Gte Large Zh GGUF
模型概述
基於thenlper/gte-large-zh轉換的GGUF格式模型,支持中文文本嵌入和多種自然語言處理任務。
模型特點
多任務評估基準
在多種中文自然語言處理任務上進行了全面評估
句子嵌入
能夠生成高質量的句子嵌入表示
GGUF格式支持
通過llama.cpp框架支持高效推理
模型能力
語義文本相似度計算
文本分類
信息檢索
重排序
文本聚類
句子對分類
使用案例
語義相似度
問答系統
計算問題和候選答案之間的語義相似度
在MTEB AFQMC數據集上達到48.94的餘弦相似度皮爾遜係數
文本匹配
判斷兩個句子是否表達相同含義
在MTEB LCQMC數據集上達到69.51的餘弦相似度皮爾遜係數
信息檢索
醫療問答檢索
從醫療知識庫中檢索相關答案
在MTEB CmedqaRetrieval數據集上前10命中率(MAP@10)達到36.50%
電商商品檢索
根據用戶查詢檢索相關商品
在MTEB EcomRetrieval數據集上前10命中率(MAP@10)達到62.17%
文本分類
情感分析
判斷文本的情感傾向
在MTEB多語言情感分析數據集上準確率達到76.48%
意圖識別
識別用戶輸入的意圖類別
在MTEB Massive意圖分類(中文)數據集上準確率達到73.31%
🚀 linlueird/gte-large-zh-GGUF
本模型是通過 llama.cpp,藉助 ggml.ai 的 GGUF-my-repo 空間,從 thenlper/gte-large-zh
轉換為 GGUF 格式的。有關該模型的更多詳細信息,請參考原始模型卡片。
🚀 快速開始
本模型可通過 llama.cpp 使用,以下是具體步驟。
📦 安裝指南
可通過 brew(適用於 Mac 和 Linux)安裝 llama.cpp:
brew install llama.cpp
💻 使用示例
基礎用法
CLI 方式
llama-cli --hf-repo linlueird/gte-large-zh-GGUF --hf-file gte-large-zh-q4_k_m.gguf -p "The meaning to life and the universe is"
Server 方式
llama-server --hf-repo linlueird/gte-large-zh-GGUF --hf-file gte-large-zh-q4_k_m.gguf -c 2048
高級用法
你也可以直接按照 Llama.cpp 倉庫中列出的使用步驟使用此檢查點。
步驟 1:從 GitHub 克隆 llama.cpp
git clone https://github.com/ggerganov/llama.cpp
步驟 2:進入 llama.cpp 文件夾,並使用 LLAMA_CURL=1
標誌以及其他特定硬件標誌(例如,在 Linux 上使用英偉達 GPU 時使用 LLAMA_CUDA=1
)進行編譯
cd llama.cpp && LLAMA_CURL=1 make
步驟 3:通過主二進制文件運行推理
./llama-cli --hf-repo linlueird/gte-large-zh-GGUF --hf-file gte-large-zh-q4_k_m.gguf -p "The meaning to life and the universe is"
或者
./llama-server --hf-repo linlueird/gte-large-zh-GGUF --hf-file gte-large-zh-q4_k_m.gguf -c 2048
📄 許可證
本項目採用 MIT 許可證。
🔍 模型指標
屬性 | 詳情 |
---|---|
模型類型 | gte-large-zh |
基礎模型 | thenlper/gte-large-zh |
許可證 | MIT |
標籤 | mteb、sentence-similarity、sentence-transformers、Sentence Transformers、llama-cpp、gguf-my-repo |
各任務指標詳情
STS 任務
數據集 | 指標類型 | 值 |
---|---|---|
MTEB AFQMC | cos_sim_pearson | 48.94131905219026 |
MTEB AFQMC | cos_sim_spearman | 54.58261199731436 |
MTEB AFQMC | euclidean_pearson | 52.73929210805982 |
MTEB AFQMC | euclidean_spearman | 54.582632097533676 |
MTEB AFQMC | manhattan_pearson | 52.73123295724949 |
MTEB AFQMC | manhattan_spearman | 54.572941830465794 |
MTEB ATEC | cos_sim_pearson | 47.292931669579005 |
MTEB ATEC | cos_sim_spearman | 54.601019783506466 |
MTEB ATEC | euclidean_pearson | 54.61393532658173 |
MTEB ATEC | euclidean_spearman | 54.60101865708542 |
MTEB ATEC | manhattan_pearson | 54.59369555606305 |
MTEB ATEC | manhattan_spearman | 54.601098593646036 |
MTEB LCQMC | cos_sim_pearson | 69.50947272908907 |
MTEB LCQMC | cos_sim_spearman | 74.40054474949213 |
MTEB LCQMC | euclidean_pearson | 73.53007373987617 |
MTEB LCQMC | euclidean_spearman | 74.40054474732082 |
