checkpoint-4400/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: 0.6571428571428571
            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: 0.6571428571428571
            name: Cosine Precision@1
          - type: cosine_precision@20
            value: 0.501904761904762
            name: Cosine Precision@20
          - type: cosine_precision@50
            value: 0.30514285714285716
            name: Cosine Precision@50
          - type: cosine_precision@100
            value: 0.18476190476190474
            name: Cosine Precision@100
          - type: cosine_precision@150
            value: 0.13238095238095238
            name: Cosine Precision@150
          - type: cosine_precision@200
            value: 0.10223809523809524
            name: Cosine Precision@200
          - type: cosine_recall@1
            value: 0.06749696615971254
            name: Cosine Recall@1
          - type: cosine_recall@20
            value: 0.5348166179254283
            name: Cosine Recall@20
          - type: cosine_recall@50
            value: 0.7176194992567407
            name: Cosine Recall@50
          - type: cosine_recall@100
            value: 0.8203546241789754
            name: Cosine Recall@100
          - type: cosine_recall@150
            value: 0.8712408549365904
            name: Cosine Recall@150
          - type: cosine_recall@200
            value: 0.8993000584751492
            name: Cosine Recall@200
          - type: cosine_ndcg@1
            value: 0.6571428571428571
            name: Cosine Ndcg@1
          - type: cosine_ndcg@20
            value: 0.6791929962471466
            name: Cosine Ndcg@20
          - type: cosine_ndcg@50
            value: 0.6958143211009435
            name: Cosine Ndcg@50
          - type: cosine_ndcg@100
            value: 0.7493655431536407
            name: Cosine Ndcg@100
          - type: cosine_ndcg@150
            value: 0.7715718645271473
            name: Cosine Ndcg@150
          - type: cosine_ndcg@200
            value: 0.7814931000676181
            name: Cosine Ndcg@200
          - type: cosine_mrr@1
            value: 0.6571428571428571
            name: Cosine Mrr@1
          - type: cosine_mrr@20
            value: 0.8026984126984127
            name: Cosine Mrr@20
          - type: cosine_mrr@50
            value: 0.8026984126984127
            name: Cosine Mrr@50
          - type: cosine_mrr@100
            value: 0.8026984126984127
            name: Cosine Mrr@100
          - type: cosine_mrr@150
            value: 0.8026984126984127
            name: Cosine Mrr@150
          - type: cosine_mrr@200
            value: 0.8026984126984127
            name: Cosine Mrr@200
          - type: cosine_map@1
            value: 0.6571428571428571
            name: Cosine Map@1
          - type: cosine_map@20
            value: 0.5371258373378305
            name: Cosine Map@20
          - type: cosine_map@50
            value: 0.5243155763407285
            name: Cosine Map@50
          - type: cosine_map@100
            value: 0.5561427452138551
            name: Cosine Map@100
          - type: cosine_map@150
            value: 0.5652920456249697
            name: Cosine Map@150
          - type: cosine_map@200
            value: 0.5681007357520309
            name: Cosine Map@200
          - type: cosine_map@500
            value: 0.5730541345190991
            name: Cosine Map@500
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: full es
          type: full_es
        metrics:
          - type: cosine_accuracy@1
            value: 0.10810810810810811
            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: 0.10810810810810811
            name: Cosine Precision@1
          - type: cosine_precision@20
            value: 0.5667567567567569
            name: Cosine Precision@20
          - type: cosine_precision@50
            value: 0.3877837837837838
            name: Cosine Precision@50
          - type: cosine_precision@100
            value: 0.25156756756756754
            name: Cosine Precision@100
          - type: cosine_precision@150
            value: 0.18954954954954953
            name: Cosine Precision@150
          - type: cosine_precision@200
            value: 0.15067567567567566
            name: Cosine Precision@200
          - type: cosine_recall@1
            value: 0.0033677005752683685
            name: Cosine Recall@1
          - type: cosine_recall@20
            value: 0.3790230473715137
