Conan Embedding V1 Q4 K M GGUF
模型概述
該模型專注於中文文本的嵌入表示生成,支持語義相似度計算、文本分類、聚類、檢索和重排序等多種任務,在多箇中文基準測試中表現出色。
模型特點
多任務支持
支持多種中文NLP任務,包括語義相似度計算、文本分類、聚類、檢索和重排序等。
高性能
在多箇中文基準測試中表現優異,特別是在醫療領域相關任務上表現突出。
中文優化
專門針對中文文本進行優化,能夠更好地捕捉中文語義特徵。
模型能力
文本嵌入生成
語義相似度計算
文本分類
文本聚類
信息檢索
搜索結果重排序
使用案例
醫療領域
醫療問答檢索
用於醫療相關問題的檢索系統,幫助用戶快速找到相關醫療信息。
在CMedQA檢索任務中,map@100達到42.495
醫療文檔重排序
對醫療文檔檢索結果進行相關性重排序,提升用戶體驗。
在CMedQAv1重排序任務中,mrr達到93.358
電子商務
商品評論分類
對電商平臺的商品評論進行情感和主題分類。
在JDReview分類任務中,準確率達到90.318%
商品檢索
提升電商平臺的商品搜索相關性。
在EcomRetrieval任務中,ndcg@10達到70.991
通用NLP
語義相似度計算
計算兩段中文文本的語義相似度。
在STSB任務中,cos_sim_spearman達到81.244
文本聚類
對中文文本進行無監督聚類分析。
在CLSClusteringP2P任務中,v_measure達到60.635
🚀 lagoon999/Conan-embedding-v1-Q4_K_M-GGUF
本模型是通過 llama.cpp 並藉助 ggml.ai 的 GGUF-my-repo 空間,從 TencentBAC/Conan-embedding-v1
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
本模型可與 llama.cpp 結合使用,以下是具體的使用步驟。
📦 安裝指南
可以通過 brew(適用於 Mac 和 Linux)安裝 llama.cpp:
brew install llama.cpp
💻 使用示例
基礎用法
可以通過 CLI 或 Server 調用該模型。
CLI
llama-cli --hf-repo lagoon999/Conan-embedding-v1-Q4_K_M-GGUF --hf-file conan-embedding-v1-q4_k_m.gguf -p "The meaning to life and the universe is"
Server
llama-server --hf-repo lagoon999/Conan-embedding-v1-Q4_K_M-GGUF --hf-file conan-embedding-v1-q4_k_m.gguf -c 2048
高級用法
也可以直接按照 Llama.cpp 倉庫中列出的 使用步驟 使用該檢查點。
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
標誌以及其他特定於硬件的標誌(例如,在 Linux 上使用 Nvidia 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 lagoon999/Conan-embedding-v1-Q4_K_M-GGUF --hf-file conan-embedding-v1-q4_k_m.gguf -p "The meaning to life and the universe is"
或者
./llama-server --hf-repo lagoon999/Conan-embedding-v1-Q4_K_M-GGUF --hf-file conan-embedding-v1-q4_k_m.gguf -c 2048
📚 詳細文檔
模型指標
以下是該模型在多個任務和數據集上的評估指標:
任務類型 | 數據集名稱 | 指標類型 | 指標值 |
---|---|---|---|
STS | MTEB AFQMC | cos_sim_pearson | 56.613572467148856 |
STS | MTEB AFQMC | cos_sim_spearman | 60.66446211824284 |
STS | MTEB AFQMC | euclidean_pearson | 58.42080485872613 |
STS | MTEB AFQMC | euclidean_spearman | 59.82750030458164 |
STS | MTEB AFQMC | manhattan_pearson | 58.39885271199772 |
STS | MTEB AFQMC | manhattan_spearman | 59.817749720366734 |
STS | MTEB ATEC | cos_sim_pearson | 56.60530380552331 |
STS | MTEB ATEC | cos_sim_spearman | 58.63822441736707 |
STS | MTEB ATEC | euclidean_pearson | 62.18551665180664 |
STS | MTEB ATEC | euclidean_spearman | 58.23168804495912 |
STS | MTEB ATEC | manhattan_pearson | 62.17191480770053 |
STS | MTEB ATEC | manhattan_spearman | 58.22556219601401 |
Classification | MTEB AmazonReviewsClassification (zh) | accuracy | 50.308 |
Classification | MTEB AmazonReviewsClassification (zh) | f1 | 46.927458607895126 |
STS | MTEB BQ | cos_sim_pearson | 72.6472074172711 |
STS | MTEB BQ | cos_sim_spearman | 74.50748447236577 |
STS | MTEB BQ | euclidean_pearson | 72.51833296451854 |
STS | MTEB BQ | euclidean_spearman | 73.9898922606105 |
STS | MTEB BQ | manhattan_pearson | 72.50184948939338 |
STS | MTEB BQ | manhattan_spearman | 73.97797921509638 |
Clustering | MTEB CLSClusteringP2P | v_measure | 60.63545326048343 |
