Add new SparseEncoder model

d9d6c87 verified 5 months ago

76.4 kB

metadata

language:
  - en
license: apache-2.0
tags:
  - sentence-transformers
  - sparse-encoder
  - sparse
  - splade
  - generated_from_trainer
  - dataset_size:99000
  - loss:SpladeLoss
  - loss:SparseMultipleNegativesRankingLoss
  - loss:FlopsLoss
base_model: distilbert/distilbert-base-uncased
widget:
  - text: >-
      The term emergent literacy signals a belief that, in a literate society,
      young children even one and two year olds, are in the process of becoming
      literate”. ... Gray (1956:21) notes: Functional literacy is used for the
      training of adults to 'meet independently the reading and writing demands
      placed on them'.
  - text: >-
      Rey is seemingly confirmed as being The Chosen One per a quote by a
      Lucasfilm production designer who worked on The Rise of Skywalker.
  - text: are union gun safes fireproof?
  - text: >-
      Fruit is an essential part of a healthy diet — and may aid weight loss.
      Most fruits are low in calories while high in nutrients and fiber, which
      can boost your fullness. Keep in mind that it's best to eat fruits whole
      rather than juiced. What's more, simply eating fruit is not the key to
      weight loss.
  - text: >-
      Treatment of suspected bacterial infection is with antibiotics, such as
      amoxicillin/clavulanate or doxycycline, given for 5 to 7 days for acute
      sinusitis and for up to 6 weeks for chronic sinusitis.
datasets:
  - sentence-transformers/gooaq
pipeline_tag: feature-extraction
library_name: sentence-transformers
metrics:
  - dot_accuracy@1
  - dot_accuracy@3
  - dot_accuracy@5
  - dot_accuracy@10
  - dot_precision@1
  - dot_precision@3
  - dot_precision@5
  - dot_precision@10
  - dot_recall@1
  - dot_recall@3
  - dot_recall@5
  - dot_recall@10
  - dot_ndcg@10
  - dot_mrr@10
  - dot_map@100
  - query_active_dims
  - query_sparsity_ratio
  - corpus_active_dims
  - corpus_sparsity_ratio
co2_eq_emissions:
  emissions: 13.27850984546157
  energy_consumed: 0.03416115647838594
  source: codecarbon
  training_type: fine-tuning
  on_cloud: false
  cpu_model: 13th Gen Intel(R) Core(TM) i7-13700K
  ram_total_size: 31.777088165283203
  hours_used: 0.142
  hardware_used: 1 x NVIDIA GeForce RTX 3090
model-index:
  - name: splade-distilbert-base-uncased trained on GooAQ
    results:
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoMSMARCO
          type: NanoMSMARCO
        metrics:
          - type: dot_accuracy@1
            value: 0.3
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.48
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.62
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.72
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.3
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.16
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.124
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.07200000000000001
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.3
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.48
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.62
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.72
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.4923029036903167
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.4212222222222222
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.4327140646742442
            name: Dot Map@100
          - type: query_active_dims
            value: 129.6999969482422
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9957506062201611
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 389.0450134277344
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9872536198994911
            name: Corpus Sparsity Ratio
          - type: dot_accuracy@1
            value: 0.32
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.48
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.62
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.7
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.32
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.16
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.124
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.07
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.32
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.48
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.62
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.7
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.497755451311688
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.4339126984126984
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.44682157418631346
            name: Dot Map@100
          - type: query_active_dims
            value: 121.5999984741211
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9960159885173278
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 397.7707824707031
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9869677353230226
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoNFCorpus
          type: NanoNFCorpus
        metrics:
          - type: dot_accuracy@1
            value: 0.36
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.44
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.46
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.54
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.36
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.2733333333333333
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.21600000000000003
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.182
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.016695854443026216
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.03106456335603726
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.039511512430362564
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.05434224491570579
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.21590295298422338
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.41285714285714287
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.07349060309007502
            name: Dot Map@100
          - type: query_active_dims
            value: 221.5399932861328
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9927416292088942
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 851.602783203125
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9720987227834637
            name: Corpus Sparsity Ratio
          - type: dot_accuracy@1
            value: 0.32
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.42
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.48
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.54
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.32
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.25999999999999995
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.22399999999999998
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.184
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.011195854443026216
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.0305213668117608
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.04137763819101165
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.054323362586038954
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.21092767617388347
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.388079365079365
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.07135787831660151
            name: Dot Map@100
          - type: query_active_dims
            value: 199.74000549316406
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9934558677185911
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 879.7700805664062
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9711758705010679
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoNQ
          type: NanoNQ
        metrics:
          - type: dot_accuracy@1
            value: 0.24
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.5
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.56
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.7
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.24
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.16666666666666663
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.11200000000000002
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.07
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.23
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.46
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.52
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.64
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.43543954013123615
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.3874365079365079
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.37672138136649846
            name: Dot Map@100
          - type: query_active_dims
            value: 103.12000274658203
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9966214532879044
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 603.1914672851562
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9802374855092997
            name: Corpus Sparsity Ratio
