metadata

language:
  - en
license: apache-2.0
tags:
  - sentence-transformers
  - sparse-encoder
  - sparse
  - csr
  - generated_from_trainer
  - dataset_size:99000
  - loss:CSRLoss
  - loss:SparseMultipleNegativesRankingLoss
base_model: mixedbread-ai/mxbai-embed-large-v1
widget:
  - text: >-
      Saudi Arabia–United Arab Emirates relations However, the UAE and Saudi
      Arabia continue to take somewhat differing stances on regional conflicts
      such the Yemeni Civil War, where the UAE opposes Al-Islah, and supports
      the Southern Movement, which has fought against Saudi-backed forces, and
      the Syrian Civil War, where the UAE has disagreed with Saudi support for
      Islamist movements.[4]
  - text: >-
      Economy of New Zealand New Zealand's diverse market economy has a sizable
      service sector, accounting for 63% of all GDP activity in 2013.[17] Large
      scale manufacturing industries include aluminium production, food
      processing, metal fabrication, wood and paper products. Mining,
      manufacturing, electricity, gas, water, and waste services accounted for
      16.5% of GDP in 2013.[17] The primary sector continues to dominate New
      Zealand's exports, despite accounting for 6.5% of GDP in 2013.[17]
  - text: >-
      who was the first president of indian science congress meeting held in
      kolkata in 1914
  - text: >-
      Get Over It (Eagles song) "Get Over It" is a song by the Eagles released
      as a single after a fourteen-year breakup. It was also the first song
      written by bandmates Don Henley and Glenn Frey when the band reunited.
      "Get Over It" was played live for the first time during their Hell Freezes
      Over tour in 1994. It returned the band to the U.S. Top 40 after a
      fourteen-year absence, peaking at No. 31 on the Billboard Hot 100 chart.
      It also hit No. 4 on the Billboard Mainstream Rock Tracks chart. The song
      was not played live by the Eagles after the "Hell Freezes Over" tour in
      1994. It remains the group's last Top 40 hit in the U.S.
  - text: >-
      Cornelius the Centurion Cornelius (Greek: Κορνήλιος) was a Roman centurion
      who is considered by Christians to be one of the first Gentiles to convert
      to the faith, as related in Acts of the Apostles.
datasets:
  - sentence-transformers/natural-questions
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: 53.0273650168183
  energy_consumed: 0.13642164181511365
  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: 0.41
  hardware_used: 1 x NVIDIA GeForce RTX 3090
model-index:
  - name: Sparse CSR model trained on Natural Questions
    results:
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoMSMARCO 128
          type: NanoMSMARCO_128
        metrics:
          - type: dot_accuracy@1
            value: 0.38
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.66
            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: 0.38
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.22
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.14400000000000002
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.08199999999999999
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.38
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.66
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.72
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.82
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.6074833126260415
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.5392698412698412
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.5478391044500884
            name: Dot Map@100
          - type: query_active_dims
            value: 128
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.96875
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 128
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.96875
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoNFCorpus 128
          type: NanoNFCorpus_128
        metrics:
          - type: dot_accuracy@1
            value: 0.44
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.54
            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: 0.3133333333333333
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.28
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.24600000000000002
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.045132854073603
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.06751477851868476
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.08765169300408888
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.12035202437952344
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.3037747903284991
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.5081904761904761
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.13867493157888547
            name: Dot Map@100
          - type: query_active_dims
            value: 128
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.96875
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 128
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.96875
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoNQ 128
          type: NanoNQ_128
        metrics:
          - type: dot_accuracy@1
            value: 0.48
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.66
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.7
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.84
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.48
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.22666666666666668
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.14800000000000002
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.08999999999999998
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.45
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.62
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.67
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.81
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.6337677207897237
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.5932936507936507
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.5761859932841973
            name: Dot Map@100
          - type: query_active_dims
            value: 128
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.96875
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 128
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.96875
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-nano-beir
          name: Sparse Nano BEIR
        dataset:
          name: NanoBEIR mean 128
          type: NanoBEIR_mean_128
        metrics:
          - type: dot_accuracy@1
            value: 0.43333333333333335
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.6200000000000001
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.6866666666666665
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.7799999999999999
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.43333333333333335
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.25333333333333335
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.19066666666666668
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.13933333333333334
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.2917109513578677
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.44917159283956165
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.49255056433469635
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.5834506747931745
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.5150086079147548
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.5469179894179893
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.42090000977105707
            name: Dot Map@100
          - type: query_active_dims
            value: 128
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.96875
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 128
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.96875
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoMSMARCO 256
          type: NanoMSMARCO_256
        metrics:
          - type: dot_accuracy@1
            value: 0.44
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.64
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.74
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.84
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.44
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.21333333333333332
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.14800000000000002
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.08399999999999999
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.44
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.64
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.74
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.84
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.6405150998246686
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.5768809523809523
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.5851061967133396
            name: Dot Map@100
          - type: query_active_dims
            value: 256
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9375
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 256
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9375
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoNFCorpus 256
          type: NanoNFCorpus_256
        metrics:
          - type: dot_accuracy@1
            value: 0.42
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.58
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.6
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.62
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.42
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.37333333333333324
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.324
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.248
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.045123947439696374
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.08083248635236362
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.0993952531376598
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.1259275313458498
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.3181127342430942
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.5041666666666667
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.15847418838222901
            name: Dot Map@100
          - type: query_active_dims
            value: 256
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9375
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 256
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9375
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoNQ 256
          type: NanoNQ_256
        metrics:
