This is a continuously updated handbook for readers to easily track the latest Text-to-SQL techniques in the literature and provide practical guidance for researchers and practitioners.

core shell functions building blocks for advanced AI pipelines

A Language Interpreter as semantiC Experiment in natural language processing

Five AI function sections for an Office Personal Data Assistant (OPDA) in Conversations, Coding, Images and Video Production, Music Composing, and Tech Support Chat.

Bird-Pandas is a cross-paradigm benchmark and evaluation pipeline that aligns Text-to-SQL with Text-to-Python (Pandas) by deriving verified Python solutions from the BIRD dataset, augmented by the Logic Completion Framework (LCF) for robust data analysis code generation.

AI-powered design tool. Describe any design and get polished HTML/CSS variations with real images

Play with characters (GUI)

Key Words: Generative AI, Machine Learning, Code-Generation

Key Words: Generative AI, Machine Learning, Code-Generation

Applications defined as Markdown specifications, executed by agentic coding harnesses. No compiled binary. Text is the code.

Self-correcting data-analyst agent: ask a dataset a question in plain English and get runnable Python, SQL, or an Excel formula. Built on LangGraph + a local Qwen2.5-Coder LLM, with a generate→execute→reflect loop that fixes its own errors (measured: 75%→94% success).

A powerful, chainable regex builder platform with comprehensive validation utilities

VS Code extension for CodeVerb

This repository contains a complex evaluation of qwen1.5-0.5B coding abilities

🚀 A fully custom Transformer built from scratch in PyTorch that converts human-written pseudocode into real C++ code. Trained on Stanford’s SPoC dataset and deployed on Hugging Face Spaces with an interactive Gradio app for instant code generation.

Fine-tuning CodeT5 for Python code generation on the MBPP dataset. Features custom TensorFlow training loops, mixed precision, XLA optimization, and distributed multi-GPU strategies.