HRVibeCheck-Hire-Recommendation-Model

Fine-tuned model for candidate-job matching (Hire / No-Hire)

This model was developed as part of the ISOM5240 Group Project — Deep Learning Business Applications with Python.

Model Details

• Base Model: BERT / JobBERT variant
• Task: Binary Text Classification (Job Description + Resume)
• Input Format: JOB DESCRIPTION: {jd} [SEP] RESUME: {resume}
• Output: Probability of Hire (0.0 - 1.0)

Intended Use

• Automated resume screening for recruiters
• Part of the HRVibeCheck Streamlit application (Pipeline 1)

Training Data

• Custom JD-Resume matching dataset
• Fine-tuned with Hugging Face Trainer

Performance

Achieved strong validation accuracy during training (exact numbers in project report).

How to Use

from transformers import pipeline

pipe = pipeline(
    "text-classification",
    model="Cheykong/HRVibeCheck-Hire-Recommendation-Model"
)