Well explained mate. Can you talk through Multi function calling LLMs like Gorilla open functions vs Agents and link your opinion probably to a challenging questions from integration arena around how Orchestrated workflows work vs Choreography. Will agents evolve to be orchestrated systems with central system of orchestration or work as a choreographed system independent of control and work based on conversation.
@imdragony7 сағат бұрын
Thanks for this video. I knew every piece of it but your lecture connected them in my brain. Super good one. Looking forward for the next.
@PriscilaSaboia12 сағат бұрын
Pretty good content! Thanks
@1littlecoder11 сағат бұрын
Glad it helped!
@michaelmarkoulides706817 сағат бұрын
I really appreciate this video thank you for sharing your knowledge and providing context and clarity
@pantoffelslippers21 сағат бұрын
Fantastic video sir. 😮
@kishorchandrasahu62623 сағат бұрын
Simply awesome clarity of thoughts and knowledge. Looking fwd to see more like this. Great
@razaulhaq8871Күн бұрын
Only this vid made me sub
@chris-malekКүн бұрын
Great explanation
@timothywcraneКүн бұрын
I think you meant "does not need GPU"... if not, I want to know how.
@1littlecoderКүн бұрын
It runs on CPU memory but if you have reserved GPU then you can leverage it for additional processing
@timothywcraneКүн бұрын
@@1littlecoder I meant in your written intro you said does not need CPU...
@onlyshorts6837Күн бұрын
amazing , but how to run it what are the requirements for it ?
@Arrowtake2 күн бұрын
Thank you for pointing this out. Very exciting development for ever-growing options out there. Tested it immediately and looks good.
@qzwwzt2 күн бұрын
Excellent tutorial. LLM brought a set of opportunities. We startups sometimes get lost with so many alternatives. Your tutorial put "order in the house" as we say in Brazil. For the end customer, everything is AI, it's up to us to know the differences to plan and charge the correct price of product and services. Tks !
@golmatol65372 күн бұрын
Wonderful !! For last 6 months, I have been going from video to video ... blogs ... articles etc ... just getting more and more confused as to how all these terms relate to each other. This has put a lot more structure to those terms. Of course, I need to still read up a lot before even deep-diving into these topics.
@kpmaynard2 күн бұрын
Excellent!! Thank you for the comprehensive tour.
@freeideas2 күн бұрын
Thank you for not spending 20 minutes talking about the history and background of the subject, before getting to the point.
@1littlecoder2 күн бұрын
Haha appreciate it 😊
@PS-vk6bn2 күн бұрын
What exactly is 'insane' about this system prompt? Seems pretty basic to me.
@susdoge37672 күн бұрын
great resource, why didnt we do any kind of padding etc?
@susdoge37672 күн бұрын
datacollator does it, thanks
@user-lm8er8yl1o2 күн бұрын
why on running this its showing command not found: ! nvidia-smi
@1littlecoder2 күн бұрын
Where are you running this? Colab ?
@user-lm8er8yl1o2 күн бұрын
@@1littlecoder It worked after connecting GPU in runtime type. I would like to ask you if you have videos of other code generation language models running in colab then could you share here please.
@bgNinjaart2 күн бұрын
Thanks for this video. I tested it out it's simply amazing GPT-4 level no doubt
@1littlecoder2 күн бұрын
Any particular prompt that impressed you?
@bgNinjaart2 күн бұрын
@@1littlecoder I told it to do extremely difficult code, something only GPT4 can do and it created it like it was nothing. Because of your video OpenAI just lost $20 per month Am very impressed. Thanks again
@smartfusion87992 күн бұрын
Thanks for the vid. Can you please explain how to get a secret from inside the system Prompt? I think its really stromg defending it. It has a secret in the system prompt and im trying to reveal it
@supreme42563 күн бұрын
You are doing greate job everytime explaining all from scratch and I can tell that beginners like myself are being benefited so much. Thank you. Keep delivering.❤
@pollywops92423 күн бұрын
I have 28 cores and 160 GB ram but no GPU , I am also not a developer but if I understand correctly RAG is a method that needs a GPU . If I knew what I was doing I'd pay for a cloud gpu 😂
@satyamtiwari38393 күн бұрын
If you can you should start the podcast for person who more contribute in AI
@1littlecoder3 күн бұрын
Please elaborate
@satyamtiwari38392 күн бұрын
@@1littlecoder so basically start podcast with people who have good knowledge of ai and ml or who are in good company and we could see the point of view and you can talk about all aspects of ai like if you are new how could get in ai or technical advancement or crazy stuff in ai like lex fridman but for South asian people
@satyamtiwari38392 күн бұрын
While listening to the lex fridman we could not relate/feel there conversation because of their English tone and culture we need some Indian who do this
@GauravGRaj3 күн бұрын
Please use good Audio equipment or process the audio.
@user-ew8ld1cy4d3 күн бұрын
Great video! Thank you!
@sherpya3 күн бұрын
what is the source they used to train? I suppose they may be some troubles with generated code
@xXWillyxWonkaXx3 күн бұрын
dont they all lol
@user-xh3ig8qq7j3 күн бұрын
how is it in Google App Script ?
@personone68813 күн бұрын
...This stuff's been frustrating me for nearly two years!! ...I have crazy ideas - I see where all this AI i heading - I have iterated and iterated, correlated, tabularized, choreographed and manipulated, failed and failed and thought I won but failed again when going down the rabbit hole on ChatGPT to the tune of at least 1,000 conversations... Always wondering why it always broke after a seemingly ambiguous number of queries? Or, throwing a curve ball question out of contextual scope - even when donating a 'side-note'... Well, All that was answered in a single mention about 2D model "CTX windows" in here... and it was so Obvious!! - All I needed was someone to say it out loud for me to go "DING! THAT'S EXACTLY IT!!" I have so many ideas and I want to work on my own AI applications and investigate rabbit holes crazy flowchart and invent concepts like "In a system diagram of a computer system: Replace the word 'CPU' with "LLM" and receive as many plaudits as possible ("Replace the word 'plaudits' with "MONEY" could even work for that) See? ...Transferable Correlations. I still can't finish a project by myself but. Just one thing... If you don't mind me asking? - and this is to every-one reading / any-one interested """ How can I help? """
@MODEST5003 күн бұрын
all LLM based chat environment need to have token count indicators. honestly how could you ignore such an important features?. without it it's hard to know when we will run out context window