techhy/mindtrack-mental-health-analyzer

MindTrack: Mental Health Sentiment Analyzer 🧠

Model Description

MindTrack is a fine-tuned DistilBERT model specifically trained for mental health sentiment analysis and risk detection in text content. The model can classify text into two categories:

• —Normal: Indicates healthy mental state or neutral content
• —Risk Detected: Indicates potential mental health concerns that may require attention

Model Details

• —Model Type: DistilBERT for Sequence Classification
• —Training Data: Curated mental health dataset with balanced samples
• —Languages: English
• —License: MIT

Performance

The model achieves the following performance on the validation set:

• —Overall Accuracy: 97.13%
• —Overall F1-Score: 97.13%
• —Normal Class F1: 97.16%
• —Risk Class F1: 97.10%

Dataset

The model was trained on a perfectly balanced dataset.

• —Total Samples: 50,000
• —Training Samples: 40,000 (80%)
• —Validation Samples: 10,000 (20%)
• —Class Balance: 50/50

Usage

from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch

# Load model and tokenizer
model_id = "techhy/mindtrack-mental-health-analyzer"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForSequenceClassification.from_pretrained(model_id)

# Example usage
text = "I am feeling great today!"
inputs = tokenizer(text, return_tensors="pt", truncation=True, padding=True, max_length=512)

with torch.no_grad():
    outputs = model(**inputs)
    probabilities = torch.softmax(outputs.logits, dim=-1)
    predicted_class = torch.argmax(probabilities, dim=-1).item()
    confidence = probabilities[0][predicted_class].item()

labels = {0: "Normal", 1: "Risk Detected"}
print(f"Prediction: {labels[predicted_class]} (Confidence: {confidence:.2%})")

Intended Use

This model is designed to assist in identifying potential mental health concerns in text content. Important: This tool should not replace professional medical advice or diagnosis. Always consult qualified healthcare professionals for mental health issues.

Limitations

• —Trained primarily on English text
• —May not capture cultural nuances in mental health expression
• —Performance may vary on text significantly different from training data
• —Should be used as a screening tool, not for final diagnosis

Training Data

The model was trained on a carefully curated dataset of mental health-related text, including:

• —Social media posts (anonymized)
• —Mental health support forum discussions
• —Clinical text samples (anonymized)
• —Balanced representation of risk and normal states

Ethical Considerations

• —Privacy: No personal information was used in training
• —Bias: Efforts were made to reduce bias, but some may remain
• —Responsible Use: Should be used to help people, not to discriminate
• —Professional Guidance: Always recommend professional help for mental health concerns

Citation

If you use this model in your research or applications, please cite:

@misc{mindtrack2024,
  title={MindTrack: Mental Health Sentiment Analyzer},
  author={Soham},
  year={2024},
  url={https://huggingface.co/techhy/mindtrack-mental-health-analyzer}
}

Contact

For questions or issues, please open an issue on the GitHub repository.