modernbert base go emotions EU AI Act Compliance Profile

ModernBERT Go Emotions 2026-03

Note: This version significantly improves the baseline 2025 model.

Overview

This model was fine-tuned from ModernBERT-base on the GoEmotions dataset for multi-label classification. It predicts emotional states in text, with a total of 28 possible labels. Each input text can have one or more associated labels, reflecting the multi-label nature of the task.

Try it out here.

Intended Use

The model is designed for emotion classification in English-language text, particularly in domains such as:

• Social media sentiment analysis
• Customer feedback evaluation
• Behavioral or psychological research

Example Usage

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

from transformers import pipeline
import torch

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

text = "I am so happy and excited about this opportunity!"
predictions = classifier(text)

# Print top 5 detected emotions
print("\nTop 5 emotions detected:")
for pred in predictions[0]:  # Note: predictions is [[...]] format
    print(f"\t{pred['label']:10s} : {pred['score']:.3f}")

# Example output
# Top 5 emotions detected:
#        excitement : 0.976
#        joy        : 0.698
#        approval   : 0.012
#        optimism   : 0.007
#        admiration : 0.006

Model Details

• Base Model: ModernBERT-base
• Fine-Tuning Dataset: GoEmotions
• Number of Labels: 28
• **Pro

Your obligations as a modernbert base go emotions deployer

Track your progress

Sign up to track obligations, receive deadline reminders, and generate audit packages.

Tags

Languages

Library

Base Model