Downloads · 30 days
4
8% of all-time downloads
moodshift-roberta-esa-tfidf-focal
Sarjinkhan2003/moodshift-roberta-esa-tfidf-focal
moodshift-roberta-esa-tfidf-focal is a text classification model from Sarjinkhan2003. Use it when you need a label for a piece of text.
Test Accuracy: 0.9230 | Macro F1: 0.8903 Architecture: Full RoBERTa-base + Emotion-Specific Attention + TF-IDF Gate + Focal Loss ICCA 2026 HCI Research
All-time downloads
49
Public
Repo size
511 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin511 MB · 99%
At a glance
- Task
- Text Classification
- Model type
- emotion_aware_roberta_esa_tfidf_focal
- Access
- Public
- Created
- Apr 30, 2026
- Updated
- Apr 30, 2026
- SHA
- d6dc84b6
Try a prompt
Datasets
Model card
From the Hugging Face model README
MoodShift — RoBERTa+ESA+TF-IDF+Focal Loss
Test Accuracy: 0.9230 | Macro F1: 0.8903 Architecture: Full RoBERTa-base + Emotion-Specific Attention + TF-IDF Gate + Focal Loss ICCA 2026 HCI Research
- Task
- Text Classification
- Type
- emotion_aware_roberta_esa_tfidf_focal
- Created
- Apr 30, 2026
- Updated
- Apr 30, 2026