%20--%3e%3cdefs%3e%3cstyle%3e%20.st0%20{%20fill:%20%23061b40;%20}%20.st1%20{%20fill:%20%23306af1;%20}%20.st2%20{%20fill:%20%235ce5cf;%20}%20%3c/style%3e%3c/defs%3e%3cg%3e%3cpath%20class='st0'%20d='M55,10.5h9v3h-9v9h12V7.5h-12v3ZM64,19.5h-6v-3h6v3Z'/%3e%3cpolygon%20class='st0'%20points='69%2016.5%2078%2016.5%2078%2019.5%2069%2019.5%2069%2022.5%2081%2022.5%2081%2013.5%2072%2013.5%2072%2010.5%2081%2010.5%2081%207.5%2069%207.5%2069%2016.5'/%3e%3cpolygon%20class='st0'%20points='95%2010.5%2095%207.5%2083%207.5%2083%2022.5%2095%2022.5%2095%2019.5%2086%2019.5%2086%2016.5%2095%2016.5%2095%2013.5%2086%2013.5%2086%2010.5%2095%2010.5'/%3e%3cpath%20class='st0'%20d='M40,1.5v21h11.6l1.4-1.4v-7.6h0c0,0-1.4-1.5-1.4-1.5l1.4-1.4V2.9l-1.4-1.4h-11.6ZM50,19.5h-7v-6h7v6ZM50,10.5h-7v-6h7v6Z'/%3e%3c/g%3e%3cpath%20class='st1'%20d='M23.1,24L14.7,4.8l-4.9,11.2h3.8l-1.8,4H3.3L12.1,0H2C.9,0,0,.9,0,2v20c0,1.1.9,2,2,2h21.1Z'/%3e%3cpath%20class='st2'%20d='M34,0h-16.8l10.6,24h6.2c1.1,0,2-.9,2-2V2C36,.9,35.1,0,34,0ZM32.5,20h-4V4h4v16Z'/%3e%3c/svg%3e)
JA
St Polish Kartonberta Base Alpha V1
S
St Polish Kartonberta Base Alpha V1
OrlikBによって開発
これはKartonBERTaアーキテクチャに基づくポーランド語の文変換モデルで、主に文の類似度計算と特徴抽出タスクに使用されます。
ダウンロード数 3,494
リリース時間 : 11/12/2023
モデル概要
このモデルはポーランド語テキスト処理に特化しており、文を高次元ベクトル表現に変換し、文の類似度計算、クラスタリング、分類などのタスクに使用できます。
モデル特徴
ポーランド語最適化
ポーランド語テキストに特化して最適化されており、優れた言語理解能力を提供します
文類似度計算
ポーランド語の文間の意味的類似度を正確に計算できます
マルチタスクサポート
クラスタリング、分類、検索など様々な自然言語処理タスクをサポートします
モデル能力
文特徴抽出
文類似度計算
テキストクラスタリング
テキスト分類
情報検索
使用事例
テキスト分析
レビュー分類
ポーランド語の製品レビューを感情分類
AllegroReviewsデータセットで40.19%の精度を達成
意味的類似度判定
2つのポーランド語の文が同じ意味を表しているか判断
CDSC-Eデータセットで89.2%の精度を達成
情報検索
ドキュメント検索
クエリに基づいて関連するポーランド語ドキュメントを検索
DBPedia-PLデータセットで54.54%のMRR@100を達成
tags:
- sentence-transformers
- feature-extraction
- sentence-similarity
- transformers
- mteb license: lgpl language:
- pl pipeline_tag: sentence-similarity model-index:
- name: st-polish-kartonberta-base-alpha-v1
results:
- task:
type: Clustering
dataset:
type: PL-MTEB/8tags-clustering
name: MTEB 8TagsClustering
config: default
split: test
revision: None
metrics:
- type: v_measure value: 32.85180358455615
- task:
type: Classification
dataset:
type: PL-MTEB/allegro-reviews
name: MTEB AllegroReviews
config: default
split: test
revision: None
metrics:
- type: accuracy value: 40.188866799204774
- type: f1 value: 34.71127012684797
- task:
type: Retrieval
dataset:
type: arguana-pl
name: MTEB ArguAna-PL
config: default
split: test
revision: None
metrics:
- type: map_at_1 value: 30.939
- type: map_at_10 value: 47.467999999999996
- type: map_at_100 value: 48.303000000000004
- type: map_at_1000 value: 48.308
- type: map_at_3 value: 43.22
- type: map_at_5 value: 45.616
- type: mrr_at_1 value: 31.863000000000003
- type: mrr_at_10 value: 47.829
- type: mrr_at_100 value: 48.664
- type: mrr_at_1000 value: 48.67
- type: mrr_at_3 value: 43.492
- type: mrr_at_5 value: 46.006
- type: ndcg_at_1 value: 30.939
- type: ndcg_at_10 value: 56.058
- type: ndcg_at_100 value: 59.562000000000005
- type: ndcg_at_1000 value: 59.69799999999999
- type: ndcg_at_3 value: 47.260000000000005
- type: ndcg_at_5 value: 51.587
- type: precision_at_1 value: 30.939
- type: precision_at_10 value: 8.329
- type: precision_at_100 value: 0.984
- type: precision_at_1000 value: 0.1
- type: precision_at_3 value: 19.654
- type: precision_at_5 value: 13.898
- type: recall_at_1 value: 30.939
- type: recall_at_10 value: 83.286
- type: recall_at_100 value: 98.43499999999999
- type: recall_at_1000 value: 99.502
- type: recall_at_3 value: 58.962
- type: recall_at_5 value: 69.488
- task:
type: Classification
dataset:
type: PL-MTEB/cbd
name: MTEB CBD
config: default
split: test
revision: None
metrics:
- type: accuracy value: 67.69000000000001
- type: ap value: 21.078799692467182
- type: f1 value: 56.80107173953953
- task:
type: PairClassification
dataset:
type: PL-MTEB/cdsce-pairclassification
name: MTEB CDSC-E
