Model Card for GoEmotions Model BERT
This model is a fine-tuned DistilBERT model for emotion detection using the simplified GoEmotions dataset provided by Google.
It is capable of classifying English text into multiple emotions from the GoEmotions label set.
Model Details
Model Description
This is a fine-tuned distilbert-base-uncased model on the simplified version of the GoEmotions dataset for multi-label emotion classification.
- Developed by: Shahzad Sohail
- Funded by [optional]: Self-funded
- Shared by [optional]: Shahzad Sohail
- Model type: DistilBERT (fine-tuned for multi-label classification)
- Language(s) (NLP): English
- License: Apache 2.0
- Finetuned from model [optional]: distilbert-base-uncased
Model Sources
- Repository: [Optional: Add link if you have a GitHub repository]
- Dataset: https://huggingface.co/datasets/google-research-datasets/go_emotions
- Demo [optional]: [Add Gradio/Hugging Face Space if applicable]
Uses
Direct Use
This model can be directly used for:
- Emotion classification in social media comments
- Sentiment analysis with fine-grained emotions
- Enhancing chatbot emotional understanding
- Social listening and brand monitoring
Downstream Use [optional]
The model can be further fine-tuned for specialized domains such as healthcare, education, or customer support.
Out-of-Scope Use
- Not suitable for clinical or psychological diagnosis.
- Not designed for non-English text.
- Should not be used in critical decision-making without human oversight.
Bias, Risks, and Limitations
The model inherits biases from the GoEmotions dataset:
- May underperform on non-standard English, slang, or low-resource dialects.
- May misclassify ambiguous or sarcastic text.
- Model's emotion predictions should be verified in sensitive contexts.
Recommendations
- Use the model with awareness of its biases and limitations.
- Apply additional human review in high-stakes or sensitive applications.
How to Get Started with the Model
You can load and use the model with the Hugging Face transformers pipeline:
from transformers import pipeline
classifier = pipeline("text-classification", model="Shahzad1/go_emotions_model_bert", top_k=None)
text = "I am feeling very happy today!"
results = classifier(text)
print(results)