          - type: dot_accuracy@1
            value: 0.54
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.7
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.8
            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: 0.24
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.16799999999999998
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.092
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.51
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.66
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.75
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.81
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.6642484604451891
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.6294126984126983
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.6162769242153361
            name: Dot Map@100
          - type: query_active_dims
            value: 256
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9375
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 256
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9375
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-nano-beir
          name: Sparse Nano BEIR
        dataset:
          name: NanoBEIR mean 256
          type: NanoBEIR_mean_256
        metrics:
          - type: dot_accuracy@1
            value: 0.4666666666666666
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.64
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.7133333333333333
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.7666666666666666
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.4666666666666666
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.2755555555555555
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.21333333333333335
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.1413333333333333
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.3317079824798988
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.46027749545078783
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.5297984177125533
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.5919758437819499
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.5409587648376507
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.570153439153439
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.4532857697703016
            name: Dot Map@100
          - type: query_active_dims
            value: 256
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9375
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 256
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9375
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoClimateFEVER
          type: NanoClimateFEVER
        metrics:
          - type: dot_accuracy@1
            value: 0.28
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.52
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.7
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.8
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.28
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.18666666666666668
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.16799999999999998
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.10799999999999997
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.12166666666666665
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.23233333333333334
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.348
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.42633333333333334
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.33235923006734097
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.43644444444444447
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.24903211945618525
            name: Dot Map@100
          - type: query_active_dims
            value: 256
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9375
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 256
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9375
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoDBPedia
          type: NanoDBPedia
        metrics:
          - type: dot_accuracy@1
            value: 0.8
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.9
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.9
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.92
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.8
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.6466666666666666
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.56
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.474
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.09128542236179474
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.17409405829521904
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.22516141018064886
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.321390285824061
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.600179050204524
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.8425
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.45264984932006563
            name: Dot Map@100
          - type: query_active_dims
            value: 256
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9375
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 256
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9375
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoFEVER
          type: NanoFEVER
        metrics:
          - type: dot_accuracy@1
            value: 0.84
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.92
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.96
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.96
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.84
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.32
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.19999999999999996
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.09999999999999998
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.7866666666666667
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.8866666666666667
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.9266666666666667
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.9266666666666667
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.8816129048397259
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.89
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.8589881484317317
            name: Dot Map@100
          - type: query_active_dims
            value: 256
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9375
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 256
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9375
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoFiQA2018
          type: NanoFiQA2018
        metrics:
          - type: dot_accuracy@1
            value: 0.48
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.6
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.64
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.74
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.48
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.3066666666666667
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.22399999999999998
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.13599999999999998
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.2592460317460317
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.39734920634920634
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.4497857142857143
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.5795634920634921
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.48812055653800884
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.5517460317460319
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.42554170336694114
            name: Dot Map@100
          - type: query_active_dims
            value: 256
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9375
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 256
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9375
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoHotpotQA
          type: NanoHotpotQA
        metrics:
          - type: dot_accuracy@1
            value: 0.84
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.96
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.96
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.98
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.84
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.5133333333333333
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.32799999999999996
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.16999999999999996
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.42
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.77
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.82
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.85
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.8106522538764799
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.8966666666666666
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.7565706035126855
            name: Dot Map@100
          - type: query_active_dims
            value: 256
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9375
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 256
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9375
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoMSMARCO
          type: NanoMSMARCO
        metrics:
          - type: dot_accuracy@1
            value: 0.44
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.62
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.74
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.84
            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: 0.14800000000000002
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.08399999999999999
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.44
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.62
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.74
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.84
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.6329477813439243
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.5677777777777777
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.5762304873870092
            name: Dot Map@100
          - type: query_active_dims
            value: 256
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9375
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 256
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9375
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoNFCorpus
          type: NanoNFCorpus
        metrics:
          - type: dot_accuracy@1
            value: 0.42
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.56
