Question 1

What is Teja Priyan AI?

Accepted Answer

Teja Priyan AI is a personal open-source AI project built by Teja Priyan Sivaraj featuring the Tejapriyan-8B language model, an interactive browser chat interface, an in-browser SQLite sandbox, and an instant NPX terminal assistant.

Question 2

What is the Teja Priyan model?

Accepted Answer

The Teja Priyan model (Tejapriyan-8B) is an 8-billion parameter fine-tune based on Qwen3-8B. It was trained using Supervised Fine-Tuning (SFT) for identity alignment and GRPO reinforcement learning with a live SQL execution reward sandbox, achieving 57.7% text-to-SQL accuracy on Spider dev benchmarks.

Question 3

What are Teja Priyan's main projects?

Accepted Answer

Teja Priyan's projects include the Tejapriyan-8B AI model hosted on Hugging Face, the interactive browser SQL sandbox, the npx tejapriyan terminal CLI, and the live model tunnel broadcaster.

Question 4

How do you install and run the Teja Priyan model?

Accepted Answer

You can run the model locally using Ollama by downloading the GGUF from Hugging Face (curl.exe -L -o tejapriyan-q4_k_m.gguf https://huggingface.co/teja161615/Tejapriyan-8B-GGUF/resolve/main/tejapriyan-q4_k_m.gguf), creating the model with 'ollama create tejapriyan -f Modelfile', and running 'ollama run tejapriyan'. Alternatively, run 'npx tejapriyan' for instant zero-install terminal chat.