fhamborg / NewsMTSC

Target-dependent sentiment classification in news articles reporting on political events. Includes a high-quality data set of over 11k sentences and a state-of-the-art classification model.
Other
140 stars 21 forks source link

Raise requirements #36

Closed t1h0 closed 11 months ago

t1h0 commented 11 months ago

Raise upper bounds to python (<3.12) and torch (<2.1). Successfully tested train and infer.