Clustering | MTEB CLSClusteringS2S | v_measure | 52.64834762325994 |
Reranking | MTEB CMedQAv1 | map | 91.38528814655234 |
Reranking | MTEB CMedQAv1 | mrr | 93.35857142857144 |
Reranking | MTEB CMedQAv2 | map | 89.72084678877096 |
Reranking | MTEB CMedQAv2 | mrr | 91.74380952380953 |
Retrieval | MTEB CmedqaRetrieval | map_at_1 | 26.987 |
Retrieval | MTEB CmedqaRetrieval | map_at_10 | 40.675 |
Retrieval | MTEB CmedqaRetrieval | map_at_100 | 42.495 |
Retrieval | MTEB CmedqaRetrieval | map_at_1000 | 42.596000000000004 |
Retrieval | MTEB CmedqaRetrieval | map_at_3 | 36.195 |
Retrieval | MTEB CmedqaRetrieval | map_at_5 | 38.704 |
Retrieval | MTEB CmedqaRetrieval | mrr_at_1 | 41.21 |
Retrieval | MTEB CmedqaRetrieval | mrr_at_10 | 49.816 |
Retrieval | MTEB CmedqaRetrieval | mrr_at_100 | 50.743 |
Retrieval | MTEB CmedqaRetrieval | mrr_at_1000 | 50.77700000000001 |
Retrieval | MTEB CmedqaRetrieval | mrr_at_3 | 47.312 |
Retrieval | MTEB CmedqaRetrieval | mrr_at_5 | 48.699999999999996 |
Retrieval | MTEB CmedqaRetrieval | ndcg_at_1 | 41.21 |
Retrieval | MTEB CmedqaRetrieval | ndcg_at_10 | 47.606 |
Retrieval | MTEB CmedqaRetrieval | ndcg_at_100 | 54.457 |
Retrieval | MTEB CmedqaRetrieval | ndcg_at_1000 | 56.16100000000001 |
Retrieval | MTEB CmedqaRetrieval | ndcg_at_3 | 42.108000000000004 |
Retrieval | MTEB CmedqaRetrieval | ndcg_at_5 | 44.393 |
Retrieval | MTEB CmedqaRetrieval | precision_at_1 | 41.21 |
Retrieval | MTEB CmedqaRetrieval | precision_at_10 | 10.593 |
Retrieval | MTEB CmedqaRetrieval | precision_at_100 | 1.609 |
Retrieval | MTEB CmedqaRetrieval | precision_at_1000 | 0.183 |
Retrieval | MTEB CmedqaRetrieval | precision_at_3 | 23.881 |
Retrieval | MTEB CmedqaRetrieval | precision_at_5 | 17.339 |
Retrieval | MTEB CmedqaRetrieval | recall_at_1 | 26.987 |
Retrieval | MTEB CmedqaRetrieval | recall_at_10 | 58.875 |
Retrieval | MTEB CmedqaRetrieval | recall_at_100 | 87.023 |
Retrieval | MTEB CmedqaRetrieval | recall_at_1000 | 98.328 |
Retrieval | MTEB CmedqaRetrieval | recall_at_3 | 42.265 |
Retrieval | MTEB CmedqaRetrieval | recall_at_5 | 49.334 |
PairClassification | MTEB Cmnli | cos_sim_accuracy | 85.91701743836441 |
PairClassification | MTEB Cmnli | cos_sim_ap | 92.53650618807644 |
PairClassification | MTEB Cmnli | cos_sim_f1 | 86.80265975431082 |
PairClassification | MTEB Cmnli | cos_sim_precision | 83.79025239338556 |
PairClassification | MTEB Cmnli | cos_sim_recall | 90.039747486556 |
PairClassification | MTEB Cmnli | dot_accuracy | 77.17378232110643 |
PairClassification | MTEB Cmnli | dot_ap | 85.40244368166546 |
PairClassification | MTEB Cmnli | dot_f1 | 79.03038001481951 |
PairClassification | MTEB Cmnli | dot_precision | 72.20502901353966 |
PairClassification | MTEB Cmnli | dot_recall | 87.2808043020809 |
PairClassification | MTEB Cmnli | euclidean_accuracy | 84.65423932651834 |
PairClassification | MTEB Cmnli | euclidean_ap | 91.47775530034588 |
PairClassification | MTEB Cmnli | euclidean_f1 | 85.64471499723298 |
