search

OSU-NLP-Group / SeeAct

[ICML'24] SeeAct is a system for generalist web agents that autonomously carry out tasks on any given website, with a focus on large multimodal models (LMMs) such as GPT-4V(ision).
https://osu-nlp-group.github.io/SeeAct/
Other
571 stars 69 forks source link

Config not loaded? #27

Closed manuel-delverme closed 3 months ago

manuel-delverme commented 3 months ago

https://github.com/OSU-NLP-Group/SeeAct/blob/9d2fba437b8bb83f34e565122dbd47e59fc319cc/seeact_package/seeact/agent.py#L68C1-L71C1

Is this a bug or am i misunderstanding the code?

boyugou commented 3 months ago

We haven't finished the documentation for the package, and also there might be some bugs.

However, let me explain the logic here.

In the initialization, you can either input a TOML file as a config file or use the default configuration. (You can change the default settings by passing the parameters.)

manuel-delverme commented 3 months ago

wouldn't you have config = toml.load(blah) in there?

boyugou commented 3 months ago

OK, I forgot about that. (probably deleted it mistakenly). You are totally right. Thanks for that and pushed a new version.