cocacola-lab / MineLand

Simulating Large-Scale Multi-Agent Interactions with Limited Multimodal Senses and Physical Needs
MIT License
55 stars 9 forks source link

Analysing LLM/VLM behaviour and use of Gemini #25

Closed jbarrerobuch closed 2 weeks ago

jbarrerobuch commented 1 month ago

Hi there team,

I had implemented the use of Gemini with multiple API keys to use the free tier from Google for training. I've implemented with Langchain so it works with the parsers etc.

I wanted test and male sure the prompt and the outputs are working fine and make sure that is not the reason why the bot is not making progress.

Any clue or experienced shared or your side on how did you test it is wellcome :)

kind regards,

Pepe /

FooLiqi commented 2 weeks ago

please see issue #24