PairClassification | MTEB Cmnli | euclidean_precision | 81.31567885666246 |
PairClassification | MTEB Cmnli | euclidean_recall | 90.46060322656068 |
PairClassification | MTEB Cmnli | manhattan_accuracy | 84.58208057726999 |
PairClassification | MTEB Cmnli | manhattan_ap | 91.46228709402014 |
PairClassification | MTEB Cmnli | manhattan_f1 | 85.6631626034444 |
PairClassification | MTEB Cmnli | manhattan_precision | 82.10075026795283 |
PairClassification | MTEB Cmnli | manhattan_recall | 89.5487491232172 |
PairClassification | MTEB Cmnli | max_accuracy | 85.91701743836441 |
PairClassification | MTEB Cmnli | max_ap | 92.53650618807644 |
PairClassification | MTEB Cmnli | max_f1 | 86.80265975431082 |
Retrieval | MTEB CovidRetrieval | map_at_1 | 83.693 |
Retrieval | MTEB CovidRetrieval | map_at_10 | 90.098 |
Retrieval | MTEB CovidRetrieval | map_at_100 | 90.145 |
Retrieval | MTEB CovidRetrieval | map_at_1000 | 90.146 |
Retrieval | MTEB CovidRetrieval | map_at_3 | 89.445 |
Retrieval | MTEB CovidRetrieval | map_at_5 | 89.935 |
Retrieval | MTEB CovidRetrieval | mrr_at_1 | 83.878 |
Retrieval | MTEB CovidRetrieval | mrr_at_10 | 90.007 |
Retrieval | MTEB CovidRetrieval | mrr_at_100 | 90.045 |
Retrieval | MTEB CovidRetrieval | mrr_at_1000 | 90.046 |
Retrieval | MTEB CovidRetrieval | mrr_at_3 | 89.34 |
Retrieval | MTEB CovidRetrieval | mrr_at_5 | 89.835 |
Retrieval | MTEB CovidRetrieval | ndcg_at_1 | 84.089 |
Retrieval | MTEB CovidRetrieval | ndcg_at_10 | 92.351 |
Retrieval | MTEB CovidRetrieval | ndcg_at_100 | 92.54599999999999 |
Retrieval | MTEB CovidRetrieval | ndcg_at_1000 | 92.561 |
Retrieval | MTEB CovidRetrieval | ndcg_at_3 | 91.15299999999999 |
Retrieval | MTEB CovidRetrieval | ndcg_at_5 | 91.968 |
Retrieval | MTEB CovidRetrieval | precision_at_1 | 84.089 |
Retrieval | MTEB CovidRetrieval | precision_at_10 | 10.011000000000001 |
Retrieval | MTEB CovidRetrieval | precision_at_100 | 1.009 |
Retrieval | MTEB CovidRetrieval | precision_at_1000 | 0.101 |
Retrieval | MTEB CovidRetrieval | precision_at_3 | 32.28 |
Retrieval | MTEB CovidRetrieval | precision_at_5 | 19.789 |
Retrieval | MTEB CovidRetrieval | recall_at_1 | 83.693 |
Retrieval | MTEB CovidRetrieval | recall_at_10 | 99.05199999999999 |
Retrieval | MTEB CovidRetrieval | recall_at_100 | 99.895 |
Retrieval | MTEB CovidRetrieval | recall_at_1000 | 100 |
Retrieval | MTEB CovidRetrieval | recall_at_3 | 95.917 |
Retrieval | MTEB CovidRetrieval | recall_at_5 | 97.893 |
Retrieval | MTEB DuRetrieval | map_at_1 | 26.924 |
Retrieval | MTEB DuRetrieval | map_at_10 | 81.392 |
Retrieval | MTEB DuRetrieval | map_at_100 | 84.209 |
Retrieval | MTEB DuRetrieval | map_at_1000 | 84.237 |
Retrieval | MTEB DuRetrieval | map_at_3 | 56.998000000000005 |
Retrieval | MTEB DuRetrieval | map_at_5 | 71.40100000000001 |
Retrieval | MTEB DuRetrieval | mrr_at_1 | 91.75 |
Retrieval | MTEB DuRetrieval | mrr_at_10 | 94.45 |
Retrieval | MTEB DuRetrieval | mrr_at_100 | 94.503 |
Retrieval | MTEB DuRetrieval | mrr_at_1000 | 94.505 |
Retrieval | MTEB DuRetrieval | mrr_at_3 | 94.258 |
