Dave Ebbelaar
Want to get started with freelancing? Let me help: https://www.datalumina.com/data-freelancer
Need help with a project? Work with me: https://www.datalumina.com/solutions
🔗 GitHub Gist
https://gist.github.com/daveebbelaar/d65f30bd539a9979d9976af80ec41f07
👋🏻 About Me
Hi there! I’m Dave, an AI Engineer and the founder of Datalumina. On this channel, I share practical coding tutorials to help you become better at building intelligent systems. If you’re interested in that, consider subscribing!
Source
Good tip thnkss
great yar bhoat zbrdst.
This is exactly what i needed !
Thanks !!
Why don't you just just use the json response from openai directly?
Gold
Maaaaaan…. You have THE BEST content, HANDS DOWN, for Gen AI Development. Clear, concise, every step explained, context…. Context is key… Bravo! And thanks a lot for this, it's inspiring.
Loved the content.
What are the advantages of using this instead of function calling?
thank you for sharing Dave, really interesting. I think you would love LangGraph, it is made for LLM accuracy/state managment/classification.
How do you deal with the objections of sending this 'sensitive' data to OpenAI? We are doing a project now where we have to clean the data before sending it to openAI which is a big challenge. Curious to hear other people thoughts on this…
Excellent video. Can you go into a bit more detail of how a database of this type of information might look and operate. Or any type of automation that would be involved? You mentioned sentiment or you mentioned doing analytics
You have 50 000 classes transcripts you need to do a recommendation engine. Best approach?
Congratulations this is just perfect!
combined it with fastapi to transform it to an endpoint and call in the frontend side ooooofff… faster development for machine learning web system
Great! would love to see more of these.
Searched for you on google and can't find it. I follow you for a long time. For example, if MrBeast searches him on Google, he can be found quickly.
So cool that you make such great content, with clear explanations, and are so transparent <3
You did an amazin job, thank you so much for sharing this.
Such great content. I was going to gist this and then i see that's even how you're sharing it! I wanted to get a use case for Instructor library as looked interesting, but wasnt sure what it added beyond pydantic. … and here it is. Thanks!
Very clear and real-world example too. Thank you
I tried following your script and I downloaded pip install -U instructor but I keep getting no module found instructor, have you faced this kind of errors any thoughts?
seems max_retries and response_model are not supported param for client.chat.completions.create
This timely for me and brilliant, just what I need. Thank you very much!
This video pretty much changed my life/direction yesterday. 15 years ago Ruby on Rails/ActiveRecord attracted me for very similar efficiencies you're showing with Pydantic and Instructor. I've started in the last couple months transitioning from being an SRE and previously backend engineer to working with AI models and related hackery… The approach you lay out so much appeals to me. I have had LangChain as an idea bouncing around in my head for the past month b/c everyone is saying it's cool yet not found through hard knock the derived value for me yet. I've already worked your pattern into my current project and it's so much nicer! for it
Just came across this. Great content and great explanations! I come from the traditional ML world and I'm starting to explore a bit this kind of approach.
Quick question: is there any way to ensure reproducibility using the same model and temperature? setting a random seed or something like that?
and digging a bit deeper, have you found the confidence scores to be reliable? or at least "calibratable"?
This is similar to my project. But how to productionize this. meaning .. how to monitor model performance as new data flows in. any help here is appreciated.
Watching this made me cringe. No awareness of any ML methodology. Software engineers just using AI for everything and without methodology or evaluations.
I know to turn away candidates applying for my ML Engineering roles when they have “AI Engineer” in their resumes
Outstanding! But how to call a simple tool? There is no tutorials about function calling with Instructor? Thanks!
Fast, clear, easy to understand, just right amount of information to understand the topic, but not too much to get overwhelmed. I love it, thanks a lot!
Thank you for this amazing content!
How would you evaluate this kind of system? What kind of metrics/tricks would you use?
Je hebt helemaal geen ai gemaakt. Je gebruikt gewoon chatgpt
How can this be deployed and would this be a better alternative to using ML models like BERT for text classification? This task seems really hard for a Transformer like BERT to do or am I missing something?
LLM as a classifier will give unreliable results. You need to ground it with your own dataset. Classic ML classifiers are more reliable and cheaper to deploy.
niiiice great explanation, thanks!!
New sub! Excellent presentation of a genuine use case! Very well done!
Great as alwayss!