          - type: dot_accuracy@1
            value: 0.28
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.48
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.56
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.72
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.28
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.16
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.11599999999999999
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.07400000000000001
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.26
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.45
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.52
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.67
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.45351079818093326
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.40330158730158727
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.3886037227739609
            name: Dot Map@100
          - type: query_active_dims
            value: 97.54000091552734
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.996804272298161
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 625.50048828125
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9795065694161179
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-nano-beir
          name: Sparse Nano BEIR
        dataset:
          name: NanoBEIR mean
          type: NanoBEIR_mean
        metrics:
          - type: dot_accuracy@1
            value: 0.3
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.47333333333333333
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.5466666666666667
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.6533333333333333
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.3
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.19999999999999998
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.1506666666666667
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.108
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.18223195148100876
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.32368818778534575
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.39317050414345417
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.47144741497190196
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.3812151322685921
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.4071719576719577
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.2943086830436059
            name: Dot Map@100
          - type: query_active_dims
            value: 151.45333099365234
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9950378962389864
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 576.610088197042
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9811083779504279
            name: Corpus Sparsity Ratio
          - type: dot_accuracy@1
            value: 0.3591522762951334
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.5289795918367346
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.6121507064364207
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.7046153846153848
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.3591522762951334
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.2408895866038723
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.19369544740973316
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.14295133437990584
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.20232124722865225
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.31785179276573494
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.3843850663750471
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.4668945859961475
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.4149846764819604
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.46609898831327395
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.34193786183886477
            name: Dot Map@100
          - type: query_active_dims
            value: 207.32511528026893
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9932073548496079
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 571.211992111076
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9812852371367842
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoClimateFEVER
          type: NanoClimateFEVER
        metrics:
          - type: dot_accuracy@1
            value: 0.18
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.32
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.4
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.52
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.18
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.12
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.08800000000000001
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.06199999999999999
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.09166666666666666
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.16333333333333333
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.19
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.24733333333333335
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.20663969061747678
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.2785238095238095
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.16465402486405106
            name: Dot Map@100
          - type: query_active_dims
            value: 243.5399932861328
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9920208376487081
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 704.9307250976562
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.976904176492443
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoDBPedia
          type: NanoDBPedia
        metrics:
          - type: dot_accuracy@1
            value: 0.54
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.74
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.82
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.9
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.54
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.4666666666666666
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.43200000000000005
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.41200000000000003
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.055010013413515094
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.11580740263202617
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.15730927980831147
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.2871622888330674
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.49104405587222694
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.6685238095238095
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.34192873131869617
            name: Dot Map@100
          - type: query_active_dims
            value: 180.3000030517578
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.994092785431762
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 549.1513671875
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9820080149666633
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoFEVER
          type: NanoFEVER
        metrics:
          - type: dot_accuracy@1
            value: 0.54
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.68
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.76
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.84
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.54
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.22666666666666666
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.15600000000000003
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.08599999999999998
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.53
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.6466666666666666
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.7266666666666666
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.8066666666666665
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.6669183841429774
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.6358571428571428
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.6246746886992789
            name: Dot Map@100
          - type: query_active_dims
            value: 255.33999633789062
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9916342311664409
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 861.5084228515625
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9717741818081527
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoFiQA2018
          type: NanoFiQA2018
        metrics:
          - type: dot_accuracy@1
            value: 0.18
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.36
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.42
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.52
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.18
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.15333333333333332
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.12800000000000003
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.086
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.10933333333333334
