Beibin Li

Hi everyone! I've noticed that the Gemini integration is becoming increasingly popular and complex. Therefore, I am planning to pause merging into the "main" branch. Instead, I will request multiple...

I am okay with this notebook. However, the agent is designed for "Spider" benchmark, which is not used generally for actual product. For instance, when users want to use a...

I really like the PR. Can we also add a really simple Notebook to illustrate the use case? Some other notes. 1. The motivation of this PR is very similar...

As others stated, my biggest concern is in introducing new classes. While they can simplify the nested chat system, users have to go over a steep learning curve and still...

> There is an issue #1069 due to the filename is being used as docker's tag name so it cannot start with a period (https://docs.docker.com/engine/reference/commandline/tag/). PR #1127 looks to fix...

@luxzoli Thanks for your great PR! I will take a look at it. @arjun-g Please also take a look at luxzoli's PR. For your function calling PR, can you add...

@afourney Yes, ideally it will work. Do you want to use GPT-4V for `MultimodalWebSurfer` or for all agents? I think using GPT-4V for all agents might produce better results. See...

@afourney Got it! Then, yes, this design would work.

> So as you know I'm a big fan of this PR -- but am super nervous about introducing a prompting DSL that diverges from other prompting template DSLs (e.g.,...

@afourney Yes, usually images are treated separately in the model. There is often an "image encoder model" (separate of the LLM) to process the image, and this "encoder" can be...