MTEB LCQMC | manhattan_pearson | 73.51396571849736 |
MTEB LCQMC | manhattan_spearman | 74.38395696630835 |
MTEB PAWSX | cos_sim_pearson | 35.301730226895955 |
MTEB PAWSX | cos_sim_spearman | 38.54612530948101 |
MTEB PAWSX | euclidean_pearson | 39.02831131230217 |
MTEB PAWSX | euclidean_spearman | 38.54612530948101 |
MTEB PAWSX | manhattan_pearson | 39.04765584936325 |
MTEB PAWSX | manhattan_spearman | 38.54455759013173 |
MTEB QBQTC | cos_sim_pearson | 32.27907454729754 |
MTEB QBQTC | cos_sim_spearman | 33.35945567162729 |
MTEB QBQTC | euclidean_pearson | 31.997628193815725 |
MTEB QBQTC | euclidean_spearman | 33.3592386340529 |
MTEB QBQTC | manhattan_pearson | 31.97117833750544 |
MTEB QBQTC | manhattan_spearman | 33.30857326127779 |
MTEB STS22 (zh) | cos_sim_pearson | 62.53712784446981 |
MTEB STS22 (zh) | cos_sim_spearman | 62.975074386224286 |
MTEB STS22 (zh) | euclidean_pearson | 61.791207731290854 |
MTEB STS22 (zh) | euclidean_spearman | 62.975073716988064 |
MTEB STS22 (zh) | manhattan_pearson | 62.63850653150875 |
MTEB STS22 (zh) | manhattan_spearman | 63.56640346497343 |
MTEB STSB | cos_sim_pearson | 79.52067424748047 |
MTEB STSB | cos_sim_spearman | 79.68425102631514 |
MTEB STSB | euclidean_pearson | 79.27553959329275 |
MTEB STSB | euclidean_spearman | 79.68450427089856 |
MTEB STSB | manhattan_pearson | 79.21584650471131 |
MTEB STSB | manhattan_spearman | 79.6419242840243 |
分類任務
數據集 | 指標類型 | 值 |
---|---|---|
MTEB AmazonReviewsClassification (zh) | accuracy | 47.233999999999995 |
MTEB AmazonReviewsClassification (zh) | f1 | 45.68998446563349 |
MTEB IFlyTek | accuracy | 49.60369372835706 |
MTEB IFlyTek | f1 | 38.24016248875209 |
MTEB JDReview | accuracy | 86.71669793621012 |
MTEB JDReview | ap | 55.75807094995178 |
MTEB JDReview | f1 | 81.59033162805417 |
MTEB MassiveIntentClassification (zh-CN) | accuracy | 73.30531271015468 |
MTEB MassiveIntentClassification (zh-CN) | f1 | 70.88091430578575 |
MTEB MassiveScenarioClassification (zh-CN) | accuracy | 75.7128446536651 |
MTEB MassiveScenarioClassification (zh-CN) | f1 | 75.06125593532262 |
MTEB MultilingualSentiment | accuracy | 76.47666666666667 |
MTEB MultilingualSentiment | f1 | 76.4808576632057 |
MTEB OnlineShopping | accuracy | 92.68 |
MTEB OnlineShopping | ap | 90.78652757815115 |
MTEB OnlineShopping | f1 | 92.67153098230253 |
MTEB TNews | accuracy | 51.979000000000006 |
MTEB TNews | f1 | 50.35658238894168 |
MTEB Waimai | accuracy | 88.36999999999999 |
MTEB Waimai | ap | 73.29590829222836 |
MTEB Waimai | f1 | 86.74250506247606 |
聚類任務
數據集 | 指標類型 | 值 |
---|---|---|
MTEB CLSClusteringP2P | v_measure | 42.098169316685045 |
