Timeline

1. LLM in a Box implementation ✔
2. Groq implementation ✔
3. Ollama implementation (offline) ✔
4. Save logs offline, then send it to LLM in a Box with chatid ✔
5. Route between offline and online models for seamless interaction ✔
6. Using websockets for ElevenLabs API # not possible
7. Use Groq token streaming ✔
8. Llama.cpp implementation (offline) C
9. Whisper.cpp (offline) ✔
10. Other offline TTS ✔
11. Gather dataset for Immy offline model ✔
12. Finetune a model for offline Immy ✔
13. LLM in a Box streaming ? # not possible