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.23572222222222222
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.2821111111111111
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.35740476190476195
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.2731926938394576
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.28719047619047616
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.22109942326282112
            name: Dot Map@100
          - type: query_active_dims
            value: 88.83999633789062
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9970893127469402
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 414.41741943359375
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9864223373490076
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoHotpotQA
          type: NanoHotpotQA
        metrics:
          - type: dot_accuracy@1
            value: 0.66
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.72
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.78
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.9
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.66
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.33333333333333326
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.23199999999999996
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.13799999999999998
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.33
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.5
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.58
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.69
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.6058525828428769
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.7181269841269841
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.5210794772668003
            name: Dot Map@100
          - type: query_active_dims
            value: 147.39999389648438
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9951706967467242
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 622.540283203125
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9796035553632422
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoQuoraRetrieval
          type: NanoQuoraRetrieval
        metrics:
          - type: dot_accuracy@1
            value: 0.4
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.58
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.72
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.82
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.4
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.19333333333333333
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.15200000000000002
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.08599999999999998
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.3866666666666667
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.5466666666666667
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.6906666666666667
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.7906666666666666
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.5774223027465384
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.5223333333333333
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.5159175593795703
            name: Dot Map@100
          - type: query_active_dims
            value: 52.779998779296875
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9982707555606023
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 59.61296081542969
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9980468854984789
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoSCIDOCS
          type: NanoSCIDOCS
        metrics:
          - type: dot_accuracy@1
            value: 0.34
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.52
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.64
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.78
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.34
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.22
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.18
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.14
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.07266666666666667
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.13666666666666666
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.18566666666666662
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.2876666666666667
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.2663751860700744
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.453611111111111
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.17985066244400083
            name: Dot Map@100
          - type: query_active_dims
            value: 197.97999572753906
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9935135313633596
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 729.0889892578125
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9761126731781072
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoArguAna
          type: NanoArguAna
        metrics:
          - type: dot_accuracy@1
            value: 0.02
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.16
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.22
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.24
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.02
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.05333333333333333
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.044000000000000004
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.024000000000000004
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.02
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.16
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.22
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.24
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.1308713212722807
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.0955
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.10586268132966514
            name: Dot Map@100
          - type: query_active_dims
            value: 798.3800048828125
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9738424741208698
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 696.6200561523438
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9771764610395012
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoSciFact
          type: NanoSciFact
        metrics:
          - type: dot_accuracy@1
            value: 0.44
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.58
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.64
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.68
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.44
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.20666666666666667
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.14
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.078
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.415
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.555
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.615
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.67
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.5536377846083319
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.5218888888888888
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.5200703583453583
            name: Dot Map@100
          - type: query_active_dims
            value: 270.4800109863281
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9911381950400915
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 766.060302734375
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9749013726907028
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoTouche2020
          type: NanoTouche2020
        metrics:
          - type: dot_accuracy@1
            value: 0.4489795918367347
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.8367346938775511
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.8979591836734694
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 1
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.4489795918367347
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.5782312925170067
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.5020408163265306
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.4183673469387756
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.028637012782604658
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.11168898095521147
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.168207833765178
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.26840587129271604
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.46065286658673993
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.652437641723356
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.34327142171812286
            name: Dot Map@100
          - type: query_active_dims
            value: 37.918365478515625
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9987576710085015
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 507.71820068359375
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9833655002724724
            name: Corpus Sparsity Ratio