MTEB CLSClusteringS2S | v_measure | 38.90716707051822 |
MTEB ThuNewsClusteringP2P | v_measure | 68.36477832710159 |
MTEB ThuNewsClusteringS2S | v_measure | 62.92080622759053 |
重排序任務
數據集 | 指標類型 | 值 |
---|---|---|
MTEB CMedQAv1 | map | 86.09191911031553 |
MTEB CMedQAv1 | mrr | 88.6747619047619 |
MTEB CMedQAv2 | map | 86.45781885502122 |
MTEB CMedQAv2 | mrr | 89.01591269841269 |
MTEB MMarcoReranking | map | 31.188333827724108 |
MTEB MMarcoReranking | mrr | 29.84801587301587 |
MTEB T2Reranking | map | 65.8563449629786 |
MTEB T2Reranking | mrr | 75.82550832339254 |
檢索任務
數據集 | 指標類型 | 值 |
---|---|---|
MTEB CmedqaRetrieval | map_at_1 | 24.215 |
MTEB CmedqaRetrieval | map_at_10 | 36.498000000000005 |
MTEB CmedqaRetrieval | map_at_100 | 38.409 |
MTEB CmedqaRetrieval | map_at_1000 | 38.524 |
MTEB CmedqaRetrieval | map_at_3 | 32.428000000000004 |
MTEB CmedqaRetrieval | map_at_5 | 34.664 |
MTEB CmedqaRetrieval | mrr_at_1 | 36.834 |
MTEB CmedqaRetrieval | mrr_at_10 | 45.196 |
MTEB CmedqaRetrieval | mrr_at_100 | 46.214 |
MTEB CmedqaRetrieval | mrr_at_1000 | 46.259 |
MTEB CmedqaRetrieval | mrr_at_3 | 42.631 |
MTEB CmedqaRetrieval | mrr_at_5 | 44.044 |
MTEB CmedqaRetrieval | ndcg_at_1 | 36.834 |
MTEB CmedqaRetrieval | ndcg_at_10 | 43.146 |
MTEB CmedqaRetrieval | ndcg_at_100 | 50.632999999999996 |
MTEB CmedqaRetrieval | ndcg_at_1000 | 52.608999999999995 |
MTEB CmedqaRetrieval | ndcg_at_3 | 37.851 |
MTEB CmedqaRetrieval | ndcg_at_5 | 40.005 |
MTEB CmedqaRetrieval | precision_at_1 | 36.834 |
MTEB CmedqaRetrieval | precision_at_10 | 9.647 |
MTEB CmedqaRetrieval | precision_at_100 | 1.574 |
MTEB CmedqaRetrieval | precision_at_1000 | 0.183 |
MTEB CmedqaRetrieval | precision_at_3 | 21.48 |
MTEB CmedqaRetrieval | precision_at_5 | 15.649 |
MTEB CmedqaRetrieval | recall_at_1 | 24.215 |
MTEB CmedqaRetrieval | recall_at_10 | 54.079 |
MTEB CmedqaRetrieval | recall_at_100 | 84.943 |
MTEB CmedqaRetrieval | recall_at_1000 | 98.098 |
MTEB CmedqaRetrieval | recall_at_3 | 38.117000000000004 |
MTEB CmedqaRetrieval | recall_at_5 | 44.775999999999996 |
MTEB CovidRetrieval | map_at_1 | 78.583 |
MTEB CovidRetrieval | map_at_10 | 85.613 |
MTEB CovidRetrieval | map_at_100 | 85.777 |
MTEB CovidRetrieval | map_at_1000 | 85.77900000000001 |
MTEB CovidRetrieval | map_at_3 | 84.58 |
MTEB CovidRetrieval | map_at_5 | 85.22800000000001 |
MTEB CovidRetrieval | mrr_at_1 | 78.925 |
MTEB CovidRetrieval | mrr_at_10 | 85.667 |
MTEB CovidRetrieval | mrr_at_100 | 85.822 |
MTEB CovidRetrieval | mrr_at_1000 | 85.824 |
MTEB CovidRetrieval | mrr_at_3 | 84.651 |
MTEB CovidRetrieval | mrr_at_5 | 85.299 |
MTEB CovidRetrieval | ndcg_at_1 | 78.925 |
MTEB CovidRetrieval | ndcg_at_10 | 88.405 |