config: default
split: test
revision: None
metrics:
- type: cos_sim_accuracy value: 89.2
- type: cos_sim_ap value: 79.11674608786898
- type: cos_sim_f1 value: 68.83468834688347
- type: cos_sim_precision value: 70.94972067039106
- type: cos_sim_recall value: 66.84210526315789
- type: dot_accuracy value: 89.2
- type: dot_ap value: 79.11674608786898
- type: dot_f1 value: 68.83468834688347
- type: dot_precision value: 70.94972067039106
- type: dot_recall value: 66.84210526315789
- type: euclidean_accuracy value: 89.2
- type: euclidean_ap value: 79.11674608786898
- type: euclidean_f1 value: 68.83468834688347
- type: euclidean_precision value: 70.94972067039106
- type: euclidean_recall value: 66.84210526315789
- type: manhattan_accuracy value: 89.1
- type: manhattan_ap value: 79.1220443374692
- type: manhattan_f1 value: 69.02173913043478
- type: manhattan_precision value: 71.34831460674157
- type: manhattan_recall value: 66.84210526315789
- type: max_accuracy value: 89.2
- type: max_ap value: 79.1220443374692
- type: max_f1 value: 69.02173913043478
- task:
type: STS
dataset:
type: PL-MTEB/cdscr-sts
name: MTEB CDSC-R
config: default
split: test
revision: None
metrics:
- type: cos_sim_pearson value: 91.41534744278998
- type: cos_sim_spearman value: 92.12681551821147
- type: euclidean_pearson value: 91.74369794485992
- type: euclidean_spearman value: 92.12685848456046
- type: manhattan_pearson value: 91.66651938751657
- type: manhattan_spearman value: 92.057603126734
- task:
type: Retrieval
dataset:
type: dbpedia-pl
name: MTEB DBPedia-PL
config: default
split: test
revision: None
metrics:
- type: map_at_1 value: 5.8709999999999996
- type: map_at_10 value: 12.486
- type: map_at_100 value: 16.897000000000002
- type: map_at_1000 value: 18.056
- type: map_at_3 value: 8.958
- type: map_at_5 value: 10.57
- type: mrr_at_1 value: 44.0
- type: mrr_at_10 value: 53.830999999999996
- type: mrr_at_100 value: 54.54
- type: mrr_at_1000 value: 54.568000000000005
- type: mrr_at_3 value: 51.87500000000001
- type: mrr_at_5 value: 53.113
- type: ndcg_at_1 value: 34.625
- type: ndcg_at_10 value: 26.996
- type: ndcg_at_100 value: 31.052999999999997
- type: ndcg_at_1000 value: 38.208
- type: ndcg_at_3 value: 29.471000000000004
- type: ndcg_at_5 value: 28.364
- type: precision_at_1 value: 44.0
- type: precision_at_10 value: 21.45
- type: precision_at_100 value: 6.837
- type: precision_at_1000 value: 1.6019999999999999
- type: precision_at_3 value: 32.333
- type: precision_at_5 value: 27.800000000000004
- type: recall_at_1 value: 5.8709999999999996
- type: recall_at_10 value: 17.318
- type: recall_at_100 value: 36.854
- type: recall_at_1000 value: 60.468999999999994
- type: recall_at_3 value: 10.213999999999999
- type: recall_at_5 value: 13.364
- task:
type: Retrieval
dataset:
type: fiqa-pl
name: MTEB FiQA-PL
config: default
split: test
revision: None
metrics:
- type: map_at_1 value: 10.289
- type: map_at_10 value: 18.285999999999998
- type: map_at_100 value: 19.743
- type: map_at_1000 value: 19.964000000000002
- type: map_at_3 value: 15.193000000000001
- type: map_at_5 value: 16.962
- type: mrr_at_1 value: 21.914
- type: mrr_at_10 value: 30.653999999999996
- type: mrr_at_100 value: 31.623
- type: mrr_at_1000 value: 31.701
- type: mrr_at_3 value: 27.855
- type: mrr_at_5 value: 29.514000000000003
- type: ndcg_at_1 value: 21.914
- type: ndcg_at_10 value: 24.733
- type: ndcg_at_100 value: 31.253999999999998
- type: ndcg_at_1000 value: 35.617
- type: ndcg_at_3 value: 20.962
- type: ndcg_at_5 value: 22.553
- type: precision_at_1 value: 21.914
- type: precision_at_10 value: 7.346
- type: precision_at_100 value: 1.389
- type: precision_at_1000 value: 0.214
- type: precision_at_3 value: 14.352
- type: precision_at_5 value: 11.42
- type: recall_at_1 value: 10.289
- type: recall_at_10 value: 31.459
- type: recall_at_100 value: 56.854000000000006
