license: apache-2.0

MentiFy – Fine-Tuned DistilBERT

Fine-tuned DistilBERT model for mental-wellness text classification into 7 categories.

Model Details

• Base Model: distilbert-base-uncased
• Task: Text Classification
• Dataset: 51,073 cleaned samples
• Classes: 7
• Accuracy: ~87.25%
• Framework: PyTorch, Hugging Face Transformers

Usage

pip install transformers torch

from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch

model_name = "anisha-14/Mentify-DistilBERT"

tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name)

text = "I have been feeling stressed lately."

inputs = tokenizer(
    text,
    return_tensors="pt",
    truncation=True,
    padding=True
)

with torch.no_grad():
    outputs = model(**inputs)

prediction = torch.argmax(outputs.logits, dim=1).item()
print(model.config.id2label[prediction])

MentiFy Integration

The model is used in MentiFy (mentalGUARDIAN) to classify user responses from a wellness questionnaire. Multiple predictions are combined using a majority-based approach to determine the final category.

Disclaimer: This model is for educational and wellness-support purposes only and is not a medical diagnostic tool.