Crew AI Full Tutorial For Beginners - Build Your Own AI Agents | Tyler AI

Поделиться
HTML-код
  • Опубликовано: 25 ноя 2024

Комментарии • 63

  • @amitbuch
    @amitbuch Месяц назад +3

    Perfect intro to CrewAI in such a short tutorial. Best 48 minutes spent in very condensed learning. Thanks a bunch.

  • @HassanAllaham
    @HassanAllaham 4 месяца назад +13

    I like Ollama, I like CrewAi, and I enjoy the great content you provide. Thank you 🌹🌹🌹.. Oh I can not understand you people, 121 views and only 14 likes !!! shame on you people .. You all know that the good content in RUclips is truly rare .. We must support the good providers such as Mr. Tyler.. Thanks again Mr. Tyler .. Keep it good 🌹🌹🌹

    • @polyacrylamide4551
      @polyacrylamide4551 4 месяца назад +1

      and now has 4.2K views but only 190 likes. hikks why good content like this has little viewers.

  • @josepservat211
    @josepservat211 Месяц назад +3

    This intro is amazing thanks , but also crew ai is amazingly easy as well ! Kudos for this framework !!!

  • @mriwilis
    @mriwilis 15 дней назад +1

    Fantastic video, congrats. It's like you were coded by CrewAI yourself 🎉

  • @geneanthony3421
    @geneanthony3421 2 месяца назад +2

    You're honestly the first guide I found which got my crewai working for both openai and ollama. Every tutorial I found I ran into some issue.

    • @TylerReedAI
      @TylerReedAI  2 месяца назад

      Thank you, appreciate that!

  • @SetordoStiep
    @SetordoStiep 2 месяца назад +3

    Hi! I'm an 14 year old developer from Brazil! I appreciate your content guy, good luck with your channel! Just a note guys at minute 9 he tells about verbose setting it to 2. But verbose needs to be a Boolean so put it as True!!!!!!

    • @TylerReedAI
      @TylerReedAI  2 месяца назад

      Thank you I’m glad you’re into this! If you have questions let me know. And thank you, it may have been updated since this video. Thanks for noting this

  • @BenjaminK123
    @BenjaminK123 4 месяца назад +3

    This is a great video thank you, i have been playing with autogen and crewai for a while now and really loving it, quick tip for people i dont think this code is on the github yet and if your lazy like me and dont want to type just take a screen shot of the code in the video send it to gpt and ask it to print it out as text and pow you can easy copy and paste

  • @zaptrading
    @zaptrading 4 месяца назад +5

    I'm new to AI and "agents," but I recently coded a poem-generating script to learn Python and the related processes. It was a steep learning curve, but I have to say, your explanations are the best I've seen by far. Keep up the good work!
    I'm 57 years old, and I'm really excited about the potential of AI and what we can achieve with it. As a full-time trader specialising in gold (XAUUSD), I want to use AI to assist me with chart analysis. 😃

    • @TylerReedAI
      @TylerReedAI  4 месяца назад

      That's awesome! It's amazing to create things like that. Happy you were able to code a poem-generating script!
      Thank you, I'm glad you were able to follow along! Hey I see so many people using AI to help with trading, I hope you get some chart analysis as well :)

  • @danielherrera3855
    @danielherrera3855 4 месяца назад +1

    Greetings from Brazil. Two weeks trying all sorts os tutorials until getting across this one here. Fantastic video! Finally made it - all thanks to you! Will now work on the Crew I really want to, using web broweser and URLs scraper tools. Let's see how it goes. Thank you Tyler.

    • @TylerReedAI
      @TylerReedAI  4 месяца назад

      So glad to here that :). And greetings from USA

  • @Sumankumar-lo1vr
    @Sumankumar-lo1vr 4 месяца назад +2

    Nice Taylor.. It's clear to understand

  • @HamzaKhan-zj6dn
    @HamzaKhan-zj6dn 3 месяца назад +1

    underrated youtuber...

