Multimodal AI Agents Are Revolutionising Image & Video Analysis!

[@RizwanRizwan2R]

Absolutely brilliant Marvin. Thanks for sharing the knowledge 🎉 👍

[@MervinPraison]

Thank you

[@RizwanRizwan2R]

@@MervinPraison please look, I am getting this error: Traceback (most recent call last): line 15, in task1 = Task( TypeError: Task.__init__() got an unexpected keyword argument 'images'

[@MervinPraison]

@@RizwanRizwan2R Please upgrade to the latest version pip install -U praisonai

[@RizwanRizwan2R]

@@MervinPraison thanks a lot mervin it worked 😃

[@SonGoku-pc7jl]

thanks, this framework is una maravila :) amazing

[@finnpoitier]

Great video, thanks! Question about analyzing a video: Do you think the LLM model could create timestamps of individual scenes it identified? This could be useful for automatic video cutting and repurposing.

[@orafaelgf]

great, but wouldn't it be better to directly use the crewai framework?

[@yazanrisheh5127]

How does it exactly understand videos? Does it transcribe the video or does it cut the video into frames and connects the frame img with text?

[@maxcurrent485]

There's probably more than one way but a good idea of how it's likely being done can be found by looking up the Divot LLM research paper that came out with Divot LLM from Tencent earlier this month on Huggingface or arXiv.

[@FREDDYHulsey]

The price movement of Web3 Infinity has been positive. It's building up speed!

[@orkutmuratyilmaz]

awesome! can you make a video about connecting a Streamlit UI with Praison agents?

[@MervinPraison]

Sure

[@60pluscrazy]

Praison 🎉🎉🎉

[@adamchan4403]

Love this ❤

[@brianWreaves]

🏆 Mate you've really built an impressive tool! I take it voice is on the way???

[@RDZ333]

Hey thanks for the info! I'm brand new to all of this, I found this because I'm looking to run a LLM locally and have fluid TTS convos while watching a YT video for example, or listening to a podcast and discussing it live together. Is this possible yet with low latency? I'm chatting with GPT about it and they say yeah, but I'd like to ask you, is doing a multi-modal split possible where they can contextually process audio and video from a cpu source while recognizing my voice separately and carrying out a fairly complex convo? I'm running a 4080 mobile card which i guess can run up to 13bP well, but I'm eyeing the new 5080 too. Although it can't handle a lot more parameters, I'm wondering if the latency differences due to the architecture will be drastically better. Hope this makes sense!

[@abdulahadashraf8142]

@MervinPraison I tried the PraisonAI on Windows 10. I set the system environment variables as well as in the terminal using the 'SET' command. I tried 'gpt-3.5-turbo-0125,' but the PraisonAI always uses gpt-4o. How can i use the different models? Thanks!

[@FutureAIUpdates]

Hi Mervin Thanks for sharing...can it do 3d segmentation like on .fbx or .obj file....can you guide please?

[@TomikoSturgeon]

Web3 Infinity is about to go off the rails.

[@verifili]

deos PraisonAI support Multivector retriever!?

[@motouman3240]

For the vision Agent, if I change the llm to llama3.2 or llava, would it still work? If it does, do I still need to use OpenAI APIs?

[@motouman3240]

Hi @MervinPraison, any answers?

[@adamchan4403]

Does the video analysis work under gpt-4o-mini ?

[@MervinPraison]

Yes

[@SejalDatta-l9u]

Hi Marvin. Is your solution capable of: 1) Dynamic Inter-Agent Communication: a mechanism for creating a dynamic, conversational flow between agents. In other words, agents talking to each other. 2) Iterative Process: to create an iterative process where agents can refine their output based on feedback from other agents and/or the user? 3) Short and long term memory? Thanks

[@brianWreaves]

In a previous video he covers #1 & #2.

[@SejalDatta-l9u]

@brianWreaves thanks Brian. Could you share the links o tje video that you're referring.. either I've missed something, haven't explained myself quite right or the solution that Mervin has made wasn't quite fitting. Either way, happy new years to you all!

[@Tech--Sphere]

Thanks for the guide! Can you also include code to integrate these with Gemini and Groq APIs?

[@MervinPraison]

Sure will do

[@ibrahimVolkan-i8e]

Web3 Infinity is paving the way in self-regulated crypto! 💎💡

[@RajSingh-of1fs]

can you give video explaining your github and how we can clone and then use the agent. ALso help me to use groq model i can't able to it. make more video on coding

[@JonJon-nc1nb]

DUDE! - This is so good~ But what choices exist for piping through a UI? Even before there is a perfect (Praison UI) solution, how can we use something visually intuitive ? I know you're working it ;)

[@brianWreaves]

I think he covered a process to add a UI in a previous video, but I'm not certain...

[@JonJon-nc1nb]

@@brianWreaves he said " Thinking of integrating to UI effectively" in comments...guessing that means it has not been done yet but is in the works

[@nasamind]

Awesome

[@rohithtechguy8991]

It's super. Can you do it with Gemini Multimodel One please

[@MervinPraison]

Yes

[@prasanthm888]

Daaammmnn!!

[@EmrahSeyfullah]

Web3 Infinity is 🔥🔥🔥

[@MireilleElkins]

Web3 Infinity is Web3 Infinity.

[@TERRENCEChrist]

There are much better and safer alternative currencies, including Web3 Infinity, thus I'm no longer interested in those ones.