splade-distilbert-base-uncased trained on GooAQ

This is a SPLADE Sparse Encoder model finetuned from distilbert/distilbert-base-uncased on the gooaq dataset using the sentence-transformers library. It maps sentences & paragraphs to a 30522-dimensional sparse vector space and can be used for semantic search and sparse retrieval.

Model Details

Model Description

Model Type: SPLADE Sparse Encoder
Base model: distilbert/distilbert-base-uncased
Maximum Sequence Length: 256 tokens
Output Dimensionality: 30522 dimensions
Similarity Function: Dot Product
Training Dataset:
- gooaq
Language: en
License: apache-2.0

Model Sources

Documentation: Sentence Transformers Documentation
Documentation: Sparse Encoder Documentation
Repository: Sentence Transformers on GitHub
Hugging Face: Sparse Encoders on Hugging Face

Full Model Architecture

SparseEncoder(
  (0): MLMTransformer({'max_seq_length': 256, 'do_lower_case': False, 'architecture': 'DistilBertForMaskedLM'})
  (1): SpladePooling({'pooling_strategy': 'max', 'activation_function': 'relu', 'word_embedding_dimension': 30522})
)

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 SparseEncoder

# Download from the 🤗 Hub
model = SparseEncoder("tomaarsen/splade-distilbert-base-uncased-gooaq-peft-r64")
# Run inference
queries = [
    "how many days for doxycycline to work on sinus infection?",
]
documents = [
    'Treatment of suspected bacterial infection is with antibiotics, such as amoxicillin/clavulanate or doxycycline, given for 5 to 7 days for acute sinusitis and for up to 6 weeks for chronic sinusitis.',
    'Most engagements typically have a cocktail dress code, calling for dresses at, or slightly above, knee-length and high heels. If your party states a different dress code, however, such as semi-formal or dressy-casual, you may need to dress up or down accordingly.',
    'The average service life of a gas furnace is about 15 years, but the actual life span of an individual unit can vary greatly. There are a number of contributing factors that determine the age a furnace reaches: The quality of the equipment.',
]
query_embeddings = model.encode_query(queries)
document_embeddings = model.encode_document(documents)
print(query_embeddings.shape, document_embeddings.shape)
# [1, 30522] [3, 30522]

# Get the similarity scores for the embeddings
similarities = model.similarity(query_embeddings, document_embeddings)
print(similarities)
# tensor([[90.6936, 25.1699, 31.3902]])

Evaluation

Metrics

Sparse Information Retrieval

Datasets: NanoMSMARCO, NanoNFCorpus, NanoNQ, NanoClimateFEVER, NanoDBPedia, NanoFEVER, NanoFiQA2018, NanoHotpotQA, NanoMSMARCO, NanoNFCorpus, NanoNQ, NanoQuoraRetrieval, NanoSCIDOCS, NanoArguAna, NanoSciFact and NanoTouche2020
Evaluated with SparseInformationRetrievalEvaluator

