# Multi-source Data Training

A Korean reranking model fine-tuned from Alibaba-NLP/gte-multilingual-reranker-base, optimized for text relevance in search and Q&A scenarios

This is a large Transformer model with 780 million parameters, specifically designed for music generation and transcription tasks, using anticipatory training methods.

This is a BERT model specifically designed for Named Entity Recognition (NER) in the field of nutrition labels, used to detect and classify different nutritional components.

Text classification model based on ELECTRA architecture for detecting expressions of suicidal tendencies in text

A sentence-transformers model developed by Hum, supporting 4096 tokens long text embedding, suitable for semantic search and clustering tasks

A text classification model based on the Roberta architecture, used to determine the truthfulness of assertions based on evidence.

A small Japanese language model based on GPT-NeoX architecture, supporting text generation tasks

Chinese speech recognition model fine-tuned based on XLSR-53 large model, supporting 16kHz sampled audio input

Swedish BERT base model released by the National Library of Sweden/KBLab, trained on multi-source texts

A sentence embedding model based on DistilBERT, optimized for question-answer similarity tasks, fine-tuned on various QA datasets through contrastive learning

A BERT model fine-tuned for Persian Named Entity Recognition (NER) tasks, supporting 10 entity categories