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.64
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.66
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.42
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.37999999999999995
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.34800000000000003
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.258
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.04486258380333274
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.08768477299713343
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.10844641112515632
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.135531563356284
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.3285187113745097
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.5009999999999999
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.16174125549238802
            name: Dot Map@100
          - type: query_active_dims
            value: 256
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9375
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 256
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9375
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoNQ
          type: NanoNQ
        metrics:
          - type: dot_accuracy@1
            value: 0.58
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.7
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.8
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.82
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.58
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.24
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.16799999999999998
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.08999999999999998
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.55
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.66
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.75
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.79
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.677342414343143
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.6521666666666666
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.6420660106369513
            name: Dot Map@100
          - type: query_active_dims
            value: 256
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9375
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 256
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9375
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoQuoraRetrieval
          type: NanoQuoraRetrieval
        metrics:
          - type: dot_accuracy@1
            value: 0.9
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 1
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 1
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 1
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.9
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.4133333333333333
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.27199999999999996
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.13799999999999998
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.7773333333333333
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.9620000000000001
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.9933333333333334
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.9966666666666666
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.9509657098958008
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.9466666666666665
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.9297051282051282
            name: Dot Map@100
          - type: query_active_dims
            value: 256
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9375
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 256
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9375
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoSCIDOCS
          type: NanoSCIDOCS
        metrics:
          - type: dot_accuracy@1
            value: 0.42
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.72
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.82
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.88
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.42
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.35333333333333333
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.3
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.20800000000000002
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.09066666666666666
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.22166666666666665
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.3096666666666667
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.42566666666666664
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.4022717287490821
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.5887222222222221
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.32075091248131626
            name: Dot Map@100
          - type: query_active_dims
            value: 256
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9375
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 256
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9375
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoArguAna
          type: NanoArguAna
        metrics:
          - type: dot_accuracy@1
            value: 0.38
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.7
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.8
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.92
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.38
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.23333333333333336
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.16
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.092
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.38
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.7
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.8
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.92
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.6550827948648061
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.5706349206349206
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.5760927960927961
            name: Dot Map@100
          - type: query_active_dims
            value: 256
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9375
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 256
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9375
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoSciFact
          type: NanoSciFact
        metrics:
          - type: dot_accuracy@1
            value: 0.62
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.72
            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: 0.62
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.26666666666666666
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.17199999999999996
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.09599999999999997
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.595
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.705
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.755
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.84
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.7193800580696723
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.6823888888888889
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.6850911930363545
            name: Dot Map@100
          - type: query_active_dims
            value: 256
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9375
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 256
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9375
            name: Corpus Sparsity Ratio
      - task:
          type: sparse-information-retrieval
          name: Sparse Information Retrieval
        dataset:
          name: NanoTouche2020
          type: NanoTouche2020
        metrics:
          - type: dot_accuracy@1
            value: 0.4897959183673469
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.8367346938775511
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.9591836734693877
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.9795918367346939
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.4897959183673469
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.5170068027210885
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.5346938775510204
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.4346938775510204
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.03422245985964837
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.10897367065265
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.18115391425134045
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.2884686031356881
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.47678328743473813
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.6784580498866212
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.3590479959667369
            name: Dot Map@100
          - type: query_active_dims
            value: 256
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9375
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 256
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9375
            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: 0.576138147566719
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.7505180533751962
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.821475667189953
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.8722762951334379
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.576138147566719
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.3525902668759811
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.27559183673469384
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.18374568288854
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.35314998700801087
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.5019821826892981
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.5697857012699635
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.6415605598240661
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.6120166524309044
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.6773209488923774
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.5379621694912531
            name: Dot Map@100
          - type: query_active_dims
            value: 256
            name: Query Active Dims
          - type: query_sparsity_ratio
            value: 0.9375
            name: Query Sparsity Ratio
          - type: corpus_active_dims
            value: 256
            name: Corpus Active Dims
          - type: corpus_sparsity_ratio
            value: 0.9375
            name: Corpus Sparsity Ratio

