Desktop GUI app for automating deep learning training on Vast.ai cloud GPUs — classification & regression with one click

This repository features an end-to-end Dyula-to-French translation system built with Joeynmt, addressing low-resource language challenges. It incorporates MLOps best practices for optimizing accuracy, latency, throughput, and cost efficiency. The project is fully documented for reproducibility and deployed on the AWS-powered HighWind platform.

A comprehensive machine learning pipeline for predicting startup success using Crunchbase data. This project demonstrates end-to-end ML workflow including data preprocessing, feature engineering, model training, and evaluation.

Open-Source PDF Assistant: This tool allows users to ask questions based on the content of a PDF by simply providing a link to the document. It leverages Docker to create a vector database using pgvector for efficient text retrieval, ensuring unlimited queries without OpenAI embedder limitations. 🚀📄

Production-style retail demand forecasting: LightGBM ensemble, FastAPI + Streamlit, 0.40 RMSLE on Kaggle Store Sales

Titanic survival prediction MLOps pipeline with Apache Airflow and MLflow

Image classification with transfer learning using PyTorch (ResNet / EfficientNet)

MLops