            name: Cosine Recall@20
          - type: cosine_recall@50
            value: 0.5587328778405388
            name: Cosine Recall@50
          - type: cosine_recall@100
            value: 0.670664457795493
            name: Cosine Recall@100
          - type: cosine_recall@150
            value: 0.7335635895457856
            name: Cosine Recall@150
          - type: cosine_recall@200
            value: 0.766278425246947
            name: Cosine Recall@200
          - type: cosine_ndcg@1
            value: 0.10810810810810811
            name: Cosine Ndcg@1
          - type: cosine_ndcg@20
            value: 0.613008635976177
            name: Cosine Ndcg@20
          - type: cosine_ndcg@50
            value: 0.5878242736285791
            name: Cosine Ndcg@50
          - type: cosine_ndcg@100
            value: 0.6148703843706662
            name: Cosine Ndcg@100
          - type: cosine_ndcg@150
            value: 0.6471060871986968
            name: Cosine Ndcg@150
          - type: cosine_ndcg@200
            value: 0.6634453873788777
            name: Cosine Ndcg@200
          - type: cosine_mrr@1
            value: 0.10810810810810811
            name: Cosine Mrr@1
          - type: cosine_mrr@20
            value: 0.5509009009009009
            name: Cosine Mrr@20
          - type: cosine_mrr@50
            value: 0.5509009009009009
            name: Cosine Mrr@50
          - type: cosine_mrr@100
            value: 0.5509009009009009
            name: Cosine Mrr@100
          - type: cosine_mrr@150
            value: 0.5509009009009009
            name: Cosine Mrr@150
          - type: cosine_mrr@200
            value: 0.5509009009009009
            name: Cosine Mrr@200
          - type: cosine_map@1
            value: 0.10810810810810811
            name: Cosine Map@1
          - type: cosine_map@20
            value: 0.48105434805966624
            name: Cosine Map@20
          - type: cosine_map@50
            value: 0.42917908716630376
            name: Cosine Map@50
          - type: cosine_map@100
            value: 0.4322285035959748
            name: Cosine Map@100
          - type: cosine_map@150
            value: 0.4473320611795549
            name: Cosine Map@150
          - type: cosine_map@200
            value: 0.45413116686066823
            name: Cosine Map@200
          - type: cosine_map@500
            value: 0.4666628908850396
            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: 0.9852216748768473
            name: Cosine Accuracy@20
          - type: cosine_accuracy@50
            value: 0.9852216748768473
            name: Cosine Accuracy@50
          - type: cosine_accuracy@100
            value: 0.9901477832512315
            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: 0.5438423645320197
            name: Cosine Precision@20
          - type: cosine_precision@50
            value: 0.3827586206896551
            name: Cosine Precision@50
          - type: cosine_precision@100
            value: 0.2493103448275862
            name: Cosine Precision@100
          - type: cosine_precision@150
            value: 0.1868965517241379
            name: Cosine Precision@150
          - type: cosine_precision@200
            value: 0.150320197044335
            name: Cosine Precision@200
          - type: cosine_recall@1
            value: 0.01108543831680986
            name: Cosine Recall@1
          - type: cosine_recall@20
            value: 0.3450860009022403
            name: Cosine Recall@20
          - type: cosine_recall@50
            value: 0.5334236440941986
            name: Cosine Recall@50
          - type: cosine_recall@100
            value: 0.6498536020861698
            name: Cosine Recall@100
          - type: cosine_recall@150
            value: 0.7091695139240046
            name: Cosine Recall@150
          - type: cosine_recall@200
            value: 0.7496224791667186
            name: Cosine Recall@200
          - type: cosine_ndcg@1
            value: 0.2955665024630542
            name: Cosine Ndcg@1
          - type: cosine_ndcg@20
            value: 0.567054203369494
            name: Cosine Ndcg@20
          - type: cosine_ndcg@50
            value: 0.5519557348354142
            name: Cosine Ndcg@50
          - type: cosine_ndcg@100
            value: 0.5786968752325107
            name: Cosine Ndcg@100
          - type: cosine_ndcg@150
            value: 0.6099446866772629
            name: Cosine Ndcg@150
          - type: cosine_ndcg@200
            value: 0.6301254755200327
            name: Cosine Ndcg@200
          - type: cosine_mrr@1
            value: 0.2955665024630542
            name: Cosine Mrr@1
          - type: cosine_mrr@20
            value: 0.5163441238564384
            name: Cosine Mrr@20
          - type: cosine_mrr@50
            value: 0.5163441238564384
            name: Cosine Mrr@50
          - type: cosine_mrr@100
            value: 0.5164370692974646
            name: Cosine Mrr@100
          - type: cosine_mrr@150
            value: 0.5164370692974646
            name: Cosine Mrr@150
          - type: cosine_mrr@200
            value: 0.5164370692974646
            name: Cosine Mrr@200
          - type: cosine_map@1
            value: 0.2955665024630542
            name: Cosine Map@1
          - type: cosine_map@20
            value: 0.4243293426584066
            name: Cosine Map@20
          - type: cosine_map@50
            value: 0.37874837593471367
            name: Cosine Map@50
          - type: cosine_map@100
            value: 0.3817891460614099
            name: Cosine Map@100
          - type: cosine_map@150
            value: 0.39643664920094024
            name: Cosine Map@150
          - type: cosine_map@200
            value: 0.40443608704984707
            name: Cosine Map@200
          - type: cosine_map@500
            value: 0.4176754500966089
            name: Cosine Map@500
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: full zh
          type: full_zh
        metrics:
          - type: cosine_accuracy@1
            value: 0.6601941747572816
            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: 0.6601941747572816
            name: Cosine Precision@1
          - type: cosine_precision@20
            value: 0.47038834951456326
            name: Cosine Precision@20
          - type: cosine_precision@50
            value: 0.27941747572815534
            name: Cosine Precision@50
          - type: cosine_precision@100
            value: 0.17242718446601943
            name: Cosine Precision@100
          - type: cosine_precision@150
            value: 0.1239482200647249
            name: Cosine Precision@150
          - type: cosine_precision@200
            value: 0.09762135922330097
            name: Cosine Precision@200
          - type: cosine_recall@1
            value: 0.06553457566936532
            name: Cosine Recall@1
          - type: cosine_recall@20
            value: 0.5048889923213504
            name: Cosine Recall@20
          - type: cosine_recall@50
            value: 0.6723480900580502
            name: Cosine Recall@50
          - type: cosine_recall@100
            value: 0.7839824295594963
            name: Cosine Recall@100
          - type: cosine_recall@150
            value: 0.8346078714936033
            name: Cosine Recall@150
          - type: cosine_recall@200
            value: 0.868005364909913
            name: Cosine Recall@200
          - type: cosine_ndcg@1
            value: 0.6601941747572816
            name: Cosine Ndcg@1
          - type: cosine_ndcg@20
            value: 0.6489154249968472
            name: Cosine Ndcg@20
          - type: cosine_ndcg@50
            value: 0.6582544798801073
            name: Cosine Ndcg@50
          - type: cosine_ndcg@100
            value: 0.7132853867809429
            name: Cosine Ndcg@100
          - type: cosine_ndcg@150
            value: 0.7351428110305336
            name: Cosine Ndcg@150
          - type: cosine_ndcg@200
            value: 0.7489033638336042
            name: Cosine Ndcg@200
          - type: cosine_mrr@1
            value: 0.6601941747572816
            name: Cosine Mrr@1
          - type: cosine_mrr@20
            value: 0.8105987055016182
            name: Cosine Mrr@20
          - type: cosine_mrr@50
            value: 0.8105987055016182
            name: Cosine Mrr@50
          - type: cosine_mrr@100
            value: 0.8105987055016182
            name: Cosine Mrr@100
          - type: cosine_mrr@150
            value: 0.8105987055016182
            name: Cosine Mrr@150
          - type: cosine_mrr@200
            value: 0.8105987055016182
            name: Cosine Mrr@200
          - type: cosine_map@1
            value: 0.6601941747572816
            name: Cosine Map@1
          - type: cosine_map@20
            value: 0.5020661402654003
            name: Cosine Map@20
          - type: cosine_map@50
            value: 0.4804116383814884
