Testing the Power of Google's Gemini PRO 1.5 : A Game-Changing 1 Million Token AI Multimodal Model!

Поділитися
Вставка
  • Опубліковано 28 лип 2024
  • Get ready for the ultimate AI breakthrough as Google drops Gemini Pro 1.5, outshining GPT-4 Turbo with a staggering 1 million tokens, setting the new standard in AI supremacy! With lightning-fast training capabilities, Gemini 1.5 Pro handles an hour of video, 11 hours of audio, or over 700,000 words in a document, leaving competitors like OpenAI and Anthropic in the dust!
    ▶ Link to the Gemini Pro 1.5 Blog : blog.google/technology/ai/goo...
    ▶ Link to the Gemini Pro Long Context Window Blog : blog.google/technology/ai/lon...
    ▬▬▬▬▬▬ VIDEO CHAPTERS & TIMESTAMPS ▬▬▬▬▬▬
    00:00 Intro to Google Gemini 1.5 Pro
    00:43 Context Window Discussion
    01:13 Querying PDF Documents
    04:02 Demo: Image Sample 01
    05:22 Demo: Image Sample 02
    07:17 Demo: Video Sample 01
    07:59 Demo: Using Vide Sample 02
    9:00 Conclusion
    ▶ Sponsor me on GitHub : github.com/sponsors/bhattbhav...
    ▶ Join this channel to get access to perks: bit.ly/BhaveshBhattJoin
    ▶ Join the Telegram channel for regular updates: t.me/bhattbhavesh91
    ▶ If you like my work, you can buy me a coffee : bit.ly/BuyBhaveshCoffee
    *I use affiliate links on the products that I recommend. These give me a small portion of the sales price at no cost to you. I appreciate the proceeds and they help me to improve my channel!
    ▶ Best Book for Python : amzn.to/3qYThqu
    ▶ Best Book for PyTorch & Machine Learning : amzn.to/3PyUkdy
    ▶ Best Book for Statistics : amzn.to/3vzvHEn
    ▶ Best Book for BERT: amzn.to/3lpX0fz
    ▶ Best Book for Machine Learning : amzn.to/2P6aZuT
    ▶ Best Book for Deep Learning : amzn.to/30UMTGl
    ▶ Best Intro Book for MLOps : amzn.to/3AoPZmM
    Equipments I use for recording the videos:
    ▶ 1st Laptop I use : amzn.to/3AqI8Fp
    ▶ 2nd Laptop I use : amzn.to/3KAiYsB
    ▶ Microphone : amzn.to/3qUPxtz
    ▶ Camera : amzn.to/3rKQsM2
    ▶ Mobile Phone : amzn.to/3nRHP1f
    ▶ Ring Light : amzn.to/33LedM5
    ▶ RGB Light : amzn.to/3KzLgmS
    ▶ Bag I use : amzn.to/3AsM3RZ
    If you do have any questions with what we covered in this video then feel free to ask in the comment section below & I'll do my best to answer those.
    If you enjoy these tutorials & would like to support them then the easiest way is to simply like the video & give it a thumbs up & also it's a huge help to share these videos with anyone who you think would find them useful.
    Please consider clicking the SUBSCRIBE button to be notified for future videos & thank you all for watching.
    You can find me on:
    ▶ Blog - bhattbhavesh91.github.io
    ▶ Twitter - / _bhaveshbhatt
    ▶ GitHub - github.com/bhattbhavesh91
    ▶ Medium - / bhattbhavesh91
    ▶ About.me - about.me/bhattbhavesh91
    ▶ Linktree - linktr.ee/bhattbhavesh91
    ▶ DEV Community - dev.to/bhattbhavesh91
    ▶ Telegram - t.me/bhattbhavesh91
    #googlegemini #googleaistudio

КОМЕНТАРІ • 33

  • @jaymishra6219
    @jaymishra6219 2 місяці тому

    "You have reached your rate limit. What should you I do?"

  • @Homer502a
    @Homer502a 5 місяців тому +1

    Thanks for the fascinating demo! You could try asking more detailed questions about the India Gate video. For example, "Describe the clothes worn by the couple who walk by holding hands." Or "What colour is the car shown on the right side and how many children are near it?"

    • @bhattbhavesh91
      @bhattbhavesh91  5 місяців тому

      Thanks for the idea! I'll include this in the future video

  • @souvickdas5564
    @souvickdas5564 5 місяців тому

    I am having one problem with input context length. For example given a research paper, I am trying to find relevant papers from the vector db containing 2000 papers. How to fit the entire research paper as the input? Is there any way to solve the problem? Also the vector db is huge. Is there any way to manage it efficiently?

