Increase max tokens to 16384 for reasoning models

Update README

Update user message about model response time.

Change kimi K2 Thinking to K2 Instruct.

Update model's response with a structured pattern targeting user learning process of coding.

Remove Qwen Coder 32B Instruct. Added DeepSeek V3.2 (novita)

Change Qwen 2.5 Coder 7B to 32B Instruct

Fix LLM Access Error on App.

Update README

Remove notebook download feature (keep Markdown and Code)

Remove deprecated type parameter from Chatbot (Gradio 5.x+)

Initial commit: GeoAI Coding Agent app

initial commit