Metric	NanoMSMARCO	NanoNFCorpus	NanoNQ	NanoClimateFEVER	NanoDBPedia	NanoFEVER	NanoFiQA2018	NanoHotpotQA	NanoQuoraRetrieval	NanoSCIDOCS	NanoArguAna	NanoSciFact	NanoTouche2020
dot_accuracy@1	0.32	0.32	0.28	0.18	0.54	0.54	0.18	0.66	0.4	0.34	0.02	0.44	0.449
dot_accuracy@3	0.48	0.42	0.48	0.32	0.74	0.68	0.36	0.72	0.58	0.52	0.16	0.58	0.8367
dot_accuracy@5	0.62	0.48	0.56	0.4	0.82	0.76	0.42	0.78	0.72	0.64	0.22	0.64	0.898
dot_accuracy@10	0.7	0.54	0.72	0.52	0.9	0.84	0.52	0.9	0.82	0.78	0.24	0.68	1.0
dot_precision@1	0.32	0.32	0.28	0.18	0.54	0.54	0.18	0.66	0.4	0.34	0.02	0.44	0.449
dot_precision@3	0.16	0.26	0.16	0.12	0.4667	0.2267	0.1533	0.3333	0.1933	0.22	0.0533	0.2067	0.5782
dot_precision@5	0.124	0.224	0.116	0.088	0.432	0.156	0.128	0.232	0.152	0.18	0.044	0.14	0.502
dot_precision@10	0.07	0.184	0.074	0.062	0.412	0.086	0.086	0.138	0.086	0.14	0.024	0.078	0.4184
dot_recall@1	0.32	0.0112	0.26	0.0917	0.055	0.53	0.1093	0.33	0.3867	0.0727	0.02	0.415	0.0286
dot_recall@3	0.48	0.0305	0.45	0.1633	0.1158	0.6467	0.2357	0.5	0.5467	0.1367	0.16	0.555	0.1117
dot_recall@5	0.62	0.0414	0.52	0.19	0.1573	0.7267	0.2821	0.58	0.6907	0.1857	0.22	0.615	0.1682
dot_recall@10	0.7	0.0543	0.67	0.2473	0.2872	0.8067	0.3574	0.69	0.7907	0.2877	0.24	0.67	0.2684
dot_ndcg@10	0.4978	0.2109	0.4535	0.2066	0.491	0.6669	0.2732	0.6059	0.5774	0.2664	0.1309	0.5536	0.4607
dot_mrr@10	0.4339	0.3881	0.4033	0.2785	0.6685	0.6359	0.2872	0.7181	0.5223	0.4536	0.0955	0.5219	0.6524
dot_map@100	0.4468	0.0714	0.3886	0.1647	0.3419	0.6247	0.2211	0.5211	0.5159	0.1799	0.1059	0.5201	0.3433
query_active_dims	121.6	199.74	97.54	243.54	180.3	255.34	88.84	147.4	52.78	197.98	798.38	270.48	37.9184
query_sparsity_ratio	0.996	0.9935	0.9968	0.992	0.9941	0.9916	0.9971	0.9952	0.9983	0.9935	0.9738	0.9911	0.9988
corpus_active_dims	397.7708	879.7701	625.5005	704.9307	549.1514	861.5084	414.4174	622.5403	59.613	729.089	696.6201	766.0603	507.7182
corpus_sparsity_ratio	0.987	0.9712	0.9795	0.9769	0.982	0.9718	0.9864	0.9796	0.998	0.9761	0.9772	0.9749	0.9834

Sparse Nano BEIR

Dataset: NanoBEIR_mean

Evaluated with SparseNanoBEIREvaluator with these parameters:

{
    "dataset_names": [
        "msmarco",
        "nfcorpus",
        "nq"
    ]
}

Metric	Value
dot_accuracy@1	0.3
dot_accuracy@3	0.4733
dot_accuracy@5	0.5467
dot_accuracy@10	0.6533
dot_precision@1	0.3
dot_precision@3	0.2
dot_precision@5	0.1507
dot_precision@10	0.108
dot_recall@1	0.1822
dot_recall@3	0.3237
dot_recall@5	0.3932
dot_recall@10	0.4714
dot_ndcg@10	0.3812
dot_mrr@10	0.4072
dot_map@100	0.2943
query_active_dims	151.4533
query_sparsity_ratio	0.995
corpus_active_dims	576.6101
corpus_sparsity_ratio	0.9811

Sparse Nano BEIR

Dataset: NanoBEIR_mean

Evaluated with SparseNanoBEIREvaluator with these parameters:

{
    "dataset_names": [
        "climatefever",
        "dbpedia",
        "fever",
        "fiqa2018",
        "hotpotqa",
        "msmarco",
        "nfcorpus",
        "nq",
        "quoraretrieval",
        "scidocs",
        "arguana",
        "scifact",
        "touche2020"
    ]
}

Metric	Value
dot_accuracy@1	0.3592
dot_accuracy@3	0.529
dot_accuracy@5	0.6122
dot_accuracy@10	0.7046
dot_precision@1	0.3592
dot_precision@3	0.2409
dot_precision@5	0.1937
dot_precision@10	0.143
dot_recall@1	0.2023
dot_recall@3	0.3179
dot_recall@5	0.3844
dot_recall@10	0.4669
dot_ndcg@10	0.415
dot_mrr@10	0.4661
dot_map@100	0.3419
query_active_dims	207.3251
query_sparsity_ratio	0.9932
corpus_active_dims	571.212
corpus_sparsity_ratio	0.9813