    • @nicolasgarcia253
      @nicolasgarcia253 5 місяців тому

      how u are testing 1.5 version? i only can choose pro 1.0 :/

  • @Junior-zf7yy
    @Junior-zf7yy 5 місяців тому

    Nice video

  • @AP-hv5dh
    @AP-hv5dh 5 місяців тому

    Excellent demo Bhavesh! 🔥 The responses were lightening fast. Was that that real time?

    • @bhattbhavesh91
      @bhattbhavesh91  5 місяців тому +2

      I'm glad you liked it! The response did take some time but the quality of response were mind-blowing :)

    • @AP-hv5dh
      @AP-hv5dh 5 місяців тому

      @@bhattbhavesh91 It really is impressive🙌🏾

    • @bhattbhavesh91
      @bhattbhavesh91  5 місяців тому

      I'm glad you liked it!

  • @tejaschheda4610
    @tejaschheda4610 5 місяців тому

    8:49 OH MY GOD!! Ab mujhe AI se dard lane laga hai!!! Itna accurate toh Police bhi nahi hai jitna Gemini 1.5 Pro hai!!

  • @01xKeven
    @01xKeven 5 місяців тому

    How did you get access to gemini 1.5 pro?

  • @DanFrederiksen
    @DanFrederiksen 5 місяців тому

    Not sure input window size is all that important when it doesn't really grasp the situation all that well. other than having degraded GPT4 recently, it makes a gigaton of mistakes on even very shallow tasks. I'll ask it to make a table with some specs on a couple of aircraft models and it will miscalculate them a lot. I need to hold its hand in multiple iterations to get anything right. And often it will just give me platitudes and tell me to find it myself. Which is a unique approach. I asked it to list the cheapest popular private jets and it mentioned a couple of the most expensive and couldn't be bothered further. There was also a time when I asked it to make a table and after the first two items it just wrote a generic placeholder for multiple lines. Like it was a recipe rather than actual data. Again a very unique approach. It's impressive but too often useless even when I cater to its shallow intelligence.

  • @RpgBlasterRpg
    @RpgBlasterRpg 5 місяців тому

    When Gemini 1.5 Pro will be released?

  • @abhishekak9619
    @abhishekak9619 5 місяців тому

    Did they take the preview away from you ir do you still have it?

    • @bhattbhavesh91
      @bhattbhavesh91  5 місяців тому

      I have it! why?

    • @abhishekak9619
      @abhishekak9619 5 місяців тому

      @@bhattbhavesh91 no reason when do you think it will be out. Is there any way to get it now. I dont need the 1m tokens really i just want to see how precise the video observation is, is it per frame or it somehow understands the video on someway that is more than frames.
      Not really no reason ig.

    • @abhishekak9619
      @abhishekak9619 5 місяців тому

      @@bhattbhavesh91 could i use it somehow?

    • @bhattbhavesh91
      @bhattbhavesh91  5 місяців тому

      You will have to wait to get access!

    • @abhishekak9619
      @abhishekak9619 5 місяців тому

      @@bhattbhavesh91 how many people get access to it out of lets say 100 people how many get access to it?

  • @Adhil_parammel
    @Adhil_parammel 5 місяців тому

    You should test some big files,like 3hr video or 20k page size book.

    • @bhattbhavesh91
      @bhattbhavesh91  5 місяців тому

      That's a good idea!

    • @elawchess
      @elawchess 5 місяців тому

      !M tokens is 1 hour video though. How is he going to do 3hr video?

    • @Adhil_parammel
      @Adhil_parammel 5 місяців тому

      @@elawchess their technical report says it can work upto 3hr.anyway we can try extend of possible limit.

  • @Yakibackk
    @Yakibackk 5 місяців тому +1

    wow!! Can you test take screenshot by your computer and tell it to find coordinates of the time for an exempel. Because today chatgpt4 fails to detect accurately the coorinates x and y. If Gemini pro can do it. its a game changer for the indusrty

    • @bhattbhavesh91
      @bhattbhavesh91  5 місяців тому +1

      Thank you for sharing your idea! I'll try this as well.

    • @Yakibackk
      @Yakibackk 5 місяців тому

      @@bhattbhavesh91 no problem, Plz update me ;)

    • @krushnachChandra
      @krushnachChandra 3 місяці тому

      @@bhattbhavesh91 what is the price ? can you explain