MTEB CovidRetrieval | ndcg_at_100 | 89.02799999999999 |
MTEB CovidRetrieval | ndcg_at_1000 | 89.093 |
MTEB CovidRetrieval | ndcg_at_3 | 86.393 |
MTEB CovidRetrieval | ndcg_at_5 | 87.5 |
MTEB CovidRetrieval | precision_at_1 | 78.925 |
MTEB CovidRetrieval | precision_at_10 | 9.789 |
MTEB CovidRetrieval | precision_at_100 | 1.005 |
MTEB CovidRetrieval | precision_at_1000 | 0.101 |
MTEB CovidRetrieval | precision_at_3 | 30.769000000000002 |
MTEB CovidRetrieval | precision_at_5 | 19.031000000000002 |
MTEB CovidRetrieval | recall_at_1 | 78.583 |
MTEB CovidRetrieval | recall_at_10 | 96.891 |
MTEB CovidRetrieval | recall_at_100 | 99.473 |
MTEB CovidRetrieval | recall_at_1000 | 100.0 |
MTEB CovidRetrieval | recall_at_3 | 91.438 |
MTEB CovidRetrieval | recall_at_5 | 94.152 |
MTEB DuRetrieval | map_at_1 | 25.604 |
MTEB DuRetrieval | map_at_10 | 77.171 |
MTEB DuRetrieval | map_at_100 | 80.033 |
MTEB DuRetrieval | map_at_1000 | 80.099 |
MTEB DuRetrieval | map_at_3 | 54.364000000000004 |
MTEB DuRetrieval | map_at_5 | 68.024 |
MTEB DuRetrieval | mrr_at_1 | 89.85 |
MTEB DuRetrieval | mrr_at_10 | 93.009 |
MTEB DuRetrieval | mrr_at_100 | 93.065 |
MTEB DuRetrieval | mrr_at_1000 | 93.068 |
MTEB DuRetrieval | mrr_at_3 | 92.72500000000001 |
MTEB DuRetrieval | mrr_at_5 | 92.915 |
MTEB DuRetrieval | ndcg_at_1 | 89.85 |
MTEB DuRetrieval | ndcg_at_10 | 85.038 |
MTEB DuRetrieval | ndcg_at_100 | 88.247 |
MTEB DuRetrieval | ndcg_at_1000 | 88.837 |
MTEB DuRetrieval | ndcg_at_3 | 85.20299999999999 |
MTEB DuRetrieval | ndcg_at_5 | 83.47 |
MTEB DuRetrieval | precision_at_1 | 89.85 |
MTEB DuRetrieval | precision_at_10 | 40.275 |
MTEB DuRetrieval | precision_at_100 | 4.709 |
MTEB DuRetrieval | precision_at_1000 | 0.486 |
MTEB DuRetrieval | precision_at_3 | 76.36699999999999 |
MTEB DuRetrieval | precision_at_5 | 63.75999999999999 |
MTEB DuRetrieval | recall_at_1 | 25.604 |
MTEB DuRetrieval | recall_at_10 | 85.423 |
MTEB DuRetrieval | recall_at_100 | 95.695 |
MTEB DuRetrieval | recall_at_1000 | 98.669 |
MTEB DuRetrieval | recall_at_3 | 56.737 |
MTEB DuRetrieval | recall_at_5 | 72.646 |
MTEB EcomRetrieval | map_at_1 | 51.800000000000004 |
MTEB EcomRetrieval | map_at_10 | 62.17 |
MTEB EcomRetrieval | map_at_100 | 62.649 |
MTEB EcomRetrieval | map_at_1000 | 62.663000000000004 |
MTEB EcomRetrieval | map_at_3 | 59.699999999999996 |
MTEB EcomRetrieval | map_at_5 | 61.23499999999999 |
MTEB EcomRetrieval | mrr_at_1 | 51.800000000000004 |
MTEB EcomRetrieval | mrr_at_10 | 62.17 |
MTEB EcomRetrieval | mrr_at_100 | 62.649 |
MTEB EcomRetrieval | mrr_at_1000 | 62.663000000000004 |
MTEB EcomRetrieval | mrr_at_3 | 59.699999999999996 |