  • @aiamartist
    @aiamartist 9 дней назад +1

    Hi. I am getting this error when I try to run with calculator tool
    Input should be a valid dictionary or instance of Basetool

  • @thelalomorales
    @thelalomorales 4 месяца назад +1

    nice! crewai for the win

  • @AliAlias
    @AliAlias 3 месяца назад +1

    Thanks, great work ❤🙏👍

  • @bambanx
    @bambanx 3 месяца назад

    suscribed and like, thanks you so much, very fun and practical , very well explained, cheers from Chile.

  • @Karthik-ln7eg
    @Karthik-ln7eg 3 месяца назад +1

    Hi Tyler. Great content. Your explanation was easy to understand for a beginner such as myself. I cannot believe that I have not yet heard of your channel. Subscribed immediately.
    @TylerReedAI As someone new to the world of AI Agents, is it better to learn Autogen first or Crew AI or something else? Which one has more functionality, easier to get into and more useful in the long run?

    • @TylerReedAI
      @TylerReedAI  3 месяца назад +1

      Well thank you for saying that! I appreciate the sub :). So, heres a small breakdown. In my opinion I think CrewAI is easier to start out with. AutoGen isn't bad starting out either, it just can get a little complicated, but has a lot of features. With that said CrewAI has been coming out with updates.
      That may have not necessarily answered your question lol, but I would have definitely said start out with AutoGen before, but they are coming close to maybe choosing what your preference is!

    • @Karthik-ln7eg
      @Karthik-ln7eg 3 месяца назад

      @@TylerReedAI Thanks for your reply. I have started with learning Crew AI. Seems a bit more intuitive.
      Coming from a python programming background, the weird thing with LLM's and Agents is being unable to debug line by line and understand what the problem is.
      Right now my results are one of two. It either works or it doesn't. I guess I will get better with time.

  • @NakedSageAstrology
    @NakedSageAstrology 3 месяца назад +1

    This is very similar to something I have been working on. Except mine makes the agents into functions of mind.

  • @adityanjsg99
    @adityanjsg99 2 месяца назад

    Thats a very information dense video!! Tyler can you tell me what advantages does it have over Langchain? I collect that langchain has similar features.

  • @tonyjoffre2909
    @tonyjoffre2909 3 месяца назад +1

    TY

  • @nufh
    @nufh 2 месяца назад +1

    Hi, I just found your channel. Thanks! It’s really useful for someone like me who's just starting out with programming. Do you have any tutorials related to RAG and Qdrant?

    • @faturismee
      @faturismee 2 месяца назад

      hi, im also on searching it. have you found them ?

    • @nufh
      @nufh 2 месяца назад

      @@faturismee Not yet.

    • @TylerReedAI
      @TylerReedAI  2 месяца назад

      Not…..YET 😁

    • @TylerReedAI
      @TylerReedAI  2 месяца назад

      I don’t but I hear you, and I planned on a video covering different tools with crew ai

    • @nufh
      @nufh 2 месяца назад

      @@TylerReedAI Thanks man, your explanation is easy to understand for a someone like me who is still new with Python.

  • @AaronChan-x2d
    @AaronChan-x2d 3 месяца назад +1

    great video. With regards to lags running on your own machine, why not Sagemaker?

    • @TylerReedAI
      @TylerReedAI  3 месяца назад

      Just haven't tried it yet! What are your thoughts on it?

    • @AaronChan-x2d
      @AaronChan-x2d 3 месяца назад

      @@TylerReedAI For this type of experimentation I think you get a lot of benefits using something like Sagemaker, it basically offers the elastic computing power you need on demand, doing away with limitations from running local. Just monitor your costs.

