Embedding openlit in GPT4All sdk

[patcher9]

trafficstars

Currently, users who want to integrate monitoring with the GPT4All stack need to install Openlit separately, as outlined in GPT4All documentation here. Some users have requested a more streamlined approach where the Openlit instrumentation is embedded directly within the GPT4ALL SDK.

Proposal:

Modularization: We modularize the GPT4ALL-specific instrumentation in openlit, making it a smaller, standalone component that can be easily integrated into the SDK.

SDK Integration:

Embed the modular Openlit instrumentation within the GPT4ALL SDK. Add an initialisation argument (e.g., enable_monitoring) that allows users to toggle Openlit monitoring on or off. Benefits:

I reached out to @mcembalest a while back about this but never opened this issue to get the discussion started. Lemme know what you'all think