Retrieval | MTEB DuRetrieval | mrr_at_5 | 94.381 |
Retrieval | MTEB DuRetrieval | ndcg_at_1 | 91.75 |
Retrieval | MTEB DuRetrieval | ndcg_at_10 | 88.53 |
Retrieval | MTEB DuRetrieval | ndcg_at_100 | 91.13900000000001 |
Retrieval | MTEB DuRetrieval | ndcg_at_1000 | 91.387 |
Retrieval | MTEB DuRetrieval | ndcg_at_3 | 87.925 |
Retrieval | MTEB DuRetrieval | ndcg_at_5 | 86.461 |
Retrieval | MTEB DuRetrieval | precision_at_1 | 91.75 |
Retrieval | MTEB DuRetrieval | precision_at_10 | 42.05 |
Retrieval | MTEB DuRetrieval | precision_at_100 | 4.827 |
Retrieval | MTEB DuRetrieval | precision_at_1000 | 0.48900000000000005 |
Retrieval | MTEB DuRetrieval | precision_at_3 | 78.55 |
Retrieval | MTEB DuRetrieval | precision_at_5 | 65.82000000000001 |
Retrieval | MTEB DuRetrieval | recall_at_1 | 26.924 |
Retrieval | MTEB DuRetrieval | recall_at_10 | 89.338 |
Retrieval | MTEB DuRetrieval | recall_at_100 | 97.856 |
Retrieval | MTEB DuRetrieval | recall_at_1000 | 99.11 |
Retrieval | MTEB DuRetrieval | recall_at_3 | 59.202999999999996 |
Retrieval | MTEB DuRetrieval | recall_at_5 | 75.642 |
Retrieval | MTEB EcomRetrieval | map_at_1 | 54.800000000000004 |
Retrieval | MTEB EcomRetrieval | map_at_10 | 65.613 |
Retrieval | MTEB EcomRetrieval | map_at_100 | 66.185 |
Retrieval | MTEB EcomRetrieval | map_at_1000 | 66.191 |
Retrieval | MTEB EcomRetrieval | map_at_3 | 62.8 |
Retrieval | MTEB EcomRetrieval | map_at_5 | 64.535 |
Retrieval | MTEB EcomRetrieval | mrr_at_1 | 54.800000000000004 |
Retrieval | MTEB EcomRetrieval | mrr_at_10 | 65.613 |
Retrieval | MTEB EcomRetrieval | mrr_at_100 | 66.185 |
Retrieval | MTEB EcomRetrieval | mrr_at_1000 | 66.191 |
Retrieval | MTEB EcomRetrieval | mrr_at_3 | 62.8 |
Retrieval | MTEB EcomRetrieval | mrr_at_5 | 64.535 |
Retrieval | MTEB EcomRetrieval | ndcg_at_1 | 54.800000000000004 |
Retrieval | MTEB EcomRetrieval | ndcg_at_10 | 70.991 |
Retrieval | MTEB EcomRetrieval | ndcg_at_100 | 73.434 |
Retrieval | MTEB EcomRetrieval | ndcg_at_1000 | 73.587 |
Retrieval | MTEB EcomRetrieval | ndcg_at_3 | 65.324 |
Retrieval | MTEB EcomRetrieval | ndcg_at_5 | 68.431 |
Retrieval | MTEB EcomRetrieval | precision_at_1 | 54.800000000000004 |
Retrieval | MTEB EcomRetrieval | precision_at_10 | 8.790000000000001 |
Retrieval | MTEB EcomRetrieval | precision_at_100 | 0.9860000000000001 |
Retrieval | MTEB EcomRetrieval | precision_at_1000 | 0.1 |
Retrieval | MTEB EcomRetrieval | precision_at_3 | 24.2 |
Retrieval | MTEB EcomRetrieval | precision_at_5 | 16.02 |
Retrieval | MTEB EcomRetrieval | recall_at_1 | 54.800000000000004 |
Retrieval | MTEB EcomRetrieval | recall_at_10 | 87.9 |
Retrieval | MTEB EcomRetrieval | recall_at_100 | 98.6 |
Retrieval | MTEB EcomRetrieval | recall_at_1000 | 99.8 |
Retrieval | MTEB EcomRetrieval | recall_at_3 | 72.6 |
Retrieval | MTEB EcomRetrieval | recall_at_5 | 80.10000000000001 |
Classification | MTEB IFlyTek | accuracy | 51.94305502116199 |
Classification | MTEB IFlyTek | f1 | 39.82197338426721 |