  • @SWV_AI9
    @SWV_AI9 3 месяца назад

    Sorry, if this is covered somewhere else but on Windows 11 when running pip install ... I get a RuntimeError: uvloop does not support Windows at the moment. Since you are doing this on Mac, you wouldn't see this issue. I updated the requirements.txt file to use winloop==0.1.6 instead of uvloop and pip install -r requirements.txt did its thing. So that's cool

  • @AjayCoding
    @AjayCoding Месяц назад

    Hey Tyler, I want to learn AI and build AI agents, but I’m unsure where to start. Should I begin with the basics of AI, or should I jump straight into building AI agents and integrating them with social media? or do i need learn AI from basics with maths!!! 😐
    Also, as a full-stack developer, should I focus on tools like LangChain or LangSmith, crewAI, etc.., or should I explore no-code platforms? I’m looking for some guidance to help me choose the best path forward. Thanks!

    • @TylerReedAI
      @TylerReedAI  Месяц назад +1

      Well tbh, I would just jump into it. For developing. But it is good to understand what like LLMs do, the different companies, what kinds of models there are, things like that. Understanding those things will help you know what you want to make! Email me or join the discord and I can help you out

    • @AjayCoding
      @AjayCoding Месяц назад

      @@TylerReedAI sure

  • @sozno4222
    @sozno4222 3 месяца назад

    In your PDF example, did you compare the output quality from the agents vs if you would have just uploaded the pdf to ChatGPT and promoted it ?

  • @bambanx
    @bambanx 3 месяца назад

    i think you can use crew with openrouter and point to cheap models

  • @geneanthony3421
    @geneanthony3421 2 месяца назад

    Like your guide, but right at the beginning your verbose statement needed to be set to True not 2. Gave an error. Also appreciate the slow start. Been running into a lot of issues with these tutorials with broken issues, dependency issues, etc. At least with a slow start I can see where errors start easier.

    • @TylerReedAI
      @TylerReedAI  2 месяца назад +1

      Yes I think this was updated, somebody else also mentioned this I will update this in the code. Thank you for noticing though!

  • @HorizonLeganes
    @HorizonLeganes 4 месяца назад

    hey Tyler right now, what do you think is going into the best direction autogen or crewai? im starting a long term project and i value your opinion on which to choose, thanks!!!!!❤

    • @BuildNewThings
      @BuildNewThings 4 месяца назад +2

      crewAI creator here, I'm biased, but we do offer a business offering, that includes deployment, monitoring and much more, so I'd say crew is better prepared for production. Last month alone we ran over 10M agents (through opensource) :)

  • @wreckball2315
    @wreckball2315 2 месяца назад +1

    Update
    use verbose = True instead of 2 while creating crew. I guess the might have updated that

    • @TylerReedAI
      @TylerReedAI  2 месяца назад

      Thank you! Yeah probably updated since then, thanks for showing this so others can have update also

  • @AbdelrahmanAzab-n3x
    @AbdelrahmanAzab-n3x 3 месяца назад +1

    Task output: Agent stopped due to iteration limit or time limit. can you tell me how to fix it , i made the same code and another code with same error

  • @OtherCoder
    @OtherCoder 2 месяца назад

    Tools don't work with ollama. Any work around for that?
    I want to use local llms along with tools.

    • @TylerReedAI
      @TylerReedAI  2 месяца назад

      It depends on the local ai model unfortunately. Ollama supports functions now though in a recent update

  • @OtherCoder
    @OtherCoder 2 месяца назад

    Ollama is not working for the calculate tool. In general ollama not working properly with tools.

    • @jjasghar
      @jjasghar 2 месяца назад

      Yep, i'm trying it myself, i get Agent stopped due to iteration limit or time limit. attempting to do the math wiz portion. I have ollama running locally, but it seems there is something that we're missing to get it to work.
      The first example with the octopus does work though.

    • @TylerReedAI
      @TylerReedAI  2 месяца назад +1

      Okay, I can take a look and see what's going on with Ollama. I believe they released an update to aid with this, but I can check

    • @screyesg
      @screyesg День назад

      @@TylerReedAI i had the same error with Ollama

  • @not_a_human_being
    @not_a_human_being 4 месяца назад +1

    can it do anything more complex though?

    • @TylerReedAI
      @TylerReedAI  4 месяца назад

      Yes it can! I just don't have the most creative brain to figure out more complex stuff to do lol