Skip to content

DeepPavlov/bert-base-multilingual-cased-sentence

DeepPavlov/bert-base-multilingual-cased-sentence is machine learning model.

About DeepPavlov/bert-base-multilingual-cased-sentence

Sentence Multilingual BERT (101 languages) is a representation‑based sentence encoder for 101 languages . Sentence representations are mean pooled token embeddings in the same manner as in Sentence‑BERT[3] Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks. It is initialized with Multilingual BERT and then fine‑tuned on english MultiNLI[1] and on dev set of multilingual XNLI. Sentence representation based sentence-based sentence encoder . It is then fine-tuned for English, Spanish, Arabic, Arabic and Arabic languages to suit each language .,
View model source

Explore

FAQ