- type: recall_at_1000 value: 83.722
- type: recall_at_3 value: 19.457
- type: recall_at_5 value: 24.767
- task:
type: Retrieval
dataset:
type: hotpotqa-pl
name: MTEB HotpotQA-PL
config: default
split: test
revision: None
metrics:
- type: map_at_1 value: 29.669
- type: map_at_10 value: 41.615
- type: map_at_100 value: 42.571999999999996
- type: map_at_1000 value: 42.662
- type: map_at_3 value: 38.938
- type: map_at_5 value: 40.541
- type: mrr_at_1 value: 59.338
- type: mrr_at_10 value: 66.93900000000001
- type: mrr_at_100 value: 67.361
- type: mrr_at_1000 value: 67.38499999999999
- type: mrr_at_3 value: 65.384
- type: mrr_at_5 value: 66.345
- type: ndcg_at_1 value: 59.338
- type: ndcg_at_10 value: 50.607
- type: ndcg_at_100 value: 54.342999999999996
- type: ndcg_at_1000 value: 56.286
- type: ndcg_at_3 value: 46.289
- type: ndcg_at_5 value: 48.581
- type: precision_at_1 value: 59.338
- type: precision_at_10 value: 10.585
- type: precision_at_100 value: 1.353
- type: precision_at_1000 value: 0.161
- type: precision_at_3 value: 28.877000000000002
- type: precision_at_5 value: 19.133
- type: recall_at_1 value: 29.669
- type: recall_at_10 value: 52.92400000000001
- type: recall_at_100 value: 67.657
- type: recall_at_1000 value: 80.628
- type: recall_at_3 value: 43.315
- type: recall_at_5 value: 47.833
- task:
type: Retrieval
dataset:
type: msmarco-pl
name: MTEB MSMARCO-PL
config: default
split: test
revision: None
metrics:
- type: map_at_1 value: 0.997
- type: map_at_10 value: 7.481999999999999
- type: map_at_100 value: 20.208000000000002
- type: map_at_1000 value: 25.601000000000003
- type: map_at_3 value: 3.055
- type: map_at_5 value: 4.853
- type: mrr_at_1 value: 55.814
- type: mrr_at_10 value: 64.651
- type: mrr_at_100 value: 65.003
- type: mrr_at_1000 value: 65.05199999999999
- type: mrr_at_3 value: 62.403
- type: mrr_at_5 value: 64.031
- type: ndcg_at_1 value: 44.186
- type: ndcg_at_10 value: 43.25
- type: ndcg_at_100 value: 40.515
- type: ndcg_at_1000 value: 48.345
- type: ndcg_at_3 value: 45.829
- type: ndcg_at_5 value: 46.477000000000004
- type: precision_at_1 value: 55.814
- type: precision_at_10 value: 50.465
- type: precision_at_100 value: 25.419000000000004
- type: precision_at_1000 value: 5.0840000000000005
- type: precision_at_3 value: 58.14
- type: precision_at_5 value: 57.67400000000001
- type: recall_at_1 value: 0.997
- type: recall_at_10 value: 8.985999999999999
- type: recall_at_100 value: 33.221000000000004
- type: recall_at_1000 value: 58.836999999999996
- type: recall_at_3 value: 3.472
- type: recall_at_5 value: 5.545
- task:
type: Classification
dataset:
type: mteb/amazon_massive_intent
name: MTEB MassiveIntentClassification (pl)
config: pl
split: test
revision: 31efe3c427b0bae9c22cbb560b8f15491cc6bed7
metrics:
- type: accuracy value: 68.19771351714861
- type: f1 value: 64.75039989217822
- task:
type: Classification
dataset:
type: mteb/amazon_massive_scenario
name: MTEB MassiveScenarioClassification (pl)
config: pl
split: test
revision: 7d571f92784cd94a019292a1f45445077d0ef634
metrics:
- type: accuracy value: 73.9677202420982
- type: f1 value: 73.72287107577753
- task:
type: Retrieval
dataset:
type: nfcorpus-pl
name: MTEB NFCorpus-PL
config: default
split: test
revision: None
metrics:
- type: map_at_1 value: 5.167
- type: map_at_10 value: 10.791
- type: map_at_100 value: 14.072999999999999
- type: map_at_1000 value: 15.568000000000001
- type: map_at_3 value: 7.847999999999999
- type: map_at_5 value: 9.112
- type: mrr_at_1 value: 42.105
- type: mrr_at_10 value: 49.933
- type: mrr_at_100 value: 50.659
- type: mrr_at_1000 value: 50.705
- type: mrr_at_3 value: 47.988
- type: mrr_at_5 value: 49.056
- type: ndcg_at_1 value: 39.938
- type: ndcg_at_10 value: 31.147000000000002
- type: ndcg_at_100 value: 29.336000000000002
- type: ndcg_at_1000 value: 38.147
- type: ndcg_at_3 value: 35.607
- type: ndcg_at_5 value: 33.725
