💡 This model is part of the collection Frontier BERT Workhorses.

Overview

This model was fine-tuned from ModernBERT-base on the Super Emotion dataset for multi-class emotion classification. It predicts emotional states in text across seven labels: joy, sadness, anger, fear, love, neutral, surprise.

Model Details

Base Model: ModernBERT-base
Fine-Tuning Dataset: Super Emotion
Number of Labels: 7
Problem Type: Multi-label classification
Language: English
License: CC-BY-4.0
Fine-Tuning Framework: Hugging Face Transformers

Example Usage

Here’s how to use the model with Hugging Face Transformers:

from transformers import pipeline

# Load the model
classifier = pipeline(
    "text-classification", 
    model="cirimus/modernbert-base-emotions",
    top_k=5
)

text = "I can't believe this just happened!"
predictions = classifier(text)

# Print top 3 detected emotions
sorted_preds = sorted(predictions[0], key=lambda x: x['score'], reverse=True)
top_3 = sorted_preds[:3]

print("\nTop 3 emotions detected:")
for pred in top_3:
    print(f"\t{pred['label']:10s} : {pred['score']:.3f}")

# Example output:
# Top 3 emotions detected:
#        SURPRISE   : 0.843
#        FEAR       : 0.059
#        NEUTRAL    : 0.049

How the Model Was Created

The model was fine-tuned for 2 epochs using the following hyperparameters:

Learning Rate: 2e-5
Batch Size: 16
Weight Decay: 0.01
Warmup Steps: Cosine decay scheduling
Optimizer: AdamW
Evaluation Metrics: Precision, Recall, F1 Score (macro), Accuracy

Evaluation Results

As evaluated on the joint test-set:

	Accuracy	Precision	Recall	F1	MCC	Support
macro avg	0.967	0.879	0.802	0.836	0.818	51950
NEUTRAL	0.946	0.609	0.702	0.652	0.625	3734
SURPRISE	0.981	0.863	0.630	0.729	0.729	2064
FEAR	0.975	0.936	0.822	0.875	0.864	5526
SADNESS	0.968	0.988	0.888	0.936	0.916	13792
JOY	0.961	0.962	0.915	0.938	0.910	16623
ANGER	0.966	0.921	0.853	0.885	0.866	7993
LOVE	0.969	0.871	0.801	0.835	0.818	5071

Intended Use

The model is designed for emotion classification in English-language text, particularly useful for:

Social media sentiment analysis
Customer feedback evaluation
Large scale behavioral or psychological research

The model is designed for fast and accurate emotion detection but struggles with subtle expressions or indirect references to emotions (e.g., "I find myself remembering the little things you say, long after you've said them.")

Limitations and Biases

Data Bias: The dataset is aggregated from multiple sources and may contain biases in annotation and class distribution.
Underrepresented Classes: Some emotions have fewer samples, affecting their classification performance.
Context Dependence: The model classifies individual sentences and may not perform well on multi-sentence contexts.

Environmental Impact

Hardware Used: NVIDIA DGX Spark (GB10)
Training Time: ~5 hours
Carbon Emissions: ~0.28 kg CO2 (estimated via ML CO2 Impact Calculator)

Citation

If you use this model, please cite:

@misc{junque2025emotion,
  title        = {Emotion and Bias Classification with Modern Encoders},
  author       = {Junqu\'e de Fortuny, Enric},
  year         = {2025},
  howpublished = {\url{https://huggingface.co/spaces/cirimus/emotion-detection}}
}