Classification | MTEB JDReview | accuracy | 90.31894934333957 |
Classification | MTEB JDReview | ap | 63.89821836499594 |
Classification | MTEB JDReview | f1 | 85.93687177603624 |
STS | MTEB LCQMC | cos_sim_pearson | 73.18906216730208 |
STS | MTEB LCQMC | cos_sim_spearman | 79.44570226735877 |
STS | MTEB LCQMC | euclidean_pearson | 78.8105072242798 |
STS | MTEB LCQMC | euclidean_spearman | 79.15605680863212 |
STS | MTEB LCQMC | manhattan_pearson | 78.80576507484064 |
STS | MTEB LCQMC | manhattan_spearman | 79.14625534068364 |
Reranking | MTEB MMarcoReranking | map | 41.58107192600853 |
Reranking | MTEB MMarcoReranking | mrr | 41.37063492063492 |
Retrieval | MTEB MMarcoRetrieval | map_at_1 | 68.33 |
Retrieval | MTEB MMarcoRetrieval | map_at_10 | 78.261 |
Retrieval | MTEB MMarcoRetrieval | map_at_100 | 78.522 |
Retrieval | MTEB MMarcoRetrieval | map_at_1000 | 78.527 |
Retrieval | MTEB MMarcoRetrieval | map_at_3 | 76.236 |
Retrieval | MTEB MMarcoRetrieval | map_at_5 | 77.557 |
Retrieval | MTEB MMarcoRetrieval | mrr_at_1 | 70.602 |
Retrieval | MTEB MMarcoRetrieval | mrr_at_10 | 78.779 |
Retrieval | MTEB MMarcoRetrieval | mrr_at_100 | 79.00500000000001 |
Retrieval | MTEB MMarcoRetrieval | mrr_at_1000 | 79.01 |
Retrieval | MTEB MMarcoRetrieval | mrr_at_3 | 77.037 |
Retrieval | MTEB MMarcoRetrieval | mrr_at_5 | 78.157 |
Retrieval | MTEB MMarcoRetrieval | ndcg_at_1 | 70.602 |
Retrieval | MTEB MMarcoRetrieval | ndcg_at_10 | 82.254 |
Retrieval | MTEB MMarcoRetrieval | ndcg_at_100 | 83.319 |
Retrieval | MTEB MMarcoRetrieval | ndcg_at_1000 | 83.449 |
Retrieval | MTEB MMarcoRetrieval | ndcg_at_3 | 78.46 |
Retrieval | MTEB MMarcoRetrieval | ndcg_at_5 | 80.679 |
Retrieval | MTEB MMarcoRetrieval | precision_at_1 | 70.602 |
Retrieval | MTEB MMarcoRetrieval | precision_at_10 | 9.989 |
Retrieval | MTEB MMarcoRetrieval | precision_at_100 | 1.05 |
Retrieval | MTEB MMarcoRetrieval | precision_at_1000 | 0.106 |
Retrieval | MTEB MMarcoRetrieval | precision_at_3 | 29.598999999999997 |
Retrieval | MTEB MMarcoRetrieval | precision_at_5 | 18.948 |
Retrieval | MTEB MMarcoRetrieval | recall_at_1 | 68.33 |
Retrieval | MTEB MMarcoRetrieval | recall_at_10 | 94.00800000000001 |
Retrieval | MTEB MMarcoRetrieval | recall_at_100 | 98.589 |
Retrieval | MTEB MMarcoRetrieval | recall_at_1000 | 99.60799999999999 |
Retrieval | MTEB MMarcoRetrieval | recall_at_3 | 84.057 |
Retrieval | MTEB MMarcoRetrieval | recall_at_5 | 89.32900000000001 |
Classification | MTEB MassiveIntentClassification (zh-CN) | accuracy | 78.13718897108272 |
Classification | MTEB MassiveIntentClassification (zh-CN) | f1 | 74.07613180855328 |
Classification | MTEB MassiveScenarioClassification (zh-CN) | accuracy | 86.20040349697376 |
Classification | MTEB MassiveScenarioClassification (zh-CN) | f1 | 85.05282136519973 |
Retrieval | MTEB MedicalRetrieval | map_at_1 | 56.8 |
Retrieval | MTEB MedicalRetrieval | map_at_10 | 64.199 |
Retrieval | MTEB MedicalRetrieval | map_at_100 | 64.89 |
Retrieval | MTEB MedicalRetrieval | map_at_1000 | 64.917 |