- type: precision_at_1 value: 41.486000000000004
- type: precision_at_10 value: 23.901
- type: precision_at_100 value: 7.960000000000001
- type: precision_at_1000 value: 2.086
- type: precision_at_3 value: 33.437
- type: precision_at_5 value: 29.598000000000003
- type: recall_at_1 value: 5.167
- type: recall_at_10 value: 14.244000000000002
- type: recall_at_100 value: 31.192999999999998
- type: recall_at_1000 value: 62.41799999999999
- type: recall_at_3 value: 8.697000000000001
- type: recall_at_5 value: 10.911
- task:
type: Retrieval
dataset:
type: nq-pl
name: MTEB NQ-PL
config: default
split: test
revision: None
metrics:
- type: map_at_1 value: 14.417
- type: map_at_10 value: 23.330000000000002
- type: map_at_100 value: 24.521
- type: map_at_1000 value: 24.604
- type: map_at_3 value: 20.076
- type: map_at_5 value: 21.854000000000003
- type: mrr_at_1 value: 16.454
- type: mrr_at_10 value: 25.402
- type: mrr_at_100 value: 26.411
- type: mrr_at_1000 value: 26.479000000000003
- type: mrr_at_3 value: 22.369
- type: mrr_at_5 value: 24.047
- type: ndcg_at_1 value: 16.454
- type: ndcg_at_10 value: 28.886
- type: ndcg_at_100 value: 34.489999999999995
- type: ndcg_at_1000 value: 36.687999999999995
- type: ndcg_at_3 value: 22.421
- type: ndcg_at_5 value: 25.505
- type: precision_at_1 value: 16.454
- type: precision_at_10 value: 5.252
- type: precision_at_100 value: 0.8410000000000001
- type: precision_at_1000 value: 0.105
- type: precision_at_3 value: 10.428999999999998
- type: precision_at_5 value: 8.019
- type: recall_at_1 value: 14.417
- type: recall_at_10 value: 44.025
- type: recall_at_100 value: 69.404
- type: recall_at_1000 value: 86.18900000000001
- type: recall_at_3 value: 26.972
- type: recall_at_5 value: 34.132
- task:
type: Classification
dataset:
type: laugustyniak/abusive-clauses-pl
name: MTEB PAC
config: default
split: test
revision: None
metrics:
- type: accuracy value: 66.55082536924412
- type: ap value: 76.44962281293184
- type: f1 value: 63.899803692180434
- task:
type: PairClassification
dataset:
type: PL-MTEB/ppc-pairclassification
name: MTEB PPC
config: default
split: test
revision: None
metrics:
- type: cos_sim_accuracy value: 86.5
- type: cos_sim_ap value: 92.65086645409387
- type: cos_sim_f1 value: 89.39157566302653
- type: cos_sim_precision value: 84.51327433628319
- type: cos_sim_recall value: 94.86754966887418
- type: dot_accuracy value: 86.5
- type: dot_ap value: 92.65086645409387
- type: dot_f1 value: 89.39157566302653
- type: dot_precision value: 84.51327433628319
- type: dot_recall value: 94.86754966887418
- type: euclidean_accuracy value: 86.5
- type: euclidean_ap value: 92.65086645409387
- type: euclidean_f1 value: 89.39157566302653
- type: euclidean_precision value: 84.51327433628319
- type: euclidean_recall value: 94.86754966887418
- type: manhattan_accuracy value: 86.5
- type: manhattan_ap value: 92.64975544736456
- type: manhattan_f1 value: 89.33852140077822
- type: manhattan_precision value: 84.28781204111601
- type: manhattan_recall value: 95.03311258278146
- type: max_accuracy value: 86.5
- type: max_ap value: 92.65086645409387
- type: max_f1 value: 89.39157566302653
- task:
type: PairClassification
dataset:
type: PL-MTEB/psc-pairclassification
name: MTEB PSC
config: default
split: test
revision: None
metrics:
- type: cos_sim_accuracy value: 95.64007421150278
- type: cos_sim_ap value: 98.42114841894346
- type: cos_sim_f1 value: 92.8895612708018
- type: cos_sim_precision value: 92.1921921921922
- type: cos_sim_recall value: 93.59756097560977
- type: dot_accuracy value: 95.64007421150278
- type: dot_ap value: 98.42114841894346
- type: dot_f1 value: 92.8895612708018
- type: dot_precision value: 92.1921921921922
- type: dot_recall value: 93.59756097560977
- type: euclidean_accuracy value: 95.64007421150278
- type: euclidean_ap value: 98.42114841894346
- type: euclidean_f1 value: 92.8895612708018