Training Details

Training Dataset

gooaq

Dataset: gooaq at b089f72
Size: 99,000 training samples
Columns: question and answer
Approximate statistics based on the first 1000 samples:
question answer
type string string
details
min: 8 tokens
mean: 11.79 tokens
max: 24 tokens

min: 14 tokens
mean: 60.02 tokens
max: 153 tokens

	question	answer
type	string	string
details	min: 8 tokens mean: 11.79 tokens max: 24 tokens	min: 14 tokens mean: 60.02 tokens max: 153 tokens

Samples:

question	answer
`what are the 5 characteristics of a star?`	`Key Concept: Characteristics used to classify stars include color, temperature, size, composition, and brightness.`
`are copic markers alcohol ink?`	`Copic Ink is alcohol-based and flammable. Keep away from direct sunlight and extreme temperatures.`
`what is the difference between appellate term and appellate division?`	`Appellate terms An appellate term is an intermediate appellate court that hears appeals from the inferior courts within their designated counties or judicial districts, and are intended to ease the workload on the Appellate Division and provide a less expensive forum closer to the people.`

Loss: SpladeLoss with these parameters:

{
    "loss": "SparseMultipleNegativesRankingLoss(scale=1.0, similarity_fct='dot_score')",
    "document_regularizer_weight": 3e-05,
    "query_regularizer_weight": 5e-05
}

Evaluation Dataset

gooaq

Dataset: gooaq at b089f72
Size: 1,000 evaluation samples
Columns: question and answer
Approximate statistics based on the first 1000 samples:
question answer
type string string
details
min: 8 tokens
mean: 11.93 tokens
max: 25 tokens

min: 14 tokens
mean: 60.84 tokens
max: 127 tokens

	question	answer
type	string	string
details	min: 8 tokens mean: 11.93 tokens max: 25 tokens	min: 14 tokens mean: 60.84 tokens max: 127 tokens

Samples:

question	answer
`should you take ibuprofen with high blood pressure?`	`In general, people with high blood pressure should use acetaminophen or possibly aspirin for over-the-counter pain relief. Unless your health care provider has said it's OK, you should not use ibuprofen, ketoprofen, or naproxen sodium. If aspirin or acetaminophen doesn't help with your pain, call your doctor.`
`how old do you have to be to work in sc?`	`The general minimum age of employment for South Carolina youth is 14, although the state allows younger children who are performers to work in show business. If their families are agricultural workers, children younger than age 14 may also participate in farm labor.`
`how to write a topic proposal for a research paper?`	`['Write down the main topic of your paper. ... ', 'Write two or three short sentences under the main topic that explain why you chose that topic. ... ', 'Write a thesis sentence that states the angle and purpose of your research paper. ... ', 'List the items you will cover in the body of the paper that support your thesis statement.']`

Loss: SpladeLoss with these parameters:

{
    "loss": "SparseMultipleNegativesRankingLoss(scale=1.0, similarity_fct='dot_score')",
    "document_regularizer_weight": 3e-05,
    "query_regularizer_weight": 5e-05
}

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: steps
per_device_train_batch_size: 32
per_device_eval_batch_size: 32
learning_rate: 2e-05
num_train_epochs: 1
bf16: True
load_best_model_at_end: 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: 32
per_device_eval_batch_size: 32
per_gpu_train_batch_size: None
per_gpu_eval_batch_size: None
gradient_accumulation_steps: 1
eval_accumulation_steps: None
torch_empty_cache_steps: None
learning_rate: 2e-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: 1
max_steps: -1
lr_scheduler_type: linear
lr_scheduler_kwargs: {}
warmup_ratio: 0.0
warmup_steps: 0
log_level: passive
log_level_replica: warning
log_on_each_node: True
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: True
fp16: False
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: False
dataloader_num_workers: 0
dataloader_prefetch_factor: None
past_index: -1
disable_tqdm: False
remove_unused_columns: True
label_names: None
load_best_model_at_end: True
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}
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: None
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
router_mapping: {}
learning_rate_mapping: {}

