🔍 Resume Job Role Classifier

A fine-tuned DistilBERT model to classify job roles based on resume content. This model is trained to predict the most likely profession from the given resume text, supporting over 14 different job categories.

🧠 Model Details

Architecture: DistilBERT (base, uncased)
Task: Multi-class text classification
Input: Raw resume text (English)
Output: Predicted job category label and score

📊 Labels Covered

This model supports classification into the following job categories:

Data Science
Java Developer
Web Designing
HR
Mechanical Engineer
Electrical Engineering
Civil Engineer
Arts
Advocate
Sales
Health and fitness
Business Analyst
SAP Developer
Automation Testing

🏋️ Training

Dataset: Custom dataset containing labeled resumes
Split: 80% train / 20% test
Metrics:
- Accuracy: 99–100%
- F1 Score: ~0.99–1.00 (macro avg)
Epochs: 3
Batch size: 8
Optimizer: AdamW

📥 How to Use

from transformers import pipeline

classifier = pipeline("text-classification", model="srivihari/resume-job-role-classifier")
result = classifier("Experienced data scientist with Python, machine learning, and statistics background.")

print(result)
# Example output:
# [{'label': 'Data Science', 'score': 0.97}]
#Note: If you face any issues like token_type_ids errors, make sure to adjust tokenizer config as below:
#from transformers import AutoTokenizer, AutoModelForSequenceClassification, pipeline

#model = AutoModelForSequenceClassification.from_pretrained("srivihari/resume-job-role-classifier")
#tokenizer = AutoTokenizer.from_pretrained("srivihari/resume-job-role-classifier")
#tokenizer.model_input_names = ["input_ids", "attention_mask"]
#classifier = pipeline("text-classification", model=model, tokenizer=tokenizer)