- type: euclidean_precision value: 92.1921921921922
- type: euclidean_recall value: 93.59756097560977
- type: manhattan_accuracy value: 95.82560296846012
- type: manhattan_ap value: 98.38712415914046
- type: manhattan_f1 value: 93.19213313161876
- type: manhattan_precision value: 92.49249249249249
- type: manhattan_recall value: 93.90243902439023
- type: max_accuracy value: 95.82560296846012
- type: max_ap value: 98.42114841894346
- type: max_f1 value: 93.19213313161876
- task:
type: Classification
dataset:
type: PL-MTEB/polemo2_in
name: MTEB PolEmo2.0-IN
config: default
split: test
revision: None
metrics:
- type: accuracy value: 68.40720221606648
- type: f1 value: 67.09084289613526
- task:
type: Classification
dataset:
type: PL-MTEB/polemo2_out
name: MTEB PolEmo2.0-OUT
config: default
split: test
revision: None
metrics:
- type: accuracy value: 38.056680161943326
- type: f1 value: 32.87731504372395
- task:
type: Retrieval
dataset:
type: quora-pl
name: MTEB Quora-PL
config: default
split: test
revision: None
metrics:
- type: map_at_1 value: 65.422
- type: map_at_10 value: 79.259
- type: map_at_100 value: 80.0
- type: map_at_1000 value: 80.021
- type: map_at_3 value: 76.16199999999999
- type: map_at_5 value: 78.03999999999999
- type: mrr_at_1 value: 75.26
- type: mrr_at_10 value: 82.39699999999999
- type: mrr_at_100 value: 82.589
- type: mrr_at_1000 value: 82.593
- type: mrr_at_3 value: 81.08999999999999
- type: mrr_at_5 value: 81.952
- type: ndcg_at_1 value: 75.3
- type: ndcg_at_10 value: 83.588
- type: ndcg_at_100 value: 85.312
- type: ndcg_at_1000 value: 85.536
- type: ndcg_at_3 value: 80.128
- type: ndcg_at_5 value: 81.962
- type: precision_at_1 value: 75.3
- type: precision_at_10 value: 12.856000000000002
- type: precision_at_100 value: 1.508
- type: precision_at_1000 value: 0.156
- type: precision_at_3 value: 35.207
- type: precision_at_5 value: 23.316
- type: recall_at_1 value: 65.422
- type: recall_at_10 value: 92.381
- type: recall_at_100 value: 98.575
- type: recall_at_1000 value: 99.85300000000001
- type: recall_at_3 value: 82.59100000000001
- type: recall_at_5 value: 87.629
- task:
type: Retrieval
dataset:
type: scidocs-pl
name: MTEB SCIDOCS-PL
config: default
split: test
revision: None
metrics:
- type: map_at_1 value: 2.52
- type: map_at_10 value: 6.814000000000001
- type: map_at_100 value: 8.267
- type: map_at_1000 value: 8.565000000000001
- type: map_at_3 value: 4.736
- type: map_at_5 value: 5.653
- type: mrr_at_1 value: 12.5
- type: mrr_at_10 value: 20.794999999999998
- type: mrr_at_100 value: 22.014
- type: mrr_at_1000 value: 22.109
- type: mrr_at_3 value: 17.8
- type: mrr_at_5 value: 19.42
- type: ndcg_at_1 value: 12.5
- type: ndcg_at_10 value: 12.209
- type: ndcg_at_100 value: 18.812
- type: ndcg_at_1000 value: 24.766
- type: ndcg_at_3 value: 10.847
- type: ndcg_at_5 value: 9.632
- type: precision_at_1 value: 12.5
- type: precision_at_10 value: 6.660000000000001
- type: precision_at_100 value: 1.6340000000000001
- type: precision_at_1000 value: 0.307
- type: precision_at_3 value: 10.299999999999999
- type: precision_at_5 value: 8.66
- type: recall_at_1 value: 2.52
- type: recall_at_10 value: 13.495
- type: recall_at_100 value: 33.188
- type: recall_at_1000 value: 62.34499999999999
- type: recall_at_3 value: 6.245
- type: recall_at_5 value: 8.76
- task:
type: PairClassification
dataset:
type: PL-MTEB/sicke-pl-pairclassification
name: MTEB SICK-E-PL
config: default
split: test
revision: None
metrics:
- type: cos_sim_accuracy value: 86.13942111699959
- type: cos_sim_ap value: 81.47480017120256
- type: cos_sim_f1 value: 74.79794268919912
- type: cos_sim_precision value: 77.2382397572079
- type: cos_sim_recall value: 72.50712250712252
- type: dot_accuracy value: 86.13942111699959
- type: dot_ap value: 81.47478531367476
- type: dot_f1 value: 74.79794268919912
- type: dot_precision value: 77.2382397572079
