The dataset consists of 5500 English sentence pairs that are scored and ranked on a relatedness scale ranging from 0 (least related) to 1 (most related).
The sentence pairs, and associated scores, are in the file sem_text_rel_ranked.csv in the root directory. The CSV file can be read using:
import pandas as pd
row = str.loc[0]
sent1, sent2 = row['Text'].split("\n")
score =… See the full description on the dataset page:
https://huggingface.co/datasets/vkpriya/str-2022.