            name: Cosine Map@50
          - type: cosine_map@100
            value: 0.5096988475054017
            name: Cosine Map@100
          - type: cosine_map@150
            value: 0.5182607426785758
            name: Cosine Map@150
          - type: cosine_map@200
            value: 0.5226490945380862
            name: Cosine Map@200
          - type: cosine_map@500
            value: 0.5274856682898562
            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: 0.9615184607384295
            name: Cosine Accuracy@20
          - type: cosine_accuracy@50
            value: 0.983359334373375
            name: Cosine Accuracy@50
          - type: cosine_accuracy@100
            value: 0.9921996879875195
            name: Cosine Accuracy@100
          - type: cosine_accuracy@150
            value: 0.9942797711908476
            name: Cosine Accuracy@150
          - type: cosine_accuracy@200
            value: 0.9947997919916797
            name: Cosine Accuracy@200
          - type: cosine_precision@1
            value: 0.7358294331773271
            name: Cosine Precision@1
          - type: cosine_precision@20
            value: 0.12477899115964639
            name: Cosine Precision@20
          - type: cosine_precision@50
            value: 0.05174206968278733
            name: Cosine Precision@50
          - type: cosine_precision@100
            value: 0.02630265210608425
            name: Cosine Precision@100
          - type: cosine_precision@150
            value: 0.017652972785578088
            name: Cosine Precision@150
          - type: cosine_precision@200
            value: 0.013294331773270933
            name: Cosine Precision@200
          - type: cosine_recall@1
            value: 0.28398831191342894
            name: Cosine Recall@1
          - type: cosine_recall@20
            value: 0.9220748829953198
            name: Cosine Recall@20
          - type: cosine_recall@50
            value: 0.9548448604610851
            name: Cosine Recall@50
          - type: cosine_recall@100
            value: 0.9711388455538221
            name: Cosine Recall@100
          - type: cosine_recall@150
            value: 0.9777604437510833
            name: Cosine Recall@150
          - type: cosine_recall@200
            value: 0.9823019587450165
            name: Cosine Recall@200
          - type: cosine_ndcg@1
            value: 0.7358294331773271
            name: Cosine Ndcg@1
          - type: cosine_ndcg@20
            value: 0.8104398530748719
            name: Cosine Ndcg@20
          - type: cosine_ndcg@50
            value: 0.8194810222604678
            name: Cosine Ndcg@50
          - type: cosine_ndcg@100
            value: 0.8230427127064399
            name: Cosine Ndcg@100
          - type: cosine_ndcg@150
            value: 0.8243283104602539
            name: Cosine Ndcg@150
          - type: cosine_ndcg@200
            value: 0.8251186561711241
            name: Cosine Ndcg@200
          - type: cosine_mrr@1
            value: 0.7358294331773271
            name: Cosine Mrr@1
          - type: cosine_mrr@20
            value: 0.8034886386855536
            name: Cosine Mrr@20
          - type: cosine_mrr@50
            value: 0.8042294348215404
            name: Cosine Mrr@50
          - type: cosine_mrr@100
            value: 0.8043610639446989
            name: Cosine Mrr@100
          - type: cosine_mrr@150
            value: 0.8043778926448901
            name: Cosine Mrr@150
          - type: cosine_mrr@200
            value: 0.8043807816493392
            name: Cosine Mrr@200
          - type: cosine_map@1
            value: 0.7358294331773271
            name: Cosine Map@1
          - type: cosine_map@20
            value: 0.742446597316252
            name: Cosine Map@20
          - type: cosine_map@50
            value: 0.7448760952950458
            name: Cosine Map@50
          - type: cosine_map@100
            value: 0.7453727938942869
            name: Cosine Map@100
          - type: cosine_map@150
            value: 0.7454980553388746
            name: Cosine Map@150
          - type: cosine_map@200
            value: 0.7455568923614244
            name: Cosine Map@200
          - type: cosine_map@500
            value: 0.7456455633479137
            name: Cosine Map@500
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: mix de
          type: mix_de
        metrics:
          - type: cosine_accuracy@1
            value: 0.6942277691107644
            name: Cosine Accuracy@1
          - type: cosine_accuracy@20
            value: 0.9667186687467498
            name: Cosine Accuracy@20
          - type: cosine_accuracy@50
            value: 0.983359334373375
            name: Cosine Accuracy@50
          - type: cosine_accuracy@100
            value: 0.9916796671866874
            name: Cosine Accuracy@100
          - type: cosine_accuracy@150
            value: 0.9932397295891836
            name: Cosine Accuracy@150
          - type: cosine_accuracy@200
            value: 0.9942797711908476
            name: Cosine Accuracy@200
          - type: cosine_precision@1
            value: 0.6942277691107644
            name: Cosine Precision@1
          - type: cosine_precision@20
            value: 0.12784711388455536
            name: Cosine Precision@20
          - type: cosine_precision@50
            value: 0.05319812792511702
            name: Cosine Precision@50
          - type: cosine_precision@100
            value: 0.0270306812272491
            name: Cosine Precision@100
          - type: cosine_precision@150
            value: 0.018110591090310275
            name: Cosine Precision@150
          - type: cosine_precision@200
            value: 0.013616744669786796
            name: Cosine Precision@200
          - type: cosine_recall@1
            value: 0.26120644825793027
            name: Cosine Recall@1
          - type: cosine_recall@20
            value: 0.927873114924597
            name: Cosine Recall@20
          - type: cosine_recall@50
            value: 0.9637285491419657
            name: Cosine Recall@50
          - type: cosine_recall@100
            value: 0.9789391575663027
            name: Cosine Recall@100
          - type: cosine_recall@150
            value: 0.9837926850407349
            name: Cosine Recall@150
          - type: cosine_recall@200
            value: 0.9862194487779511
            name: Cosine Recall@200
          - type: cosine_ndcg@1
            value: 0.6942277691107644
            name: Cosine Ndcg@1
          - type: cosine_ndcg@20
            value: 0.7952836406043297
            name: Cosine Ndcg@20
          - type: cosine_ndcg@50
            value: 0.8052399503452229
            name: Cosine Ndcg@50
          - type: cosine_ndcg@100
            value: 0.8086752401344494
            name: Cosine Ndcg@100
          - type: cosine_ndcg@150
            value: 0.8096382458419952
            name: Cosine Ndcg@150
          - type: cosine_ndcg@200
            value: 0.810085192105751
            name: Cosine Ndcg@200
          - type: cosine_mrr@1
            value: 0.6942277691107644
            name: Cosine Mrr@1
          - type: cosine_mrr@20
            value: 0.7761581892584265
            name: Cosine Mrr@20
          - type: cosine_mrr@50
            value: 0.7766868481375114
            name: Cosine Mrr@50
          - type: cosine_mrr@100
            value: 0.7768104145556238
            name: Cosine Mrr@100
          - type: cosine_mrr@150
            value: 0.7768244234791826
            name: Cosine Mrr@150
          - type: cosine_mrr@200
            value: 0.7768305684544853
            name: Cosine Mrr@200
          - type: cosine_map@1
            value: 0.6942277691107644
            name: Cosine Map@1
          - type: cosine_map@20
            value: 0.7188197545745756
            name: Cosine Map@20
          - type: cosine_map@50
            value: 0.7215707141808124
            name: Cosine Map@50
          - type: cosine_map@100
            value: 0.7220898692554206
            name: Cosine Map@100
          - type: cosine_map@150
            value: 0.7221900369972237
            name: Cosine Map@150
          - type: cosine_map@200
            value: 0.7222223600003219
            name: Cosine Map@200
          - type: cosine_map@500
            value: 0.7222810622423789
            name: Cosine Map@500
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: mix zh
          type: mix_zh
        metrics:
          - type: cosine_accuracy@1
            value: 0.18200728029121166
            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: 0.18200728029121166
            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: 0.05850234009360374
            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: 0.18200728029121166
            name: Cosine Ndcg@1
          - type: cosine_ndcg@20
            value: 0.5450053067257837
            name: Cosine Ndcg@20
