1from sentence_transformers import SentenceTransformer
2
3model = SentenceTransformer('ThanhLe0125/e5-base-math')
4
5# Encode queries (add prefix for better performance)
6queries = ["query: Định nghĩa hàm số đồng biến là gì?"]
7query_embeddings = model.encode(queries)
8
9# Encode passages/documents
10passages = ["passage: Hàm số đồng biến trên khoảng (a;b) là hàm số mà với mọi x1 < x2 thì f(x1) < f(x2)"]
11passage_embeddings = model.encode(passages)
12
13# Calculate similarity
14from sklearn.metrics.pairwise import cosine_similarity
15similarity = cosine_similarity(query_embeddings, passage_embeddings)
1# Recommended usage for RAG
2def encode_query(query_text):
3 return model.encode([f"query: {query_text}"])
4
5def encode_passage(passage_text):
6 return model.encode([f"passage: {passage_text}"])
7
8# Example usage
9query_emb = encode_query("Định nghĩa hàm số đồng biến")
10passage_emb = encode_passage("Hàm số đồng biến là...")
11
12# Calculate similarity
13similarity = cosine_similarity(query_emb, passage_emb)[0][0]
14print(f"Similarity: {similarity:.4f}")
This model has been fine-tuned specifically for Vietnamese mathematical content and should perform better than the base model for math-related queries in Vietnamese.
1@misc{e5-base-math,
2 author = {ThanhLe},
3 title = {E5-Base-Math: Fine-tuned Vietnamese Math Embedding Model},
4 year = {2025},
5 publisher = {Hugging Face},
6 howpublished = {\url{https://huggingface.co/ThanhLe0125/e5-base-math}}
7}
For questions or issues, please contact via the repository discussions.