Sparse CSR model trained on Natural Questions

This is a CSR Sparse Encoder model finetuned from mixedbread-ai/mxbai-embed-large-v1 on the natural-questions dataset using the sentence-transformers library. It maps sentences & paragraphs to a 4096-dimensional sparse vector space with 256 maximum active dimensions and can be used for semantic search and sparse retrieval.

Model Details

Model Description

Model Type: CSR Sparse Encoder
Base model: mixedbread-ai/mxbai-embed-large-v1
Maximum Sequence Length: 512 tokens
Output Dimensionality: 4096 dimensions (trained with 256 maximum active dimensions)
Similarity Function: Dot Product
Training Dataset:
- natural-questions
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): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel 
  (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): CSRSparsity({'input_dim': 1024, 'hidden_dim': 4096, 'k': 256, 'k_aux': 512, 'normalize': False, 'dead_threshold': 30})
)

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/csr-mxbai-embed-large-v1-nq-updated-reconstruction-2")
# Run inference
queries = [
    "who is cornelius in the book of acts",
]
documents = [
    'Cornelius the Centurion Cornelius (Greek: Κορνήλιος) was a Roman centurion who is considered by Christians to be one of the first Gentiles to convert to the faith, as related in Acts of the Apostles.',
    "Joe Ranft Ranft reunited with Lasseter when he was hired by Pixar in 1991 as their head of story.[1] There he worked on all of their films produced up to 2006; this included Toy Story (for which he received an Academy Award nomination) and A Bug's Life, as the co-story writer and others as story supervisor. His final film was Cars. He also voiced characters in many of the films, including Heimlich the caterpillar in A Bug's Life, Wheezy the penguin in Toy Story 2, and Jacques the shrimp in Finding Nemo.[1]",
    'Wonderful Tonight "Wonderful Tonight" is a ballad written by Eric Clapton. It was included on Clapton\'s 1977 album Slowhand. Clapton wrote the song about Pattie Boyd.[1] The female vocal harmonies on the song are provided by Marcella Detroit (then Marcy Levy) and Yvonne Elliman.',
]
query_embeddings = model.encode_query(queries)
document_embeddings = model.encode_document(documents)
print(query_embeddings.shape, document_embeddings.shape)
# [1, 4096] [3, 4096]

# Get the similarity scores for the embeddings
similarities = model.similarity(query_embeddings, document_embeddings)
print(similarities)
# tensor([[118.6570,  32.2072,  21.3971]])