MTEB EcomRetrieval | mrr_at_5 | 61.23499999999999 |
MTEB EcomRetrieval | ndcg_at_1 | 51.800000000000004 |
MTEB EcomRetrieval | ndcg_at_10 | 67.246 |
MTEB EcomRetrieval | ndcg_at_100 | 69.58 |
MTEB EcomRetrieval | ndcg_at_1000 | 69.925 |
MTEB EcomRetrieval | ndcg_at_3 | 62.197 |
MTEB EcomRetrieval | ndcg_at_5 | 64.981 |
MTEB EcomRetrieval | precision_at_1 | 51.800000000000004 |
MTEB EcomRetrieval | precision_at_10 | 8.32 |
MTEB EcomRetrieval | precision_at_100 | 0.941 |
MTEB EcomRetrieval | precision_at_1000 | 0.097 |
MTEB EcomRetrieval | precision_at_3 | 23.133 |
MTEB EcomRetrieval | precision_at_5 | 15.24 |
MTEB EcomRetrieval | recall_at_1 | 51.800000000000004 |
MTEB EcomRetrieval | recall_at_10 | 83.2 |
MTEB EcomRetrieval | recall_at_100 | 94.1 |
MTEB EcomRetrieval | recall_at_1000 | 96.8 |
MTEB EcomRetrieval | recall_at_3 | 69.39999999999999 |
MTEB EcomRetrieval | recall_at_5 | 76.2 |
MTEB MMarcoRetrieval | map_at_1 | 64.685 |
MTEB MMarcoRetrieval | map_at_10 | 73.803 |
MTEB MMarcoRetrieval | map_at_100 | 74.153 |
MTEB MMarcoRetrieval | map_at_1000 | 74.167 |
MTEB MMarcoRetrieval | map_at_3 | 71.98 |
MTEB MMarcoRetrieval | map_at_5 | 73.21600000000001 |
MTEB MMarcoRetrieval | mrr_at_1 | 66.891 |
MTEB MMarcoRetrieval | mrr_at_10 | 74.48700000000001 |
MTEB MMarcoRetrieval | mrr_at_100 | 74.788 |
MTEB MMarcoRetrieval | mrr_at_1000 | 74.801 |
MTEB MMarcoRetrieval | mrr_at_3 | 72.918 |
MTEB MMarcoRetrieval | mrr_at_5 | 73.965 |
MTEB MMarcoRetrieval | ndcg_at_1 | 66.891 |
MTEB MMarcoRetrieval | ndcg_at_10 | 77.534 |
MTEB MMarcoRetrieval | ndcg_at_100 | 79.106 |
MTEB MMarcoRetrieval | ndcg_at_1000 | 79.494 |
MTEB MMarcoRetrieval | ndcg_at_3 | 74.13499999999999 |
MTEB MMarcoRetrieval | ndcg_at_5 | 76.20700000000001 |
MTEB MMarcoRetrieval | precision_at_1 | 66.891 |
MTEB MMarcoRetrieval | precision_at_10 | 9.375 |
MTEB MMarcoRetrieval | precision_at_100 | 1.0170000000000001 |
MTEB MMarcoRetrieval | precision_at_1000 | 0.105 |
MTEB MMarcoRetrieval | precision_at_3 | 27.932000000000002 |
MTEB MMarcoRetrieval | precision_at_5 | 17.86 |
MTEB MMarcoRetrieval | recall_at_1 | 64.685 |
MTEB MMarcoRetrieval | recall_at_10 | 88.298 |
MTEB MMarcoRetrieval | recall_at_100 | 95.426 |
MTEB MMarcoRetrieval | recall_at_1000 | 98.48700000000001 |
MTEB MMarcoRetrieval | recall_at_3 | 79.44200000000001 |
MTEB MMarcoRetrieval | recall_at_5 | 84.358 |
MTEB MedicalRetrieval | map_at_1 | 52.7 |
MTEB MedicalRetrieval | map_at_10 | 59.532 |
MTEB MedicalRetrieval | map_at_100 | 60.085 |
MTEB MedicalRetrieval | map_at_1000 | 60.126000000000005 |
MTEB MedicalRetrieval | map_at_3 | 57.767 |
MTEB MedicalRetrieval | map_at_5 | 58.952000000000005 |
