Phase 2 Quick Fix: Switch to clean Qwen2.5-0.5B-Instruct base model

Phase 1 Quick Wins: Optimize inference for 3-5x speed boost

Remove OAuth config and update README for direct model loading

Fix app.py: Replace InferenceClient with direct model loading using transformers

Add requirements.txt with transformers dependencies

Update app.py

Remove qwen references and update to Anki 2.5

initial commit

Duplicate from gradio-templates/chatbot