Evaluation

Metrics

Sparse Information Retrieval

Datasets: NanoMSMARCO_128, NanoNFCorpus_128 and NanoNQ_128
Evaluated with SparseInformationRetrievalEvaluator with these parameters:
```
{
    "max_active_dims": 128
}
```

Metric	NanoMSMARCO_128	NanoNFCorpus_128	NanoNQ_128
dot_accuracy@1	0.38	0.44	0.48
dot_accuracy@3	0.66	0.54	0.66
dot_accuracy@5	0.72	0.64	0.7
dot_accuracy@10	0.82	0.68	0.84
dot_precision@1	0.38	0.44	0.48
dot_precision@3	0.22	0.3133	0.2267
dot_precision@5	0.144	0.28	0.148
dot_precision@10	0.082	0.246	0.09
dot_recall@1	0.38	0.0451	0.45
dot_recall@3	0.66	0.0675	0.62
dot_recall@5	0.72	0.0877	0.67
dot_recall@10	0.82	0.1204	0.81
dot_ndcg@10	0.6075	0.3038	0.6338
dot_mrr@10	0.5393	0.5082	0.5933
dot_map@100	0.5478	0.1387	0.5762
query_active_dims	128.0	128.0	128.0
query_sparsity_ratio	0.9688	0.9688	0.9688
corpus_active_dims	128.0	128.0	128.0
corpus_sparsity_ratio	0.9688	0.9688	0.9688

Sparse Nano BEIR

Dataset: NanoBEIR_mean_128

Evaluated with SparseNanoBEIREvaluator with these parameters:

{
    "dataset_names": [
        "msmarco",
        "nfcorpus",
        "nq"
    ],
    "max_active_dims": 128
}

Metric	Value
dot_accuracy@1	0.4333
dot_accuracy@3	0.62
dot_accuracy@5	0.6867
dot_accuracy@10	0.78
dot_precision@1	0.4333
dot_precision@3	0.2533
dot_precision@5	0.1907
dot_precision@10	0.1393
dot_recall@1	0.2917
dot_recall@3	0.4492
dot_recall@5	0.4926
dot_recall@10	0.5835
dot_ndcg@10	0.515
dot_mrr@10	0.5469
dot_map@100	0.4209
query_active_dims	128.0
query_sparsity_ratio	0.9688
corpus_active_dims	128.0
corpus_sparsity_ratio	0.9688

Sparse Information Retrieval

Datasets: NanoMSMARCO_256, NanoNFCorpus_256 and NanoNQ_256
Evaluated with SparseInformationRetrievalEvaluator with these parameters:
```
{
    "max_active_dims": 256
}
```

Metric	NanoMSMARCO_256	NanoNFCorpus_256	NanoNQ_256
dot_accuracy@1	0.44	0.42	0.54
dot_accuracy@3	0.64	0.58	0.7
dot_accuracy@5	0.74	0.6	0.8
dot_accuracy@10	0.84	0.62	0.84
dot_precision@1	0.44	0.42	0.54
dot_precision@3	0.2133	0.3733	0.24
dot_precision@5	0.148	0.324	0.168
dot_precision@10	0.084	0.248	0.092
dot_recall@1	0.44	0.0451	0.51
dot_recall@3	0.64	0.0808	0.66
dot_recall@5	0.74	0.0994	0.75
dot_recall@10	0.84	0.1259	0.81
dot_ndcg@10	0.6405	0.3181	0.6642
dot_mrr@10	0.5769	0.5042	0.6294
dot_map@100	0.5851	0.1585	0.6163
query_active_dims	256.0	256.0	256.0
query_sparsity_ratio	0.9375	0.9375	0.9375
corpus_active_dims	256.0	256.0	256.0
corpus_sparsity_ratio	0.9375	0.9375	0.9375

Sparse Nano BEIR

Dataset: NanoBEIR_mean_256

Evaluated with SparseNanoBEIREvaluator with these parameters:

{
    "dataset_names": [
        "msmarco",
        "nfcorpus",
        "nq"
    ],
    "max_active_dims": 256
}