- type: dot_recall value: 72.50712250712252
- type: euclidean_accuracy value: 86.13942111699959
- type: euclidean_ap value: 81.47478531367476
- type: euclidean_f1 value: 74.79794268919912
- type: euclidean_precision value: 77.2382397572079
- type: euclidean_recall value: 72.50712250712252
- type: manhattan_accuracy value: 86.15980432123929
- type: manhattan_ap value: 81.40798042612397
- type: manhattan_f1 value: 74.86116253239543
- type: manhattan_precision value: 77.9491133384734
- type: manhattan_recall value: 72.00854700854701
- type: max_accuracy value: 86.15980432123929
- type: max_ap value: 81.47480017120256
- type: max_f1 value: 74.86116253239543
- task:
type: STS
dataset:
type: PL-MTEB/sickr-pl-sts
name: MTEB SICK-R-PL
config: default
split: test
revision: None
metrics:
- type: cos_sim_pearson value: 84.27525342551935
- type: cos_sim_spearman value: 79.50631730805885
- type: euclidean_pearson value: 82.07169123942028
- type: euclidean_spearman value: 79.50631887406465
- type: manhattan_pearson value: 81.98288826317463
- type: manhattan_spearman value: 79.4244081650332
- task:
type: STS
dataset:
type: mteb/sts22-crosslingual-sts
name: MTEB STS22 (pl)
config: pl
split: test
revision: 6d1ba47164174a496b7fa5d3569dae26a6813b80
metrics:
- type: cos_sim_pearson value: 35.59400236598834
- type: cos_sim_spearman value: 36.782560207852846
- type: euclidean_pearson value: 28.546177668542942
- type: euclidean_spearman value: 36.68394223635756
- type: manhattan_pearson value: 28.45606963909248
- type: manhattan_spearman value: 36.475975118547524
- task:
type: Retrieval
dataset:
type: scifact-pl
name: MTEB SciFact-PL
config: default
split: test
revision: None
metrics:
- type: map_at_1 value: 41.028
- type: map_at_10 value: 52.23799999999999
- type: map_at_100 value: 52.905
- type: map_at_1000 value: 52.945
- type: map_at_3 value: 49.102000000000004
- type: map_at_5 value: 50.992000000000004
- type: mrr_at_1 value: 43.333
- type: mrr_at_10 value: 53.551
- type: mrr_at_100 value: 54.138
- type: mrr_at_1000 value: 54.175
- type: mrr_at_3 value: 51.056000000000004
- type: mrr_at_5 value: 52.705999999999996
- type: ndcg_at_1 value: 43.333
- type: ndcg_at_10 value: 57.731
- type: ndcg_at_100 value: 61.18599999999999
- type: ndcg_at_1000 value: 62.261
- type: ndcg_at_3 value: 52.276999999999994
- type: ndcg_at_5 value: 55.245999999999995
- type: precision_at_1 value: 43.333
- type: precision_at_10 value: 8.267
- type: precision_at_100 value: 1.02
- type: precision_at_1000 value: 0.11100000000000002
- type: precision_at_3 value: 21.444
- type: precision_at_5 value: 14.533
- type: recall_at_1 value: 41.028
- type: recall_at_10 value: 73.111
- type: recall_at_100 value: 89.533
- type: recall_at_1000 value: 98.0
- type: recall_at_3 value: 58.744
- type: recall_at_5 value: 66.106
- task:
type: Retrieval
dataset:
type: trec-covid-pl
name: MTEB TRECCOVID-PL
config: default
split: test
revision: None
metrics:
- type: map_at_1 value: 0.146
- type: map_at_10 value: 1.09
- type: map_at_100 value: 6.002
- type: map_at_1000 value: 15.479999999999999
- type: map_at_3 value: 0.41000000000000003
- type: map_at_5 value: 0.596
- type: mrr_at_1 value: 54.0
- type: mrr_at_10 value: 72.367
- type: mrr_at_100 value: 72.367
- type: mrr_at_1000 value: 72.367
- type: mrr_at_3 value: 70.333
- type: mrr_at_5 value: 72.033
- type: ndcg_at_1 value: 48.0
- type: ndcg_at_10 value: 48.827
- type: ndcg_at_100 value: 38.513999999999996
- type: ndcg_at_1000 value: 37.958
- type: ndcg_at_3 value: 52.614000000000004
- type: ndcg_at_5 value: 51.013
- type: precision_at_1 value: 54.0
- type: precision_at_10 value: 53.6
- type: precision_at_100 value: 40.300000000000004
- type: precision_at_1000 value: 17.276
- type: precision_at_3 value: 57.333
- type: precision_at_5 value: 55.60000000000001
- type: recall_at_1 value: 0.146
- type: recall_at_10 value: 1.438
