Illustrative cover: Machine Elf 1735 · Public domain · Resized and converted to WebP.
Created by ADHAYAN cofounder Shaidozzaman Araf.
Recognizing emotion in English text
SATA stands for Sentiment Analyzer Trained by Araf. Its model card describes fine-tuning roberta-base on the Hugging Face Emotion dataset using a custom PyTorch training loop.
The model predicts sadness, joy, love, anger, fear and surprise. The repository provides model weights, tokenizer assets and configuration, while the model card documents loading it with the Transformers text-classification pipeline.
The cover is an illustrative emotion wheel; it is not the model’s output or label diagram.
Contributors
- Shaidozzaman Araf
Resources & links