metadata
language: en
tags:
- emotion-detection
- text-classification
- transformers
- distilbert
datasets:
- nelgiriyewithana/emotions
model-index:
- name: DistilBERT Emotion Classifier
results:
- task:
type: text-classification
name: Emotion Detection
dataset:
name: Kaggle Emotions Dataset
type: text
metrics:
- name: Accuracy
type: accuracy
value: 0.94
- name: F1
type: f1
value: 0.94
- name: Precision
type: precision
value: 0.94
- name: Recall
type: recall
value: 0.94
license: apache-2.0
metrics:
- accuracy
- recall
- precision
DistilBERT Emotion Classifier 🎭
This model classifies English text into one of six emotions: sadness, joy, love, anger, fear, surprise.
- Base model:
distilbert-base-uncased - Framework: Hugging Face Transformers
- Dataset: Kaggle Emotions Dataset
- Task: Multi-class emotion detection
📊 Evaluation
| Class | Precision | Recall | F1-score | Support |
|---|---|---|---|---|
| 0 (sadness) | 0.99 | 0.96 | 0.98 | 24,121 |
| 1 (joy) | 0.93 | 0.99 | 0.96 | 28,220 |
| 2 (love) | 1.00 | 0.71 | 0.83 | 6,824 |
| 3 (anger) | 0.95 | 0.94 | 0.95 | 11,448 |
| 4 (fear) | 0.90 | 0.91 | 0.91 | 9,574 |
| 5 (surprise) | 0.74 | 0.99 | 0.85 | 3,038 |
Overall Performance:
- Accuracy: 94%
- Macro F1: 0.91
- Weighted F1: 0.94
🧑💻 Usage
from transformers import AutoTokenizer, AutoModelForSequenceClassification, pipeline
model_name = "YamenRM/distilbert-emotion-classifier"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name)
nlp = pipeline("text-classification", model=model, tokenizer=tokenizer)
print(nlp("I feel so happy and excited today!"))
# [{'label': 'joy', 'score': 0.98}]