sampler: no_duplicates

All Hyperparameters

Click to expand

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: steps
• prediction_loss_only: True
• per_device_train_batch_size: 64
• per_device_eval_batch_size: 128
• per_gpu_train_batch_size: None
• per_gpu_eval_batch_size: None
• gradient_accumulation_steps: 2
• eval_accumulation_steps: None
• torch_empty_cache_steps: None
• learning_rate: 5e-05
• weight_decay: 0.0
• adam_beta1: 0.9
• adam_beta2: 0.999
• adam_epsilon: 1e-08
• max_grad_norm: 1.0
• num_train_epochs: 5
• max_steps: -1
• lr_scheduler_type: linear
• lr_scheduler_kwargs: {}
• warmup_ratio: 0.05
• warmup_steps: 0
• log_level: passive
• log_level_replica: warning
• log_on_each_node: False
• logging_nan_inf_filter: True
• save_safetensors: True
• save_on_each_node: False
• save_only_model: False
• restore_callback_states_from_checkpoint: False
• no_cuda: False
• use_cpu: False
• use_mps_device: False
• seed: 42
• data_seed: None
• jit_mode_eval: False
• use_ipex: False
• bf16: False
• fp16: True
• fp16_opt_level: O1
• half_precision_backend: auto
• bf16_full_eval: False
• fp16_full_eval: False
• tf32: None
• local_rank: 0
• ddp_backend: None
• tpu_num_cores: None
• tpu_metrics_debug: False
• debug: []
• dataloader_drop_last: True
• dataloader_num_workers: 4
• dataloader_prefetch_factor: None
• past_index: -1
• disable_tqdm: False
• remove_unused_columns: True
• label_names: None
• load_best_model_at_end: False
• ignore_data_skip: False
• fsdp: []
• fsdp_min_num_params: 0
• fsdp_config: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
• tp_size: 0
• fsdp_transformer_layer_cls_to_wrap: None
• accelerator_config: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
• deepspeed: None
• label_smoothing_factor: 0.0
• optim: adamw_torch
• optim_args: None
• adafactor: False
• group_by_length: False
• length_column_name: length
• ddp_find_unused_parameters: True
• ddp_bucket_cap_mb: None
• ddp_broadcast_buffers: False
• dataloader_pin_memory: True
• dataloader_persistent_workers: False
• skip_memory_metrics: True
• use_legacy_prediction_loop: False
• push_to_hub: False
• resume_from_checkpoint: None
• hub_model_id: None
• hub_strategy: every_save
• hub_private_repo: None
• hub_always_push: False
• gradient_checkpointing: False
• gradient_checkpointing_kwargs: None
• include_inputs_for_metrics: False
• include_for_metrics: []
• eval_do_concat_batches: True
• fp16_backend: auto
• push_to_hub_model_id: None
• push_to_hub_organization: None
• mp_parameters:
• auto_find_batch_size: False
• full_determinism: False
• torchdynamo: None
• ray_scope: last
• ddp_timeout: 1800
• torch_compile: False
• torch_compile_backend: None
• torch_compile_mode: None
• include_tokens_per_second: False
• include_num_input_tokens_seen: False
• neftune_noise_alpha: None
• optim_target_modules: None
• batch_eval_metrics: False
• eval_on_start: False
• use_liger_kernel: False
• eval_use_gather_object: False
• average_tokens_across_devices: False
• prompts: None
• batch_sampler: no_duplicates
• multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	full_en_cosine_ndcg@200	full_es_cosine_ndcg@200	full_de_cosine_ndcg@200	full_zh_cosine_ndcg@200	mix_es_cosine_ndcg@200	mix_de_cosine_ndcg@200	mix_zh_cosine_ndcg@200
-1	-1	-	0.6856	0.5207	0.4655	0.6713	0.6224	0.5604	0.5548
0.0010	1	5.3354	-	-	-	-	-	-	-
0.1027	100	2.665	-	-	-	-	-	-	-
0.2053	200	1.3375	0.7691	0.6530	0.6298	0.7517	0.7513	0.7393	0.5490
0.3080	300	1.1101	-	-	-	-	-	-	-
0.4107	400	0.9453	0.7802	0.6643	0.6246	0.7531	0.7610	0.7441	0.5493
0.5133	500	0.9202	-	-	-	-	-	-	-
0.6160	600	0.7887	0.7741	0.6549	0.6171	0.7542	0.7672	0.7540	0.5482
0.7187	700	0.7604	-	-	-	-	-	-	-
0.8214	800	0.7219	0.7846	0.6674	0.6244	0.7648	0.7741	0.7592	0.5497
0.9240	900	0.6965	-	-	-	-	-	-	-
1.0267	1000	0.6253	0.7646	0.6391	0.6122	0.7503	0.7825	0.7704	0.5463
1.1294	1100	0.4737	-	-	-	-	-	-	-
1.2320	1200	0.5055	0.7758	0.6582	0.6178	0.7514	0.7857	0.7764	0.5501
1.3347	1300	0.5042	-	-	-	-	-	-	-
1.4374	1400	0.5073	0.7613	0.6578	0.6178	0.7505	0.7829	0.7762	0.5452
1.5400	1500	0.4975	-	-	-	-	-	-	-
1.6427	1600	0.5242	0.7736	0.6673	0.6279	0.7555	0.7940	0.7859	0.5477
1.7454	1700	0.4713	-	-	-	-	-	-	-
1.8480	1800	0.4814	0.7845	0.6733	0.6285	0.7642	0.7992	0.7904	0.5449
1.9507	1900	0.4526	-	-	-	-	-	-	-
2.0544	2000	0.36	0.7790	0.6639	0.6252	0.7500	0.8032	0.7888	0.5499
2.1571	2100	0.3744	-	-	-	-	-	-	-
2.2598	2200	0.3031	0.7787	0.6614	0.6190	0.7537	0.7993	0.7811	0.5476
2.3624	2300	0.3638	-	-	-	-	-	-	-
2.4651	2400	0.358	0.7798	0.6615	0.6258	0.7497	0.8018	0.7828	0.5481
2.5678	2500	0.3247	-	-	-	-	-	-	-
2.6704	2600	0.3247	0.7854	0.6663	0.6248	0.7560	0.8081	0.7835	0.5452
2.7731	2700	0.3263	-	-	-	-	-	-	-
2.8758	2800	0.3212	0.7761	0.6681	0.6250	0.7517	0.8121	0.7927	0.5458
2.9784	2900	0.3291	-	-	-	-	-	-	-
3.0821	3000	0.2816	0.7727	0.6604	0.6163	0.7370	0.8163	0.7985	0.5473
3.1848	3100	0.2698	-	-	-	-	-	-	-
3.2875	3200	0.2657	0.7757	0.6615	0.6247	0.7417	0.8117	0.8004	0.5436
3.3901	3300	0.2724	-	-	-	-	-	-	-
3.4928	3400	0.2584	0.7850	0.6583	0.6320	0.7458	0.8120	0.7980	0.5454
3.5955	3500	0.2573	-	-	-	-	-	-	-
3.6982	3600	0.2744	0.7796	0.6552	0.6237	0.7409	0.8193	0.8018	0.5466
3.8008	3700	0.3054	-	-	-	-	-	-	-
3.9035	3800	0.2727	0.7825	0.6642	0.6293	0.7504	0.8213	0.8058	0.5463
4.0062	3900	0.2353	-	-	-	-	-	-	-
4.1088	4000	0.2353	0.7747	0.6628	0.6263	0.7384	0.8239	0.8065	0.5447

Framework Versions

• Python: 3.11.11
• Sentence Transformers: 4.1.0
• Transformers: 4.51.2
• PyTorch: 2.6.0+cu124
• Accelerate: 1.6.0
• Datasets: 3.5.0
• Tokenizers: 0.21.1

Citation

BibTeX

Sentence Transformers

@inproceedings{reimers-2019-sentence-bert,
    title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
    author = "Reimers, Nils and Gurevych, Iryna",
    booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
    month = "11",
    year = "2019",
    publisher = "Association for Computational Linguistics",
    url = "https://arxiv.org/abs/1908.10084",
}

GISTEmbedLoss

@misc{solatorio2024gistembed,
    title={GISTEmbed: Guided In-sample Selection of Training Negatives for Text Embedding Fine-tuning},
    author={Aivin V. Solatorio},
    year={2024},
    eprint={2402.16829},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}