| #import rapidfuzz | |
| import pandas as pd | |
| from sentence_transformers import SentenceTransformer | |
| cmp_model = SentenceTransformer('uer/sbert-base-chinese-nli') | |
| def compare(text0, text1): | |
| ''' | |
| val = rapidfuzz.fuzz.ratio(text0, text1) | |
| if val >= 75: | |
| return val | |
| ''' | |
| embeddings = cmp_model.encode([text0, text1]) | |
| return int(pd.DataFrame(embeddings.T).corr().iloc[0, 1] * 100) |