Enhance UI styling and layout for Next-Token Predictor with gradient backgrounds, improved button designs, and updated tooltip functionality

Refactor to use local model inference for next token prediction and enhance token display functionality

Update to correct Inference Providers chat completions API format

Fix 404 errors: Update to new Inference Providers API endpoint (router.huggingface.co/hf-inference)

Switch to openai-community/gpt2 with gpt2-medium fallback for better serverless inference ability

Switch to GPT-2 for better serverless inference

Remove index.html and style.css files; add requirements.txt for dependencies