MTEB MedicalRetrieval | mrr_at_1 | 52.900000000000006 |
MTEB MedicalRetrieval | mrr_at_10 | 59.648999999999994 |
MTEB MedicalRetrieval | mrr_at_100 | 60.20100000000001 |
MTEB MedicalRetrieval | mrr_at_1000 | 60.242 |
MTEB MedicalRetrieval | mrr_at_3 | 57.882999999999996 |
MTEB MedicalRetrieval | mrr_at_5 | 59.068 |
MTEB MedicalRetrieval | ndcg_at_1 | 52.7 |
MTEB MedicalRetrieval | ndcg_at_10 | 62.883 |
MTEB MedicalRetrieval | ndcg_at_100 | 65.714 |
MTEB MedicalRetrieval | ndcg_at_1000 | 66.932 |
MTEB MedicalRetrieval | ndcg_at_3 | 59.34700000000001 |
MTEB MedicalRetrieval | ndcg_at_5 | 61.486 |
MTEB MedicalRetrieval | precision_at_1 | 52.7 |
MTEB MedicalRetrieval | precision_at_10 | 7.340000000000001 |
MTEB MedicalRetrieval | precision_at_100 | 0.8699999999999999 |
MTEB MedicalRetrieval | precision_at_1000 | 0.097 |
MTEB MedicalRetrieval | precision_at_3 | 21.3 |
MTEB MedicalRetrieval | precision_at_5 | 13.819999999999999 |
MTEB MedicalRetrieval | recall_at_1 | 52.7 |
MTEB MedicalRetrieval | recall_at_10 | 73.4 |
MTEB MedicalRetrieval | recall_at_100 | 87.0 |
MTEB MedicalRetrieval | recall_at_1000 | 96.8 |
MTEB MedicalRetrieval | recall_at_3 | 63.9 |
MTEB MedicalRetrieval | recall_at_5 | 69.1 |
MTEB T2Retrieval | map_at_1 | 27.889999999999997 |
MTEB T2Retrieval | map_at_10 | 72.878 |
MTEB T2Retrieval | map_at_100 | 76.737 |
MTEB T2Retrieval | map_at_1000 | 76.836 |
MTEB T2Retrieval | map_at_3 | 52.738 |
MTEB T2Retrieval | map_at_5 | 63.726000000000006 |
MTEB T2Retrieval | mrr_at_1 | 89.35600000000001 |
MTEB T2Retrieval | mrr_at_10 | 92.622 |
MTEB T2Retrieval | mrr_at_100 | 92.692 |
MTEB T2Retrieval | mrr_at_1000 | 92.694 |
MTEB T2Retrieval | mrr_at_3 | 92.13799999999999 |
MTEB T2Retrieval | mrr_at_5 | 92.452 |
MTEB T2Retrieval | ndcg_at_1 | 89.35600000000001 |
MTEB T2Retrieval | ndcg_at_10 | 81.932 |
MTEB T2Retrieval | ndcg_at_100 | 86.351 |
MTEB T2Retrieval | ndcg_at_1000 | 87.221 |
MTEB T2Retrieval | ndcg_at_3 | 84.29100000000001 |
MTEB T2Retrieval | ndcg_at_5 | 82.279 |
MTEB T2Retrieval | precision_at_1 | 89.35600000000001 |
MTEB T2Retrieval | precision_at_10 | 39.511 |
MTEB T2Retrieval | precision_at_100 | 4.901 |
MTEB T2Retrieval | precision_at_1000 | 0.513 |
MTEB T2Retrieval | precision_at_3 | 72.62100000000001 |
MTEB T2Retrieval | precision_at_5 | 59.918000000000006 |
MTEB T2Retrieval | recall_at_1 | 27.889999999999997 |
MTEB T2Retrieval | recall_at_10 | 80.636 |
MTEB T2Retrieval | recall_at_100 | 94.333 |
MTEB T2Retrieval | recall_at_1000 | 98.39099999999999 |
MTEB T2Retrieval | recall_at_3 | 54.797 |
MTEB T2Retrieval | recall_at_5 | 67.824 |
MTEB VideoRetrieval | map_at_1 | 59.3 |
