Team Status Report for 3/30/24

Risks:

• Kria KV260 acceleration documentation doesn’t seem the most up to date

Changes:

• Using ChatGPT instead of Llama2 or Gemma
  • ChatGPT has more variety of recipes while Llama2 and Gemma do not work as intended, and we do not have enough time to fine-tune and train the LLM
  • Our LLM will now be cloud-based instead of local, so the LLM is not in our server. We will use API requests from the flutter app to the LLM in the cloud.
  • This came with the tradeoff of using a cloud-based LLM, which is not as secure as a local LLM.

Progress:

• OpenAI API set up on Flutter app
• Additional UI elements: swipe to delete, filters pop up