Web Scraping With GPT-4 Vision AI & Playwright Is Ridiculously EASY - I Can't Believe This Works

Поділитися
Вставка
  • Опубліковано 11 гру 2024

КОМЕНТАРІ • 11

  • @uzimaborehole
    @uzimaborehole 17 годин тому +1

    short and precise

  • @rdnexta
    @rdnexta 17 годин тому

    3rd comment😇, always good infromative video as usual.❤

  • @JaiRaj26
    @JaiRaj26 15 годин тому

    How to go about it if there are a lot of tabs and you need to fetch all the data from each tab? Example: A grocery site where there are several tabs such as "Fruits", "Vegetables", "Dairy", "Eggs, Meat", "Snacks" and so on and you need to fetch all the products from every tab?

    • @ByteGrad
      @ByteGrad  15 годин тому

      If those tabs have a dedicated url you’ll need some way to find those url’s. Without dedicated url you’ll need to simulate a click on the tab most likely

  • @devklepacki
    @devklepacki 16 годин тому

    What about just feeding entire html page into LLM and making it extract the data? 🙂 Or to optimize this a little bit, instead of pushing entire page's html, just get eg. wrapping element's html like #productList or sth.

    • @ByteGrad
      @ByteGrad  16 годин тому +1

      Possible, but can be costly

  • @anonanon7368
    @anonanon7368 15 годин тому

    do a video on how to scrap authenticated websites

  • @Lapatate-s1l
    @Lapatate-s1l 17 годин тому +1

    1st comment

    • @ByteGrad
      @ByteGrad  17 годин тому +1

      Nice!

    • @Lapatate-s1l
      @Lapatate-s1l 17 годин тому +1

      @ really can’t wait to see some react 19 / next 15 / tailwind videos for a real world project . Thanks for all

  • @Empathy2125
    @Empathy2125 14 годин тому

    Great video. Few things:
    1. You're repeating the same info multiple times. First 4 minutes of the video can be skipped.
    2. This is good to show case what OpenAI apis can do, but the application is completely unnecessary and much would be more efficient to use just Playwright.