![John Capobianco](/img/default-banner.jpg)
- 303
- 116 908
John Capobianco
Canada
Registrace 8. 07. 2013
The Bob Ross of Network Automation
Screen Speak - A multimodal AI Assistant that transforms screenshots into AI analysis
Screen Speak - Transform Your Screenshots into AI-Driven Insights
Discover the power of multimodal AI with Screen Speak, an innovative assistant that turns your screenshots into comprehensive AI analyses. In this video, we introduce Screen Speak's capabilities, showcasing both a premium and free implementation:
Introduction to Multimodal AI: Explore the potential of combining image recognition, text-to-speech, and AI analysis.
ChatGPT Implementation: Experience top-tier AI analysis with paid token-based services for unparalleled accuracy in image and text-to-speech conversion.
Ollama Implementation: Leverage free, local AI with the Llava model and Google Text-to-Speech to achieve a similar experience at no cost.
Getting Started:
Clone the Repository: Access the codebase and start your journey with Screen Speak.
Setup Your Environment: Create a virtual environment and install dependencies with pip install -r requirements.txt.
Launch Screen Speak: Start the service and begin transforming your screenshots.
Capture and Analyze: Use Print Screen or Snipping Tool to capture images and get instant AI analysis.
Automatic Audio Feedback: Enjoy seamless text-to-speech playback with a simple push of a button.
Whether you're using the free, open-source version or opting for the cloud-enabled ChatGPT experience, Screen Speak offers a flexible solution that's easy to fork and adapt for other AI models like Claude by Anthropic.
Github:
github.com/automateyournetwork/screenspeak
Discover the power of multimodal AI with Screen Speak, an innovative assistant that turns your screenshots into comprehensive AI analyses. In this video, we introduce Screen Speak's capabilities, showcasing both a premium and free implementation:
Introduction to Multimodal AI: Explore the potential of combining image recognition, text-to-speech, and AI analysis.
ChatGPT Implementation: Experience top-tier AI analysis with paid token-based services for unparalleled accuracy in image and text-to-speech conversion.
Ollama Implementation: Leverage free, local AI with the Llava model and Google Text-to-Speech to achieve a similar experience at no cost.
Getting Started:
Clone the Repository: Access the codebase and start your journey with Screen Speak.
Setup Your Environment: Create a virtual environment and install dependencies with pip install -r requirements.txt.
Launch Screen Speak: Start the service and begin transforming your screenshots.
Capture and Analyze: Use Print Screen or Snipping Tool to capture images and get instant AI analysis.
Automatic Audio Feedback: Enjoy seamless text-to-speech playback with a simple push of a button.
Whether you're using the free, open-source version or opting for the cloud-enabled ChatGPT experience, Screen Speak offers a flexible solution that's easy to fork and adapt for other AI models like Claude by Anthropic.
Github:
github.com/automateyournetwork/screenspeak
zhlédnutí: 221
Video
Reaction to Google Cloud Report The ROI of GenAI
zhlédnutí 390Před měsícem
My enthusastic reaction to Google Cloud and the National Research Group's report about ROI from GenAI Get the report here: inthecloud.withgoogle.com/roi-of-generative-ai/dl-cd.html
A free, local, automated, pyATS AI Agent
zhlédnutí 547Před měsícem
Using LangChain, Ollama, and pyATS we can 'chat' with our Cisco IOS XE devices in 25 lines of code ! No RAG or anything aside from our Agent!
Easy, free, local, AI Agents with Ollama and Langchain
zhlédnutí 1,1KPřed měsícem
AI Agents are here! And they are *easy* to get started with ! In this video I will show you how in less than 20 lines of code you can make a Math agent! And more! We will even have an AI agent make a txt file locally with "created by an agent" inside! Check this out!
Microsoft graphRAG Graphing Text and Chatting with it for free
zhlédnutí 3,4KPřed měsícem
Exploring the theory and implementing in practice Microsoft graphRAG - the next evolution of RAG ! References: Microsoft Project: GraphRAG: www.microsoft.com/en-us/research/project/graphrag/ microsoft.github.io/graphrag/ github.com/microsoft/graphrag www.microsoft.com/en-us/research/blog/graphrag-unlocking-llm-discovery-on-narrative-private-data/ www.microsoft.com/en-us/research/blog/graphrag-n...
Fine Tuning Llama3 with ReFT and Cisco IOS XE Running Configuration
zhlédnutí 948Před 2 měsíci
Fine Tuning Llama3 with ReFT and Cisco IOS XE Running Configuration 🚀 Dive into the cutting-edge world of AI fine-tuning with our latest video, "Fine Tuning Llama3 with ReFT and Cisco IOS XE Running Configuration"! 🌟 Discover the revolutionary Representational Fine-Tuning (ReFT) approach from Stanford that transforms AI fine-tuning from a process taking tens of hours to mere minutes or a few ho...
Tracer Packet licking an AP at Cisco Live animated
zhlédnutí 43Před 2 měsíci
Tracer Packet licking an AP at Cisco Live animated
Get started with Free Local AI Image Generation with StableSwarmUI
zhlédnutí 1,1KPřed 2 měsíci
Get started with Free Local AI Image Generation with StableSwarmUI
Retrieval Augmented Consensus with Kubernetes
zhlédnutí 454Před 2 měsíci
Retrieval Augmented Consensus with Kubernetes
Artificial Intelligence Microservices
zhlédnutí 1,1KPřed 2 měsíci
Artificial Intelligence Microservices
Using natural language at the command line interface to configure an interface
zhlédnutí 361Před 3 měsíci
Using natural language at the command line interface to configure an interface
Dude, it is not free by any measure! The number of calls it makes to the OpenAI service for generating graph means costs quickly mount. It is misleading to call this free for anything but toy problems. There is currently no way to get this to work fully replacing OpenAI models with local and open-source models. Please avoid misleading language in otherwise good video.