Training Logs

Epoch	Step	Training Loss	Validation Loss	NanoMSMARCO_dot_ndcg@10	NanoNFCorpus_dot_ndcg@10	NanoNQ_dot_ndcg@10	NanoBEIR_mean_dot_ndcg@10	NanoClimateFEVER_dot_ndcg@10	NanoDBPedia_dot_ndcg@10	NanoFEVER_dot_ndcg@10	NanoFiQA2018_dot_ndcg@10	NanoHotpotQA_dot_ndcg@10	NanoQuoraRetrieval_dot_ndcg@10	NanoSCIDOCS_dot_ndcg@10	NanoArguAna_dot_ndcg@10	NanoSciFact_dot_ndcg@10	NanoTouche2020_dot_ndcg@10
0.0323	100	132.3714	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.0646	200	17.6927	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.0970	300	2.8291	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.1293	400	0.9365	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.1616	500	0.5878	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.1939	600	0.4041	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.1972	610	-	0.2847	0.3951	0.1965	0.3212	0.3043	-	-	-	-	-	-	-	-	-	-
0.2262	700	0.3624	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.2586	800	0.295	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.2909	900	0.282	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.3232	1000	0.2947	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.3555	1100	0.2664	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.3878	1200	0.2387	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.3943	1220	-	0.2030	0.4818	0.2086	0.4084	0.3663	-	-	-	-	-	-	-	-	-	-
0.4202	1300	0.2512	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.4525	1400	0.2117	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.4848	1500	0.2367	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.5171	1600	0.2085	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.5495	1700	0.1745	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.5818	1800	0.1958	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.5915	1830	-	0.1449	0.4978	0.2109	0.4535	0.3874	-	-	-	-	-	-	-	-	-	-
0.6141	1900	0.1946	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.6464	2000	0.1801	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.6787	2100	0.1854	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.7111	2200	0.232	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.7434	2300	0.1798	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.7757	2400	0.1725	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.7886	2440	-	0.1437	0.4885	0.2206	0.4230	0.3774	-	-	-	-	-	-	-	-	-	-
0.8080	2500	0.1598	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.8403	2600	0.1586	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.8727	2700	0.172	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.9050	2800	0.1875	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.9373	2900	0.1691	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.9696	3000	0.1473	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.9858	3050	-	0.1358	0.4923	0.2159	0.4354	0.3812	-	-	-	-	-	-	-	-	-	-
-1	-1	-	-	0.4978	0.2109	0.4535	0.4150	0.2066	0.4910	0.6669	0.2732	0.6059	0.5774	0.2664	0.1309	0.5536	0.4607

The bold row denotes the saved checkpoint.

Environmental Impact

Carbon emissions were measured using CodeCarbon.

Energy Consumed: 0.034 kWh
Carbon Emitted: 0.013 kg of CO2
Hours Used: 0.142 hours

Training Hardware

On Cloud: No
GPU Model: 1 x NVIDIA GeForce RTX 3090
CPU Model: 13th Gen Intel(R) Core(TM) i7-13700K
RAM Size: 31.78 GB

Framework Versions

Python: 3.11.6
Sentence Transformers: 4.2.0.dev0
Transformers: 4.52.4
PyTorch: 2.7.1+cu126
Accelerate: 1.5.1
Datasets: 2.21.0
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",
}

SpladeLoss

@misc{formal2022distillationhardnegativesampling,
      title={From Distillation to Hard Negative Sampling: Making Sparse Neural IR Models More Effective},
      author={Thibault Formal and Carlos Lassance and Benjamin Piwowarski and Stéphane Clinchant},
      year={2022},
      eprint={2205.04733},
      archivePrefix={arXiv},
      primaryClass={cs.IR},
      url={https://arxiv.org/abs/2205.04733},
}

SparseMultipleNegativesRankingLoss

@misc{henderson2017efficient,
    title={Efficient Natural Language Response Suggestion for Smart Reply},
    author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
    year={2017},
    eprint={1705.00652},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}

FlopsLoss

@article{paria2020minimizing,
    title={Minimizing flops to learn efficient sparse representations},
    author={Paria, Biswajit and Yeh, Chih-Kuan and Yen, Ian EH and Xu, Ning and Ravikumar, Pradeep and P{'o}czos, Barnab{'a}s},
    journal={arXiv preprint arXiv:2004.05665},
    year={2020}
}