MTEB VideoRetrieval | map_at_10 | 69.299 |
MTEB VideoRetrieval | map_at_100 | 69.669 |
MTEB VideoRetrieval | map_at_1000 | 69.682 |
MTEB VideoRetrieval | map_at_3 | 67.583 |
MTEB VideoRetrieval | map_at_5 | 68.57799999999999 |
MTEB VideoRetrieval | mrr_at_1 | 59.3 |
MTEB VideoRetrieval | mrr_at_10 | 69.299 |
MTEB VideoRetrieval | mrr_at_100 | 69.669 |
MTEB VideoRetrieval | mrr_at_1000 | 69.682 |
MTEB VideoRetrieval | mrr_at_3 | 67.583 |
MTEB VideoRetrieval | mrr_at_5 | 68.57799999999999 |
MTEB VideoRetrieval | ndcg_at_1 | 59.3 |
MTEB VideoRetrieval | ndcg_at_10 | 73.699 |
MTEB VideoRetrieval | ndcg_at_100 | 75.626 |
MTEB VideoRetrieval | ndcg_at_1000 | 75.949 |
MTEB VideoRetrieval | ndcg_at_3 | 70.18900000000001 |
MTEB VideoRetrieval | ndcg_at_5 | 71.992 |
MTEB VideoRetrieval | precision_at_1 | 59.3 |
MTEB VideoRetrieval | precision_at_10 | 8.73 |
MTEB VideoRetrieval | precision_at_100 | 0.9650000000000001 |
MTEB VideoRetrieval | precision_at_1000 | 0.099 |
MTEB VideoRetrieval | precision_at_3 | 25.900000000000002 |
MTEB VideoRetrieval | precision_at_5 | 16.42 |
MTEB VideoRetrieval | recall_at_1 | 59.3 |
MTEB VideoRetrieval | recall_at_10 | 87.3 |
MTEB VideoRetrieval | recall_at_100 | 96.5 |
MTEB VideoRetrieval | recall_at_1000 | 99.0 |
MTEB VideoRetrieval | recall_at_3 | 77.7 |
MTEB VideoRetrieval | recall_at_5 | 82.1 |
成對分類任務
數據集 | 指標類型 | 值 |
---|---|---|
MTEB Cmnli | cos_sim_accuracy | 82.51352976548407 |
MTEB Cmnli | cos_sim_ap | 89.49905141462749 |
MTEB Cmnli | cos_sim_f1 | 83.89334489486234 |
MTEB Cmnli | cos_sim_precision | 78.19761567993534 |
MTEB Cmnli | cos_sim_recall | 90.48398410100538 |
MTEB Cmnli | dot_accuracy | 82.51352976548407 |
MTEB Cmnli | dot_ap | 89.49108293121158 |
MTEB Cmnli | dot_f1 | 83.89334489486234 |
MTEB Cmnli | dot_precision | 78.19761567993534 |
MTEB Cmnli | dot_recall | 90.48398410100538 |
MTEB Cmnli | euclidean_accuracy | 82.51352976548407 |
MTEB Cmnli | euclidean_ap | 89.49904709975154 |
MTEB Cmnli | euclidean_f1 | 83.89334489486234 |
MTEB Cmnli | euclidean_precision | 78.19761567993534 |
MTEB Cmnli | euclidean_recall | 90.48398410100538 |
MTEB Cmnli | manhattan_accuracy | 82.48947684906794 |
MTEB Cmnli | manhattan_ap | 89.49231995962901 |
MTEB Cmnli | manhattan_f1 | 83.84681215233205 |
MTEB Cmnli | manhattan_precision | 77.28258726089528 |
MTEB Cmnli | manhattan_recall | 91.62964694879588 |
MTEB Cmnli | max_accuracy | 82.51352976548407 |
MTEB Cmnli | max_ap | 89.49905141462749 |
MTEB Cmnli | max_f1 | 83.89334489486234 |
MTEB Ocnli | cos_sim_accuracy | 77.58527341635084 |
MTEB Ocnli | cos_sim_ap | 79.32131557636497 |
MTEB Ocnli | cos_sim_f1 | 80.51948051948052 |