Did you watch the video? It is can be done for free with ollama
Can you please share the GitHub repo URL, where this code is hosted ? I want to try it as I have a Cisco NSO installed in Ubuntu Guest VM.
Super interesting topic, beautifully delivered. Thanks for taking the time to go through it and sharing your expertise!
Great video John! just here to show some support!
Great video! Thanks for sharing. Could you please share some models that can be used to generate images for commercial use.
Congrats on your new
Fantastic, I am very excited about the human growth AI represents. You slay the topic my friend, this is very understandable and to the point!
23:09 You can take the man out of Cisco but not the Cisco out of the man. Thanks for the up.
Thanks for the amazing content. How do we add an eval_dataset to the trainer, so we can get an idea of the validation loss aswell. Hope you can help me out here.
Thanks John!, voice is choppy but we can follow you!🙂
I have a php code project. How can I use graphRag for asking code questions and make documentation out of it.
The ai that can generator for long video or short video locally..?
I'm a newbie. My question: is Retrieval done by LLM ? I thought its done by langchain
It's both langchain is the framework and the LLM us used in the conversational retrieval chain function
where's the rest?
I subscribed! Thanks for putting out so much good information John!
I get that "not a valid tool error alot when running CrewAI agents.
Hi John - I am a network engineer trying to get into AI for network automation. Where do I start please? Do you have an introductory article or youtube video please?
I have tons of videos on my channel yes!
Thanks much
That is sooo awesome. I gonna try it with Gemma. I tried self rag once using langgraph and local model but it fail though
Any thoughts, on RAGGraph vs RAPTOR? Using qdrant db as private db?
I made a video about graphRAG yes! Great topic thank you - you were way ahead on that
what is the issue here "Error during GraphRAG setup: Command 'python -m graphrag.index --root ./ragtest' returned non-zero exit status 1.."
If you index different documents at different point of time. We end up with multiple artifacts in the output folder. How should one do a search over all outputs. Like a production level application
I just wonder if it is possible to use local embeddings running on a separate machine? By the way, it took you one hour to 'graph' your txt file on 4090, what do you think it might take on h100?
I was able to setup a proxy and use local embeddings from ollama (proxy.py script in github repo) and it worked for Global configs but not Local; I fell back to openAI for embeddings but they are more or less 'free' anyway Not sure on an H100 or A100 card I will try to find one and test
Did you override the default entity_types to extract entities specific to your domain? there are default entity_types ( [organization,person,geo,event]) defined in settings.yaml....
Not for this use case - but if you have more documentation about these settings that would be wonderful
Wow this is very cool!!!!! Thanks for that.
Hi there, how about doing a video on xRAG it is using a compression of embeddings, would really be helpful if you can explain of its workings. Thnx in advance
Cool
Great idea. Yet I have to ask if it is possible to replace this online wanDB with something local?
Yes wanDB is totally optional; I set my environment variable but there is a menu of make and account and sign in; sign in; work offline if you dont set the environment variable; you can choose option 3 and run offline - great question!
@@johncapobianco2527 so, I can imagine i can run wanDB in a local docker and use it for this?
@@artur50 I am not sure if wandb can be run locally in a container but that would be cool
@@johncapobianco2527I’ll try to fix that and let u know.
Good Job
At the start - for simple RAG You mentioned that LLM is involved in retrieval process , is that a correct statement?
Correct with the ConversationalRetreivalChain.from_llm()
This is so fucking cursed
Is that AI powered?
Powered by*😊
@@TorsTechTalk I took a still image of the rocket and used AI to animate it and AI to generate the audio of a rocket taking off!
@johncapobianco2527 very cool stuff!
First
😧
Nice stuff John!
Awesome work, thanks for sharing!
Thank you, grandpa!
The audio is cutting out a lot. Is it just me?
The audio has issues not sure what happened apologies
i will love to collaborate with you and help/assist you in any way possible. thank you for the great video. already starred your repo
I really need to say this that I look forward to testing your idea. I remember your raptor repo and it was great. Cheers!
😊
When you say ‘write the exam’ it sounds like you’re saying you made the exam for Cisco.
Apologies for the confusion - good point
Thanks for sharing ❤
Great Job
This is great. Thanks for sharing. I wish I could work on this stuff full time! Your work is very inspiring to me. For now working on setting up some local LLM with ollama on my workhorse machine. So far so good. Going to experiment with fabric next to prompt engineer. This video is giving me some ideas though too
I'm glad that at least some people could follow this, but I could not.
Very nice and easy to understand video.. I shared it on my LI
Ollama mentioned!
This was a master class. Thanks for sharing.
I really appreciate that thank you
For comparison, while you see a response time of 1m38s for the "generate" API endpoint, in my solution, I am getting 179 ms, using a single NVIDIA 3080 and a 2.3GB SLM.
I was doing this three months ago and I have been trying to reach out to you to connect and engage with you on these topics. I have lots of questions about your GitHub repos.