          - type: cosine_ndcg@50
            value: 0.5450053067257837
            name: Cosine Ndcg@50
          - type: cosine_ndcg@100
            value: 0.5450053067257837
            name: Cosine Ndcg@100
          - type: cosine_ndcg@150
            value: 0.5450053067257837
            name: Cosine Ndcg@150
          - type: cosine_ndcg@200
            value: 0.5450053067257837
            name: Cosine Ndcg@200
          - type: cosine_mrr@1
            value: 0.18200728029121166
            name: Cosine Mrr@1
          - type: cosine_mrr@20
            value: 0.40246777114951904
            name: Cosine Mrr@20
          - type: cosine_mrr@50
            value: 0.40246777114951904
            name: Cosine Mrr@50
          - type: cosine_mrr@100
            value: 0.40246777114951904
            name: Cosine Mrr@100
          - type: cosine_mrr@150
            value: 0.40246777114951904
            name: Cosine Mrr@150
          - type: cosine_mrr@200
            value: 0.40246777114951904
            name: Cosine Mrr@200
          - type: cosine_map@1
            value: 0.18200728029121166
            name: Cosine Map@1
          - type: cosine_map@20
            value: 0.3277096647667185
            name: Cosine Map@20
          - type: cosine_map@50
            value: 0.3277096647667185
            name: Cosine Map@50
          - type: cosine_map@100
            value: 0.3277096647667185
            name: Cosine Map@100
          - type: cosine_map@150
            value: 0.3277096647667185
            name: Cosine Map@150
          - type: cosine_map@200
            value: 0.3277096647667185
            name: Cosine Map@200
          - type: cosine_map@500
            value: 0.3277096647667185
            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	0.6571	0.1081	0.2956	0.6602	0.7358	0.6942	0.182
cosine_accuracy@20	0.9905	1.0	0.9852	0.9903	0.9615	0.9667	1.0
cosine_accuracy@50	0.9905	1.0	0.9852	0.9903	0.9834	0.9834	1.0
cosine_accuracy@100	0.9905	1.0	0.9901	0.9903	0.9922	0.9917	1.0
cosine_accuracy@150	0.9905	1.0	0.9901	0.9903	0.9943	0.9932	1.0
cosine_accuracy@200	0.9905	1.0	0.9901	0.9903	0.9948	0.9943	1.0
cosine_precision@1	0.6571	0.1081	0.2956	0.6602	0.7358	0.6942	0.182
cosine_precision@20	0.5019	0.5668	0.5438	0.4704	0.1248	0.1278	0.1544
cosine_precision@50	0.3051	0.3878	0.3828	0.2794	0.0517	0.0532	0.0618
cosine_precision@100	0.1848	0.2516	0.2493	0.1724	0.0263	0.027	0.0309
cosine_precision@150	0.1324	0.1895	0.1869	0.1239	0.0177	0.0181	0.0206
cosine_precision@200	0.1022	0.1507	0.1503	0.0976	0.0133	0.0136	0.0154
cosine_recall@1	0.0675	0.0034	0.0111	0.0655	0.284	0.2612	0.0585
cosine_recall@20	0.5348	0.379	0.3451	0.5049	0.9221	0.9279	1.0
cosine_recall@50	0.7176	0.5587	0.5334	0.6723	0.9548	0.9637	1.0
cosine_recall@100	0.8204	0.6707	0.6499	0.784	0.9711	0.9789	1.0
cosine_recall@150	0.8712	0.7336	0.7092	0.8346	0.9778	0.9838	1.0
cosine_recall@200	0.8993	0.7663	0.7496	0.868	0.9823	0.9862	1.0
cosine_ndcg@1	0.6571	0.1081	0.2956	0.6602	0.7358	0.6942	0.182
cosine_ndcg@20	0.6792	0.613	0.5671	0.6489	0.8104	0.7953	0.545
cosine_ndcg@50	0.6958	0.5878	0.552	0.6583	0.8195	0.8052	0.545
cosine_ndcg@100	0.7494	0.6149	0.5787	0.7133	0.823	0.8087	0.545
cosine_ndcg@150	0.7716	0.6471	0.6099	0.7351	0.8243	0.8096	0.545
cosine_ndcg@200	0.7815	0.6634	0.6301	0.7489	0.8251	0.8101	0.545
cosine_mrr@1	0.6571	0.1081	0.2956	0.6602	0.7358	0.6942	0.182
cosine_mrr@20	0.8027	0.5509	0.5163	0.8106	0.8035	0.7762	0.4025
cosine_mrr@50	0.8027	0.5509	0.5163	0.8106	0.8042	0.7767	0.4025
cosine_mrr@100	0.8027	0.5509	0.5164	0.8106	0.8044	0.7768	0.4025
cosine_mrr@150	0.8027	0.5509	0.5164	0.8106	0.8044	0.7768	0.4025
cosine_mrr@200	0.8027	0.5509	0.5164	0.8106	0.8044	0.7768	0.4025
cosine_map@1	0.6571	0.1081	0.2956	0.6602	0.7358	0.6942	0.182
cosine_map@20	0.5371	0.4811	0.4243	0.5021	0.7424	0.7188	0.3277
cosine_map@50	0.5243	0.4292	0.3787	0.4804	0.7449	0.7216	0.3277
cosine_map@100	0.5561	0.4322	0.3818	0.5097	0.7454	0.7221	0.3277
cosine_map@150	0.5653	0.4473	0.3964	0.5183	0.7455	0.7222	0.3277
cosine_map@200	0.5681	0.4541	0.4044	0.5226	0.7456	0.7222	0.3277
cosine_map@500	0.5731	0.4667	0.4177	0.5275	0.7456	0.7223	0.3277

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
4.2115	4100	0.2385	-	-	-	-	-	-	-
4.3142	4200	0.231	0.7811	0.6608	0.6254	0.7463	0.8226	0.8051	0.5442
4.4168	4300	0.2115	-	-	-	-	-	-	-
4.5195	4400	0.2151	0.7815	0.6634	0.6301	0.7489	0.8251	0.8101	0.5450

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}
}