Model Trained Using AutoTrain

Problem type: Text Classification

Validation Metrics

loss: 0.39284127950668335

f1: 0.91

precision: 0.8921568627450981

recall: 0.9285714285714286

auc: 0.957133746355685

accuracy: 0.8831168831168831

Fine-tuned from

cardiffnlp/twitter-roberta-base-sentiment-latest

Trained on

Twitter_Non-Ads from the paper "Stress Detection from Social Media Articles: New Dataset Benchmark and Analytical Study" - 2022 International Joint Conference on Neural Networks (IJCNN)

All credit relating to the dataset goes to authors Aryan Rastogi; Qian Liu; Erik Cambria

Access to dataset: https://github.com/aryan-r22/Stress-Detection_Social-Media-Articles/tree/main