Metric	Value
dot_accuracy@1	0.4667
dot_accuracy@3	0.64
dot_accuracy@5	0.7133
dot_accuracy@10	0.7667
dot_precision@1	0.4667
dot_precision@3	0.2756
dot_precision@5	0.2133
dot_precision@10	0.1413
dot_recall@1	0.3317
dot_recall@3	0.4603
dot_recall@5	0.5298
dot_recall@10	0.592
dot_ndcg@10	0.541
dot_mrr@10	0.5702
dot_map@100	0.4533
query_active_dims	256.0
query_sparsity_ratio	0.9375
corpus_active_dims	256.0
corpus_sparsity_ratio	0.9375

Sparse Information Retrieval

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

Metric	NanoClimateFEVER	NanoDBPedia	NanoFEVER	NanoFiQA2018	NanoHotpotQA	NanoMSMARCO	NanoNFCorpus	NanoNQ	NanoQuoraRetrieval	NanoSCIDOCS	NanoArguAna	NanoSciFact	NanoTouche2020
dot_accuracy@1	0.28	0.8	0.84	0.48	0.84	0.44	0.42	0.58	0.9	0.42	0.38	0.62	0.4898
dot_accuracy@3	0.52	0.9	0.92	0.6	0.96	0.62	0.56	0.7	1.0	0.72	0.7	0.72	0.8367
dot_accuracy@5	0.7	0.9	0.96	0.64	0.96	0.74	0.64	0.8	1.0	0.82	0.8	0.76	0.9592
dot_accuracy@10	0.8	0.92	0.96	0.74	0.98	0.84	0.66	0.82	1.0	0.88	0.92	0.84	0.9796
dot_precision@1	0.28	0.8	0.84	0.48	0.84	0.44	0.42	0.58	0.9	0.42	0.38	0.62	0.4898
dot_precision@3	0.1867	0.6467	0.32	0.3067	0.5133	0.2067	0.38	0.24	0.4133	0.3533	0.2333	0.2667	0.517
dot_precision@5	0.168	0.56	0.2	0.224	0.328	0.148	0.348	0.168	0.272	0.3	0.16	0.172	0.5347
dot_precision@10	0.108	0.474	0.1	0.136	0.17	0.084	0.258	0.09	0.138	0.208	0.092	0.096	0.4347
dot_recall@1	0.1217	0.0913	0.7867	0.2592	0.42	0.44	0.0449	0.55	0.7773	0.0907	0.38	0.595	0.0342
dot_recall@3	0.2323	0.1741	0.8867	0.3973	0.77	0.62	0.0877	0.66	0.962	0.2217	0.7	0.705	0.109
dot_recall@5	0.348	0.2252	0.9267	0.4498	0.82	0.74	0.1084	0.75	0.9933	0.3097	0.8	0.755	0.1812
dot_recall@10	0.4263	0.3214	0.9267	0.5796	0.85	0.84	0.1355	0.79	0.9967	0.4257	0.92	0.84	0.2885
dot_ndcg@10	0.3324	0.6002	0.8816	0.4881	0.8107	0.6329	0.3285	0.6773	0.951	0.4023	0.6551	0.7194	0.4768
dot_mrr@10	0.4364	0.8425	0.89	0.5517	0.8967	0.5678	0.501	0.6522	0.9467	0.5887	0.5706	0.6824	0.6785
dot_map@100	0.249	0.4526	0.859	0.4255	0.7566	0.5762	0.1617	0.6421	0.9297	0.3208	0.5761	0.6851	0.359
query_active_dims	256.0	256.0	256.0	256.0	256.0	256.0	256.0	256.0	256.0	256.0	256.0	256.0	256.0
query_sparsity_ratio	0.9375	0.9375	0.9375	0.9375	0.9375	0.9375	0.9375	0.9375	0.9375	0.9375	0.9375	0.9375	0.9375
corpus_active_dims	256.0	256.0	256.0	256.0	256.0	256.0	256.0	256.0	256.0	256.0	256.0	256.0	256.0
corpus_sparsity_ratio	0.9375	0.9375	0.9375	0.9375	0.9375	0.9375	0.9375	0.9375	0.9375	0.9375	0.9375	0.9375	0.9375