Retrieval | MTEB MedicalRetrieval | map_at_3 | 62.383 |
Retrieval | MTEB MedicalRetrieval | map_at_5 | 63.378 |
Retrieval | MTEB MedicalRetrieval | mrr_at_1 | 56.8 |
Retrieval | MTEB MedicalRetrieval | mrr_at_10 | 64.199 |
Retrieval | MTEB MedicalRetrieval | mrr_at_100 | 64.89 |
Retrieval | MTEB MedicalRetrieval | mrr_at_1000 | 64.917 |
Retrieval | MTEB MedicalRetrieval | mrr_at_3 | 62.383 |
Retrieval | MTEB MedicalRetrieval | mrr_at_5 | 63.378 |
Retrieval | MTEB MedicalRetrieval | ndcg_at_1 | 56.8 |
Retrieval | MTEB MedicalRetrieval | ndcg_at_10 | 67.944 |
Retrieval | MTEB MedicalRetrieval | ndcg_at_100 | 71.286 |
Retrieval | MTEB MedicalRetrieval | ndcg_at_1000 | 71.879 |
Retrieval | MTEB MedicalRetrieval | ndcg_at_3 | 64.163 |
Retrieval | MTEB MedicalRetrieval | ndcg_at_5 | 65.96600000000001 |
Retrieval | MTEB MedicalRetrieval | precision_at_1 | 56.8 |
Retrieval | MTEB MedicalRetrieval | precision_at_10 | 7.9799999999999995 |
Retrieval | MTEB MedicalRetrieval | precision_at_100 | 0.954 |
Retrieval | MTEB MedicalRetrieval | precision_at_1000 | 0.1 |
Retrieval | MTEB MedicalRetrieval | precision_at_3 | 23.1 |
Retrieval | MTEB MedicalRetrieval | precision_at_5 | 14.74 |
Retrieval | MTEB MedicalRetrieval | recall_at_1 | 56.8 |
Retrieval | MTEB MedicalRetrieval | recall_at_10 | 79.80000000000001 |
Retrieval | MTEB MedicalRetrieval | recall_at_100 | 95.39999999999999 |
Retrieval | MTEB MedicalRetrieval | recall_at_1000 | 99.8 |
Retrieval | MTEB MedicalRetrieval | recall_at_3 | 69.3 |
Retrieval | MTEB MedicalRetrieval | recall_at_5 | 73.7 |
Classification | MTEB MultilingualSentiment | accuracy | 78.57666666666667 |
Classification | MTEB MultilingualSentiment | f1 | 78.23373528202681 |
PairClassification | MTEB Ocnli | cos_sim_accuracy | 85.43584190579317 |
PairClassification | MTEB Ocnli | cos_sim_ap | 90.76665640338129 |
PairClassification | MTEB Ocnli | cos_sim_f1 | 86.5021770682148 |
PairClassification | MTEB Ocnli | cos_sim_precision | 79.82142857142858 |
PairClassification | MTEB Ocnli | cos_sim_recall | 94.40337909186906 |
PairClassification | MTEB Ocnli | dot_accuracy | 78.66811044937737 |
PairClassification | MTEB Ocnli | dot_ap | 85.84084363880804 |
PairClassification | MTEB Ocnli | dot_f1 | 80.10075566750629 |
PairClassification | MTEB Ocnli | dot_precision | 76.58959537572254 |
PairClassification | MTEB Ocnli | dot_recall | 83.9493136219641 |
PairClassification | MTEB Ocnli | euclidean_accuracy | 84.46128857606931 |
PairClassification | MTEB Ocnli | euclidean_ap | 88.62351100230491 |
PairClassification | MTEB Ocnli | euclidean_f1 | 85.7709469509172 |
PairClassification | MTEB Ocnli | euclidean_precision | 80.8411214953271 |
PairClassification | MTEB Ocnli | euclidean_recall | 91.34107708553326 |
PairClassification | MTEB Ocnli | manhattan_accuracy | 84.51543042772063 |
PairClassification | MTEB Ocnli | manhattan_ap | 88.53975607870393 |
PairClassification | MTEB Ocnli | manhattan_f1 | 85.75697211155378 |
