search

regostar / eventique

AI Enhanced Event Planning
MIT License
0 stars 1 forks source link

Spike - Research and decide the right LLM to use in Eventique #1

Open regostar opened 4 months ago

regostar commented 4 months ago

To evaluate different Large Language Models for potential integration into Eventique. The evaluation criteria will prioritize cost (specifically the availability of free API credits), ease of usability (including the complexity of API integration)

Parameters which we consider in the order specified -

  1. Cost - Free credits for API access
  2. Ease of Usability - API integration complexity

(We do not consider speed at this stage in the development.)

Various LLM's to explore :-

Expected Outcome:-

A comparison report outlining the advantages and limitations of each considered LLM. A recommendation for the most suitable LLM for Eventique, balancing cost, usability, and speed, to facilitate informed decision-making for the integration process.

regostar commented 3 months ago
  1. https://github.com/LAION-AI/Open-Assistant
  2. https://github.com/ianmkim/gpt_llm?referral=text-to-speech-python - GPT-NeoX-20B
  3. Microsoft for Startups team -

Image

Proceeding with Azure AI studio which has access to all these models which can be integrated with ease. Provided with 1000$ credit helps.

Image

regostar commented 3 months ago

Alternate is Gemini -Tested API access using Python