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

som branch, strange issues with interacting with javascript (does not exist) #42

Open mlin12321 opened 2 months ago

mlin12321 commented 2 months ago

Title, for some reason calling playwright page.evaluate("function") sometimes throws a not defined error on some pages, e.g. for page.evaluate("unmarkPage()") and page.evaluate(..."drawBoxes"...)

duz-sg commented 1 month ago

One possible reason is the page has certain redirect after the functions are evaluated. These functions need to be used after the mark_page.js been evaluated. If some page, for example, jumps to a new pages after few seconds, the whole page has been changed, then these functions become unavalable, unless the mark_page.js get evaluated again.