alime-embedding-large-zh
The alime embedding model.
Usage (Sentence-Transformers)
Using this model becomes easy when you have sentence-transformers installed:
pip install -U sentence-transformers
Then you can use the model like this:
from sentence_transformers import SentenceTransformer
sentences = ["西湖在哪?", "西湖风景名胜区位于浙江省杭州市"]
model = SentenceTransformer('Pristinenlp/alime-embedding-large-zh')
embeddings = model.encode(sentences, normalize_embeddings=True)
print(embeddings)
- Downloads last month
- 704
Evaluation results
- cos_sim_pearson on MTEB AFQMCvalidation set self-reported49.648
- cos_sim_spearman on MTEB AFQMCvalidation set self-reported54.733
- euclidean_pearson on MTEB AFQMCvalidation set self-reported53.063
- euclidean_spearman on MTEB AFQMCvalidation set self-reported54.733
- manhattan_pearson on MTEB AFQMCvalidation set self-reported53.048
- manhattan_spearman on MTEB AFQMCvalidation set self-reported54.729
- cos_sim_pearson on MTEB ATECtest set self-reported48.659
- cos_sim_spearman on MTEB ATECtest set self-reported55.125
- euclidean_pearson on MTEB ATECtest set self-reported55.734
- euclidean_spearman on MTEB ATECtest set self-reported55.125
- manhattan_pearson on MTEB ATECtest set self-reported55.712
- manhattan_spearman on MTEB ATECtest set self-reported55.122
- accuracy on MTEB AmazonReviewsClassification (zh)test set self-reported46.950
- f1 on MTEB AmazonReviewsClassification (zh)test set self-reported45.344
- cos_sim_pearson on MTEB BQtest set self-reported62.927
- cos_sim_spearman on MTEB BQtest set self-reported64.888
- euclidean_pearson on MTEB BQtest set self-reported63.314
- euclidean_spearman on MTEB BQtest set self-reported64.888
- manhattan_pearson on MTEB BQtest set self-reported63.222
- manhattan_spearman on MTEB BQtest set self-reported64.798
- v_measure on MTEB CLSClusteringP2Ptest set self-reported42.519
- v_measure on MTEB CLSClusteringS2Stest set self-reported39.729
- map on MTEB CMedQAv1test set self-reported86.519
- mrr on MTEB CMedQAv1test set self-reported89.020
- map on MTEB CMedQAv2test set self-reported87.114
- mrr on MTEB CMedQAv2test set self-reported89.634
- map_at_1 on MTEB CmedqaRetrievalself-reported24.883
- map_at_10 on MTEB CmedqaRetrievalself-reported37.246
- map_at_100 on MTEB CmedqaRetrievalself-reported39.110
- map_at_1000 on MTEB CmedqaRetrievalself-reported39.222
- map_at_3 on MTEB CmedqaRetrievalself-reported32.956
- map_at_5 on MTEB CmedqaRetrievalself-reported35.411
- mrr_at_1 on MTEB CmedqaRetrievalself-reported37.834
- mrr_at_10 on MTEB CmedqaRetrievalself-reported46.031
- mrr_at_100 on MTEB CmedqaRetrievalself-reported47.033
- mrr_at_1000 on MTEB CmedqaRetrievalself-reported47.077
- mrr_at_3 on MTEB CmedqaRetrievalself-reported43.415
- mrr_at_5 on MTEB CmedqaRetrievalself-reported44.938
- ndcg_at_1 on MTEB CmedqaRetrievalself-reported37.834
- ndcg_at_10 on MTEB CmedqaRetrievalself-reported43.928
- ndcg_at_100 on MTEB CmedqaRetrievalself-reported51.313
- ndcg_at_1000 on MTEB CmedqaRetrievalself-reported53.230
- ndcg_at_3 on MTEB CmedqaRetrievalself-reported38.397
- ndcg_at_5 on MTEB CmedqaRetrievalself-reported40.848
- precision_at_1 on MTEB CmedqaRetrievalself-reported37.834
- precision_at_10 on MTEB CmedqaRetrievalself-reported9.782
- precision_at_100 on MTEB CmedqaRetrievalself-reported1.583
- precision_at_1000 on MTEB CmedqaRetrievalself-reported0.183
- precision_at_3 on MTEB CmedqaRetrievalself-reported21.664
- precision_at_5 on MTEB CmedqaRetrievalself-reported15.934
- recall_at_1 on MTEB CmedqaRetrievalself-reported24.883
- recall_at_10 on MTEB CmedqaRetrievalself-reported54.911
- recall_at_100 on MTEB CmedqaRetrievalself-reported85.419
- recall_at_1000 on MTEB CmedqaRetrievalself-reported98.160
- recall_at_3 on MTEB CmedqaRetrievalself-reported38.416
- recall_at_5 on MTEB CmedqaRetrievalself-reported45.778
- cos_sim_accuracy on MTEB Cmnlivalidation set self-reported82.562
- cos_sim_ap on MTEB Cmnlivalidation set self-reported89.361
- cos_sim_f1 on MTEB Cmnlivalidation set self-reported83.934
- cos_sim_precision on MTEB Cmnlivalidation set self-reported79.424
- cos_sim_recall on MTEB Cmnlivalidation set self-reported88.988
- dot_accuracy on MTEB Cmnlivalidation set self-reported82.562
- dot_ap on MTEB Cmnlivalidation set self-reported89.382
- dot_f1 on MTEB Cmnlivalidation set self-reported83.934
- dot_precision on MTEB Cmnlivalidation set self-reported79.424
- dot_recall on MTEB Cmnlivalidation set self-reported88.988
- euclidean_accuracy on MTEB Cmnlivalidation set self-reported82.562
- euclidean_ap on MTEB Cmnlivalidation set self-reported89.361
- euclidean_f1 on MTEB Cmnlivalidation set self-reported83.934
- euclidean_precision on MTEB Cmnlivalidation set self-reported79.424
- euclidean_recall on MTEB Cmnlivalidation set self-reported88.988
- manhattan_accuracy on MTEB Cmnlivalidation set self-reported82.429
- manhattan_ap on MTEB Cmnlivalidation set self-reported89.366
- manhattan_f1 on MTEB Cmnlivalidation set self-reported83.948
- manhattan_precision on MTEB Cmnlivalidation set self-reported78.663
- manhattan_recall on MTEB Cmnlivalidation set self-reported89.993
- max_accuracy on MTEB Cmnlivalidation set self-reported82.562
- max_ap on MTEB Cmnlivalidation set self-reported89.382
- max_f1 on MTEB Cmnlivalidation set self-reported83.948