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	0.5761
dot_accuracy@3	0.7505
dot_accuracy@5	0.8215
dot_accuracy@10	0.8723
dot_precision@1	0.5761
dot_precision@3	0.3526
dot_precision@5	0.2756
dot_precision@10	0.1837
dot_recall@1	0.3531
dot_recall@3	0.502
dot_recall@5	0.5698
dot_recall@10	0.6416
dot_ndcg@10	0.612
dot_mrr@10	0.6773
dot_map@100	0.538
query_active_dims	256.0
query_sparsity_ratio	0.9375
corpus_active_dims	256.0
corpus_sparsity_ratio	0.9375

Training Details

Training Dataset

natural-questions

Dataset: natural-questions at f9e894e
Size: 99,000 training samples
Columns: query and answer
Approximate statistics based on the first 1000 samples:
query answer
type string string
details
min: 10 tokens
mean: 11.71 tokens
max: 26 tokens

min: 4 tokens
mean: 131.81 tokens
max: 450 tokens

	query	answer
type	string	string
details	min: 10 tokens mean: 11.71 tokens max: 26 tokens	min: 4 tokens mean: 131.81 tokens max: 450 tokens

Samples:

query	answer
`who played the father in papa don't preach`	`Alex McArthur Alex McArthur (born March 6, 1957) is an American actor.`
`where was the location of the battle of hastings`	`Battle of Hastings The Battle of Hastings[a] was fought on 14 October 1066 between the Norman-French army of William, the Duke of Normandy, and an English army under the Anglo-Saxon King Harold Godwinson, beginning the Norman conquest of England. It took place approximately 7 miles (11 kilometres) northwest of Hastings, close to the present-day town of Battle, East Sussex, and was a decisive Norman victory.`
`how many puppies can a dog give birth to`	`Canine reproduction The largest litter size to date was set by a Neapolitan Mastiff in Manea, Cambridgeshire, UK on November 29, 2004; the litter was 24 puppies.[22]`

Loss: CSRLoss with these parameters:

{
    "beta": 0.1,
    "gamma": 1.0,
    "loss": "SparseMultipleNegativesRankingLoss(scale=1.0, similarity_fct='dot_score')"
}

Evaluation Dataset

natural-questions

Dataset: natural-questions at f9e894e
Size: 1,000 evaluation samples
Columns: query and answer
Approximate statistics based on the first 1000 samples:
query answer
type string string
details
min: 10 tokens
mean: 11.69 tokens
max: 23 tokens

min: 15 tokens
mean: 134.01 tokens
max: 512 tokens

	query	answer
type	string	string
details	min: 10 tokens mean: 11.69 tokens max: 23 tokens	min: 15 tokens mean: 134.01 tokens max: 512 tokens

Samples:

query	answer
`where is the tiber river located in italy`	Tiber The Tiber (/ˈtaɪbər/, Latin: Tiberis,[1] Italian: Tevere [ˈteːvere])[2] is the third-longest river in Italy, rising in the Apennine Mountains in Emilia-Romagna and flowing 406 kilometres (252 mi) through Tuscany, Umbria and Lazio, where it is joined by the river Aniene, to the Tyrrhenian Sea, between Ostia and Fiumicino.[3] It drains a basin estimated at 17,375 square kilometres (6,709 sq mi). The river has achieved lasting fame as the main watercourse of the city of Rome, founded on its eastern banks.
`what kind of car does jay gatsby drive`	`Jay Gatsby At the Buchanan home, Jordan Baker, Nick, Jay, and the Buchanans decide to visit New York City. Tom borrows Gatsby's yellow Rolls Royce to drive up to the city. On the way to New York City, Tom makes a detour at a gas station in "the Valley of Ashes", a run-down part of Long Island. The owner, George Wilson, shares his concern that his wife, Myrtle, may be having an affair. This unnerves Tom, who has been having an affair with Myrtle, and he leaves in a hurry.`
`who sings if i can dream about you`	`I Can Dream About You "I Can Dream About You" is a song performed by American singer Dan Hartman on the soundtrack album of the film Streets of Fire. Released in 1984 as a single from the soundtrack, and included on Hartman's album I Can Dream About You, it reached number 6 on the Billboard Hot 100.[1]`

