Merge Development Branch

[KillianLucas]

Added

• Added starlette to pyproject.toml dependencies (@Notnaton)
• Implemented handling of server {STOP} token (@KillianLucas)
• Added support for ollama/llama3.1 model (@CyanideByte)
• Updated list of ollama models (@CyanideByte)
• Added SuppressDebugFilter to hide missing cost completion map warnings from LiteLLM (@CyanideByte)

Changed

• Updated YouTube video links in Setup and Profiles documentation (@MikeBirdTech)
• Refreshed list of ollama models (@CyanideByte)
• Modified chat completion function to handle the last message instead of all user messages (@KillianLucas)
• Updated model loading process for ollama models (@CyanideByte)
• Changed priority models list in local setup (@KillianLucas)
• Updated litellm completion function to remove ":latest" from model names (@CyanideByte)

Fixed

• Fixed ollama not working with llama3.1 and potentially other models (@CyanideByte)
• Fixed documentation for custom model usage (@rbrisita)

Documentation

• Updated custom models documentation with new function signature (@rbrisita)
• Added new video embed for Profiles documentation (@MikeBirdTech)

Contributors

• @KillianLucas
• @CyanideByte
• @MikeBirdTech
• @Notnaton
• @rbrisita

Thanks all! Great fixes and docs updates on this one.