MTEB Ocnli | cos_sim_precision | 71.7948717948718 |
MTEB Ocnli | cos_sim_recall | 91.65786694825766 |
MTEB Ocnli | dot_accuracy | 77.58527341635084 |
MTEB Ocnli | dot_ap | 79.32131557636497 |
MTEB Ocnli | dot_f1 | 80.51948051948052 |
MTEB Ocnli | dot_precision | 71.7948717948718 |
MTEB Ocnli | dot_recall | 91.65786694825766 |
MTEB Ocnli | euclidean_accuracy | 77.58527341635084 |
MTEB Ocnli | euclidean_ap | 79.32131557636497 |
MTEB Ocnli | euclidean_f1 | 80.51948051948052 |
MTEB Ocnli | euclidean_precision | 71.7948717948718 |
MTEB Ocnli | euclidean_recall | 91.65786694825766 |
MTEB Ocnli | manhattan_accuracy | 77.15213860314023 |
MTEB Ocnli | manhattan_ap | 79.26178519246496 |
MTEB Ocnli | manhattan_f1 | 80.22028453418999 |
MTEB Ocnli | manhattan_precision | 70.94155844155844 |
MTEB Ocnli | manhattan_recall | 92.29144667370645 |
MTEB Ocnli | max_accuracy | 77.58527341635084 |
MTEB Ocnli | max_ap | 79.32131557636497 |
MTEB Ocnli | max_f1 | 80.51948051948052 |
Jina Embeddings V3
Jina Embeddings V3 是一個多語言句子嵌入模型,支持超過100種語言,專注於句子相似度和特徵提取任務。
文本嵌入
Transformers 支持多種語言

J
jinaai
3.7M
911
Ms Marco MiniLM L6 V2
Apache-2.0
基於MS Marco段落排序任務訓練的交叉編碼器模型,用於信息檢索中的查詢-段落相關性評分
文本嵌入 英語
M
cross-encoder
2.5M
86
Opensearch Neural Sparse Encoding Doc V2 Distill
Apache-2.0
基於蒸餾技術的稀疏檢索模型,專為OpenSearch優化,支持免推理文檔編碼,在搜索相關性和效率上優於V1版本
文本嵌入
Transformers 英語

O
opensearch-project
1.8M
7
Sapbert From PubMedBERT Fulltext
Apache-2.0
基於PubMedBERT的生物醫學實體表徵模型,通過自對齊預訓練優化語義關係捕捉
文本嵌入 英語
S
cambridgeltl
1.7M
49
Gte Large
MIT
GTE-Large 是一個強大的句子轉換器模型,專注於句子相似度和文本嵌入任務,在多個基準測試中表現出色。
文本嵌入 英語
G
thenlper
1.5M
278
Gte Base En V1.5
Apache-2.0
GTE-base-en-v1.5 是一個英文句子轉換器模型,專注於句子相似度任務,在多個文本嵌入基準測試中表現優異。
文本嵌入
Transformers 支持多種語言

G
Alibaba-NLP
1.5M
63
Gte Multilingual Base
Apache-2.0
GTE Multilingual Base 是一個多語言的句子嵌入模型,支持超過50種語言,適用於句子相似度計算等任務。
文本嵌入
Transformers 支持多種語言

G
Alibaba-NLP
1.2M
246
Polybert
polyBERT是一個化學語言模型,旨在實現完全由機器驅動的超快聚合物信息學。它將PSMILES字符串映射為600維密集指紋,以數值形式表示聚合物化學結構。
文本嵌入
Transformers

P
kuelumbus
1.0M
5
Bert Base Turkish Cased Mean Nli Stsb Tr
Apache-2.0
基於土耳其語BERT的句子嵌入模型,專為語義相似度任務優化
文本嵌入
Transformers 其他

B
emrecan
1.0M
40
GIST Small Embedding V0
MIT
基於BAAI/bge-small-en-v1.5模型微調的文本嵌入模型,通過MEDI數據集與MTEB分類任務數據集訓練,優化了檢索任務的查詢編碼能力。
文本嵌入
Safetensors 英語
G
avsolatorio
945.68k
29
精選推薦AI模型
Llama 3 Typhoon V1.5x 8b Instruct
專為泰語設計的80億參數指令模型,性能媲美GPT-3.5-turbo,優化了應用場景、檢索增強生成、受限生成和推理任務
大型語言模型
Transformers 支持多種語言

L
scb10x
3,269
16
Cadet Tiny
Openrail
Cadet-Tiny是一個基於SODA數據集訓練的超小型對話模型,專為邊緣設備推理設計,體積僅為Cosmo-3B模型的2%左右。
對話系統
Transformers 英語

C
ToddGoldfarb
2,691
6
Roberta Base Chinese Extractive Qa
基於RoBERTa架構的中文抽取式問答模型,適用於從給定文本中提取答案的任務。
問答系統 中文
R
uer
2,694
98