PairClassification | MTEB Ocnli | manhattan_precision | 81.14985862393968 |
PairClassification | MTEB Ocnli | manhattan_recall | 90.91869060190075 |
PairClassification | MTEB Ocnli | max_accuracy | 85.43584190579317 |
PairClassification | MTEB Ocnli | max_ap | 90.76665640338129 |
PairClassification | MTEB Ocnli | max_f1 | 86.5021770682148 |
Classification | MTEB OnlineShopping | accuracy | 95.06999999999998 |
Classification | MTEB OnlineShopping | ap | 93.45104559324996 |
Classification | MTEB OnlineShopping | f1 | 95.06036329426092 |
STS | MTEB PAWSX | cos_sim_pearson | 40.01998290519605 |
STS | MTEB PAWSX | cos_sim_spearman | 46.5989769986853 |
STS | MTEB PAWSX | euclidean_pearson | 45.37905883182924 |
STS | MTEB PAWSX | euclidean_spearman | 46.22213849806378 |
STS | MTEB PAWSX | manhattan_pearson | 45.40925124776211 |
STS | MTEB PAWSX | manhattan_spearman | 46.250705124226386 |
STS | MTEB QBQTC | cos_sim_pearson | 42.719516197112526 |
STS | MTEB QBQTC | cos_sim_spearman | 44.57507789581106 |
STS | MTEB QBQTC | euclidean_pearson | 35.73062264160721 |
STS | MTEB QBQTC | euclidean_spearman | 40.473523909913695 |
STS | MTEB QBQTC | manhattan_pearson | 35.69868964086357 |
STS | MTEB QBQTC | manhattan_spearman | 40.46349925372903 |
STS | MTEB STS22 (zh) | cos_sim_pearson | 62.340118285801104 |
STS | MTEB STS22 (zh) | cos_sim_spearman | 67.72781908620632 |
STS | MTEB STS22 (zh) | euclidean_pearson | 63.161965746091596 |
STS | MTEB STS22 (zh) | euclidean_spearman | 67.36825684340769 |
STS | MTEB STS22 (zh) | manhattan_pearson | 63.089863788261425 |
STS | MTEB STS22 (zh) | manhattan_spearman | 67.40868898995384 |
STS | MTEB STSB | cos_sim_pearson | 79.1646360962365 |
STS | MTEB STSB | cos_sim_spearman | 81.24426700767087 |
STS | MTEB STSB | euclidean_pearson | 79.43826409936123 |
STS | MTEB STSB | euclidean_spearman | 79.71787965300125 |
STS | MTEB STSB | manhattan_pearson | 79.43377784961737 |
STS | MTEB STSB | manhattan_spearman | 79.69348376886967 |
Reranking | MTEB T2Reranking | map | 68.35595092507496 |
Reranking | MTEB T2Reranking | mrr | 79.00244892585788 |
Retrieval | MTEB T2Retrieval | map_at_1 | 26.588 |
Retrieval | MTEB T2Retrieval | map_at_10 | 75.327 |
Retrieval | MTEB T2Retrieval | map_at_100 | 79.095 |
Retrieval | MTEB T2Retrieval | map_at_1000 | 79.163 |
Retrieval | MTEB T2Retrieval | map_at_3 | 52.637 |
Retrieval | MTEB T2Retrieval | map_at_5 | 64.802 |
Retrieval | MTEB T2Retrieval | mrr_at_1 | 88.103 |
Retrieval | MTEB T2Retrieval | mrr_at_10 | 91.29899999999999 |
Retrieval | MTEB T2Retrieval | mrr_at_100 | 91.408 |
Retrieval | MTEB T2Retrieval | mrr_at_1000 | 91.411 |
Retrieval | MTEB T2Retrieval | mrr_at_3 | 90.801 |
Retrieval | MTEB T2Retrieval | mrr_at_5 | 91.12700000000001 |
Retrieval | MTEB T2Retrieval | ndcg_at_1 | 88.103 |
Retrieval | MTEB T2Retrieval | ndcg_at_10 | 83.314 |
Retrieval | MTEB T2Retrieval | ndcg_at_100 | 87.201 |
Retrieval | MTEB T2Retrieval | ndcg_at_1000 | 87.83999999999999 |
Retrieval | MTEB T2Retrieval | ndcg_at_3 | 84.408 |