- type: recall_at_100 value: 9.673
- type: recall_at_1000 value: 36.870999999999995
- type: recall_at_3 value: 0.47400000000000003
- type: recall_at_5 value: 0.721
- task:
type: Clustering
dataset:
type: PL-MTEB/8tags-clustering
name: MTEB 8TagsClustering
config: default
split: test
revision: None
metrics:
Model Card for st-polish-kartonberta-base-alpha-v1
This sentence transformer model is designed to convert text content into a 768-float vector space, ensuring an effective representation. It aims to be proficient in tasks involving sentence / document similarity.
The model has been released in its alpha version. Numerous potential enhancements could boost its performance, such as adjusting training hyperparameters or extending the training duration (currently limited to only one epoch). The main reason is limited GPU.
Model Description
- Developed by: Bartłomiej Orlik, https://www.linkedin.com/in/bartłomiej-orlik/
- Model type: RoBERTa Sentence Transformer
- Language: Polish
- License: LGPL-3.0
- Trained from model: sdadas/polish-roberta-base-v2: https://huggingface.co/sdadas/polish-roberta-base-v2
How to Get Started with the Model
Use the code below to get started with the model.
Using Sentence-Transformers
You can use the model with sentence-transformers:
pip install -U sentence-transformers
from sentence_transformers import SentenceTransformer
model = SentenceTransformer('OrlikB/st-polish-kartonberta-base-alpha-v1')
text_1 = 'Jestem wielkim fanem opakowań tekturowych'
text_2 = 'Bardzo podobają mi się kartony'
embeddings_1 = model.encode(text_1, normalize_embeddings=True)
embeddings_2 = model.encode(text_2, normalize_embeddings=True)
similarity = embeddings_1 @ embeddings_2.T
print(similarity)
Using HuggingFace Transformers
from transformers import AutoTokenizer, AutoModel
import torch
import numpy as np
def encode_text(text):
encoded_input = tokenizer(text, padding=True, truncation=True, return_tensors='pt', max_length=512)
with torch.no_grad():
model_output = model(**encoded_input)
sentence_embeddings = model_output[0][:, 0]
sentence_embeddings = torch.nn.functional.normalize(sentence_embeddings, p=2, dim=1)
return sentence_embeddings.squeeze().numpy()
cosine_similarity = lambda a, b: np.dot(a, b) / (np.linalg.norm(a) * np.linalg.norm(b))
tokenizer = AutoTokenizer.from_pretrained('OrlikB/st-polish-kartonberta-base-alpha-v1')
model = AutoModel.from_pretrained('OrlikB/st-polish-kartonberta-base-alpha-v1')
model.eval()
text_1 = 'Jestem wielkim fanem opakowań tekturowych'
text_2 = 'Bardzo podobają mi się kartony'
embeddings_1 = encode_text(text_1)
embeddings_2 = encode_text(text_2)
print(cosine_similarity(embeddings_1, embeddings_2))
*Note: You can use the encode_text function for demonstration purposes. For the best experience, it's recommended to process text in batches.
Evaluation
MTEB for Polish Language
| Rank | Model | Model Size (GB) | Embedding Dimensions | Sequence Length | Average (26 datasets) | Classification Average (7 datasets) | Clustering Average (1 datasets) | Pair Classification Average (4 datasets) | Retrieval Average (11 datasets) | STS Average (3 datasets) |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | multilingual-e5-large | 2.24 | 1024 | 514 | 58.25 | 60.51 | 24.06 | 84.58 | 47.82 | 67.52 |
| 2 | st-polish-kartonberta-base-alpha-v1 | 0.5 | 768 | 514 | 56.92 | 60.44 | 32.85 | 87.92 | 42.19 | 69.47 |
| 3 | multilingual-e5-base | 1.11 | 768 | 514 | 54.18 | 57.01 | 18.62 | 82.08 | 42.5 | 65.07 |
| 4 | multilingual-e5-small | 0.47 | 384 | 512 | 53.15 | 54.35 | 19.64 | 81.67 | 41.52 | 66.08 |
| 5 | st-polish-paraphrase-from-mpnet | 0.5 | 768 | 514 | 53.06 | 57.49 | 25.09 | 87.04 | 36.53 | 67.39 |
| 6 | st-polish-paraphrase-from-distilroberta | 0.5 | 768 | 514 | 52.65 | 58.55 | 31.11 | 87 | 33.96 | 68.78 |
More Information
I developed this model as a personal scientific initiative.