Loss: CSRLoss with these parameters:

{
    "beta": 0.1,
    "gamma": 1.0,
    "loss": "SparseMultipleNegativesRankingLoss(scale=1.0, similarity_fct='dot_score')"
}

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: steps
per_device_train_batch_size: 64
per_device_eval_batch_size: 64
learning_rate: 4e-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: 64
per_device_eval_batch_size: 64
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: 4e-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_128_dot_ndcg@10	NanoNFCorpus_128_dot_ndcg@10	NanoNQ_128_dot_ndcg@10	NanoBEIR_mean_128_dot_ndcg@10	NanoMSMARCO_256_dot_ndcg@10	NanoNFCorpus_256_dot_ndcg@10	NanoNQ_256_dot_ndcg@10	NanoBEIR_mean_256_dot_ndcg@10	NanoClimateFEVER_dot_ndcg@10	NanoDBPedia_dot_ndcg@10	NanoFEVER_dot_ndcg@10	NanoFiQA2018_dot_ndcg@10	NanoHotpotQA_dot_ndcg@10	NanoMSMARCO_dot_ndcg@10	NanoNFCorpus_dot_ndcg@10	NanoNQ_dot_ndcg@10	NanoQuoraRetrieval_dot_ndcg@10	NanoSCIDOCS_dot_ndcg@10	NanoArguAna_dot_ndcg@10	NanoSciFact_dot_ndcg@10	NanoTouche2020_dot_ndcg@10	NanoBEIR_mean_dot_ndcg@10
0.0646	100	0.3565	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.1293	200	0.3568	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.1939	300	0.3545	0.3458	0.6322	0.2796	0.5893	0.5004	0.6232	0.3253	0.6548	0.5345	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.2586	400	0.3393	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.3232	500	0.3484	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.3878	600	0.3567	0.3452	0.6245	0.3038	0.5719	0.5000	0.6385	0.3375	0.6496	0.5419	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.4525	700	0.3471	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.5171	800	0.3582	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.5818	900	0.3758	0.3417	0.5849	0.3074	0.5866	0.4929	0.6147	0.3310	0.6729	0.5395	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.6464	1000	0.3515	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.7111	1100	0.3287	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.7757	1200	0.3486	0.3314	0.5937	0.2998	0.6317	0.5084	0.6309	0.3303	0.6773	0.5462	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.8403	1300	0.3527	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.9050	1400	0.3161	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.9696	1500	0.3279	0.3244	0.6075	0.3038	0.6338	0.5150	0.6405	0.3181	0.6642	0.5410	-	-	-	-	-	-	-	-	-	-	-	-	-	-
-1	-1	-	-	-	-	-	-	-	-	-	-	0.3324	0.6002	0.8816	0.4881	0.8107	0.6329	0.3285	0.6773	0.9510	0.4023	0.6551	0.7194	0.4768	0.6120

The bold row denotes the saved checkpoint.

Environmental Impact

Carbon emissions were measured using CodeCarbon.

Energy Consumed: 0.136 kWh
Carbon Emitted: 0.053 kg of CO2
Hours Used: 0.41 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: 2.6.0+cu124
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",
}

CSRLoss

@misc{wen2025matryoshkarevisitingsparsecoding,
      title={Beyond Matryoshka: Revisiting Sparse Coding for Adaptive Representation},
      author={Tiansheng Wen and Yifei Wang and Zequn Zeng and Zhong Peng and Yudi Su and Xinyang Liu and Bo Chen and Hongwei Liu and Stefanie Jegelka and Chenyu You},
      year={2025},
      eprint={2503.01776},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2503.01776},
}

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}
}