README.md

metadata

license: mit
language:
  - en
metrics:
  - accuracy 97%
base_model:
  - distilbert/distilbert-base-uncased
pipeline_tag: text-classification

Last Name Classification Model

A Transformer-based classifier that checks if a provided last name is likely to be real (LABEL_1) or fake (LABEL_0). This can be helpful in validating contact form submissions, preventing bot entries, or for general name classification tasks.

Table of Contents

• Project Structure
• Installation
• Usage
• Support Me
• License

Project Structure

Last_Name_Prediction/
├── .gitattributes
├── README.md
├── config.json
├── model.safetensors
├── requirements.txt
├── special_tokens_map.json
├── tokenizer.json
├── tokenizer_config.json
└── vocab.txt

Installation

1. Clone the Repository:

git clone https://github.com/Vishodi/First-Name-Classification.git

2. Set Up the Environment: Install the required packages using pip:

pip install -r requirements.txt

Usage

from transformers import pipeline

# Replace with your model repository
model_dir = "vishodi/First-Name-Classification"

# Load the model pipeline with authentication
classifier = pipeline(
    "text-classification",
    model=model_dir,
    tokenizer=model_dir,
)

# Test the model
test_names = ["Mark", "vcbcvb", "uhyhu", "elon"]
for name in test_names:
    result = classifier(name)
    label = result[0]['label']
    score = result[0]['score']
    print(f"Name: {name} => Prediction: {label}, Score: {score:.4f}")

Output:

Name: Mark => Prediction: LABEL_1, Score: 0.9994
Name: vcbcvb => Prediction: LABEL_0, Score: 0.9985
Name: uhyhu => Prediction: LABEL_0, Score: 0.9982
Name: elon => Prediction: LABEL_1, Score: 0.9987

Support Us

License

This project is licensed under the MIT License.