I plan to start the development on a new ST model. However, due to limited computational resources, I suspended further work to create a larger or enhanced version of current model.
Jina Embeddings V3
Jina Embeddings V3 は100以上の言語をサポートする多言語文埋め込みモデルで、文の類似度と特徴抽出タスクに特化しています。
テキスト埋め込み
Transformers 複数言語対応
Transformers 複数言語対応J
jinaai
3.7M
911
Ms Marco MiniLM L6 V2
Apache-2.0
MS Marcoパッセージランキングタスクで訓練されたクロスエンコーダモデル、情報検索におけるクエリ-パッセージ関連性スコアリング用
テキスト埋め込み 英語
M
cross-encoder
2.5M
86
Opensearch Neural Sparse Encoding Doc V2 Distill
Apache-2.0
蒸留技術に基づくスパース検索モデルで、OpenSearch向けに最適化されており、推論不要のドキュメントエンコーディングをサポートし、検索関連性と効率性においてV1版を上回ります
テキスト埋め込み Transformers 英語
Transformers 英語O
opensearch-project
1.8M
7
Sapbert From PubMedBERT Fulltext
Apache-2.0
PubMedBERTに基づく生物医学エンティティ表現モデルで、自己アライメント事前学習により意味関係の捕捉を最適化します。
テキスト埋め込み 英語
S
cambridgeltl
1.7M
49
Gte Large
MIT
GTE-Largeは強力なセンテンストランスフォーマーモデルで、文の類似度とテキスト埋め込みタスクに特化しており、複数のベンチマークテストで優れた性能を発揮します。
テキスト埋め込み 英語
G
thenlper
1.5M
278
Gte Base En V1.5
Apache-2.0
GTE-base-en-v1.5 は英語の文章変換モデルで、文章類似度タスクに特化しており、複数のテキスト埋め込みベンチマークで優れた性能を発揮します。
テキスト埋め込み Transformers 複数言語対応
Transformers 複数言語対応G
Alibaba-NLP
1.5M
63
Gte Multilingual Base
Apache-2.0
GTE Multilingual Base は50以上の言語をサポートする多言語文埋め込みモデルで、文類似度計算などのタスクに適しています。
テキスト埋め込み Transformers 複数言語対応
Transformers 複数言語対応G
Alibaba-NLP
1.2M
246
Polybert
polyBERTは、完全に機械駆動の超高速ポリマー情報学を実現するための化学言語モデルです。PSMILES文字列を600次元の密なフィンガープリントにマッピングし、ポリマー化学構造を数値形式で表現します。
テキスト埋め込み Transformers
TransformersP
kuelumbus
1.0M
5
Bert Base Turkish Cased Mean Nli Stsb Tr
Apache-2.0
トルコ語BERTベースの文埋め込みモデルで、意味的類似性タスクに最適化
テキスト埋め込み Transformers その他
Transformers その他B
emrecan
1.0M
40
GIST Small Embedding V0
MIT
BAAI/bge-small-en-v1.5モデルを微調整したテキスト埋め込みモデルで、MEDIデータセットとMTEB分類タスクデータセットで訓練され、検索タスクのクエリエンコーディング能力を最適化しました。
テキスト埋め込み
Safetensors 英語
Safetensors 英語G
avsolatorio
945.68k
29
おすすめAIモデル
Llama 3 Typhoon V1.5x 8b Instruct
タイ語専用に設計された80億パラメータの命令モデルで、GPT-3.5-turboに匹敵する性能を持ち、アプリケーションシナリオ、検索拡張生成、制限付き生成、推論タスクを最適化
大規模言語モデル Transformers 複数言語対応
Transformers 複数言語対応L
scb10x
3,269
16
Cadet Tiny
Openrail
Cadet-TinyはSODAデータセットでトレーニングされた超小型対話モデルで、エッジデバイス推論向けに設計されており、体積はCosmo-3Bモデルの約2％です。
対話システム Transformers 英語
Transformers 英語C
ToddGoldfarb
2,691
6
Roberta Base Chinese Extractive Qa
RoBERTaアーキテクチャに基づく中国語抽出型QAモデルで、与えられたテキストから回答を抽出するタスクに適しています。
質問応答システム 中国語
R
uer
2,694
98