mteb/

UrduRomanSentimentClassification

UrduRomanSentimentClassification An MTEB dataset Massive Text Embedding Benchmark

The Roman Urdu dataset is a data corpus comprising of more than 20000 records tagged for sentiment (Positive, Negative, Neutral)

Task category t2c

Domains Social, Written Reference https://archive.ics.uci.edu/dataset/458/roman+urdu+data+set

You can evaluate an embedding model on this dataset using the following code: import mteb

task =… See the full description on the dataset page: https://huggingface.co/datasets/mteb/UrduRomanSentimentClassification.