# Weakly Supervised Contrastive Learning

A locally sparse global version based on intfloat/multilingual-e5-small, supporting multilingual text embedding models with approximately 4k tokens

A multilingual text embedding model fine-tuned on XNLI and ANLI datasets based on intfloat/multilingual-e5-base, suitable for zero-shot classification tasks

A fine-tuned version of the multilingual-e5-large model on XNLI and ANLI datasets, supporting multilingual zero-shot classification tasks

A multilingual text classification model fine-tuned on the XNLI dataset based on multilingual-e5-large, supporting zero-shot classification in 15 languages

E5-base-multilingual-4096 is a locally sparse global version based on intfloat/multilingual-e5-base, supporting multilingual text embedding models that can process up to 4096 tokens.

The E5 Base Unsupervised Model is a text embedding model based on contrastive pre-training, suitable for sentence similarity and transformation tasks.

The unsupervised version of E5-small, generating text embeddings through weakly supervised contrastive pre-training, suitable for tasks like text similarity calculation