Retrieval | MTEB T2Retrieval | ndcg_at_5 | 83.078 |
Retrieval | MTEB T2Retrieval | precision_at_1 | 88.103 |
Retrieval | MTEB T2Retrieval | precision_at_10 | 41.638999999999996 |
Retrieval | MTEB T2Retrieval | precision_at_100 | 5.006 |
Retrieval | MTEB T2Retrieval | precision_at_1000 | 0.516 |
Retrieval | MTEB T2Retrieval | precision_at_3 | 73.942 |
Retrieval | MTEB T2Retrieval | precision_at_5 | 62.056 |
Retrieval | MTEB T2Retrieval | recall_at_1 | 26.588 |
Retrieval | MTEB T2Retrieval | recall_at_10 | 82.819 |
Retrieval | MTEB T2Retrieval | recall_at_100 | 95.334 |
Retrieval | MTEB T2Retrieval | recall_at_1000 | 98.51299999999999 |
Retrieval | MTEB T2Retrieval | recall_at_3 | 54.74 |
Retrieval | MTEB T2Retrieval | recall_at_5 | 68.864 |
Classification | MTEB TNews | accuracy | 55.029 |
Classification | MTEB TNews | f1 | 53.043617905026764 |
Clustering | MTEB ThuNewsClusteringP2P | v_measure | 77.83675116835911 |
Clustering | MTEB ThuNewsClusteringS2S | v_measure | 74.19701455865277 |
Retrieval | MTEB VideoRetrieval | map_at_1 | 64.7 |
Retrieval | MTEB VideoRetrieval | map_at_10 | 75.593 |
Retrieval | MTEB VideoRetrieval | map_at_100 | 75.863 |
Retrieval | MTEB VideoRetrieval | map_at_1000 | 75.863 |
Retrieval | MTEB VideoRetrieval | map_at_3 | 73.63300000000001 |
Retrieval | MTEB VideoRetrieval | map_at_5 | 74.923 |
Retrieval | MTEB VideoRetrieval | mrr_at_1 | 64.7 |
Retrieval | MTEB VideoRetrieval | mrr_at_10 | 75.593 |
Retrieval | MTEB VideoRetrieval | mrr_at_100 | 75.863 |
Retrieval | MTEB VideoRetrieval | mrr_at_1000 | 75.863 |
Retrieval | MTEB VideoRetrieval | mrr_at_3 | 73.63300000000001 |
Retrieval | MTEB VideoRetrieval | mrr_at_5 | 74.923 |
Retrieval | MTEB VideoRetrieval | ndcg_at_1 | 64.7 |
Retrieval | MTEB VideoRetrieval | ndcg_at_10 | 80.399 |
Retrieval | MTEB VideoRetrieval | ndcg_at_100 | 81.517 |
Retrieval | MTEB VideoRetrieval | ndcg_at_1000 | 81.517 |
Retrieval | MTEB VideoRetrieval | ndcg_at_3 | 76.504 |
Retrieval | MTEB VideoRetrieval | ndcg_at_5 | 78.79899999999999 |
Retrieval | MTEB VideoRetrieval | precision_at_1 | 64.7 |
Retrieval | MTEB VideoRetrieval | precision_at_10 | 9.520000000000001 |
Retrieval | MTEB VideoRetrieval | precision_at_100 | 1 |
Retrieval | MTEB VideoRetrieval | precision_at_1000 | 0.1 |
Retrieval | MTEB VideoRetrieval | precision_at_3 | 28.266999999999996 |
Retrieval | MTEB VideoRetrieval | precision_at_5 | 18.060000000000002 |
Retrieval | MTEB VideoRetrieval | recall_at_1 | 64.7 |
Retrieval | MTEB VideoRetrieval | recall_at_10 | 95.19999999999999 |
Retrieval | MTEB VideoRetrieval | recall_at_100 | 100 |
Retrieval | MTEB VideoRetrieval | recall_at_1000 | 100 |
Retrieval | MTEB VideoRetrieval | recall_at_3 | 84.8 |
Retrieval | MTEB VideoRetrieval | recall_at_5 | 90.3 |
Classification | MTEB Waimai | accuracy | 89.69999999999999 |
Classification | MTEB Waimai | ap | 75.91371640164184 |
Classification | MTEB Waimai | f1 | 88.34067777698694 |
📄 許可證
本模型採用 CC BY-NC 4.0
許可證。
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