Skip to content

silsgah/realtime_training

Repository files navigation

📩Real time cryto prediction

This project is an SMS spam classifier built on top of a fine-tuned GPT-2-based language model. It leverages modern tooling in Python, such as pydantic, loguru, ruff, and make, to deliver a clean, maintainable, and scalable machine learning pipeline.

🚀 Features

  • 🧠 Fine-tuned GPT-2 model for binary SMS classification (spam vs ham)
  • 🔄 Balanced dataset preprocessing
  • 🛡️ Pydantic for robust data validation
  • 📊 Dataset metrics and confusion matrix visualization
  • 📦 Tooling includes ruff, loguru, and make for linting, logging, and workflow automation

🧰 Tech Stack

  • Python 3.12+
  • Transformers (HuggingFace)
  • Datasets
  • Pydantic
  • Loguru
  • Ruff
  • Make

📁 Project Structure

About

training on realtime datapipeline

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors