Another well constructed, concise and massively informative video. I'm brand new to both AI and Python and I find your videos extremely useful. Thank you!
Thanks for sharing. This video has been super informative and helpful! I was easily able to follow along but I ran into issues with implementing Airtable, so I omitted that step and the bot worked like a charm. Question: Is there a way that I could edit the script so the bot would include a random photo that I have in a folder (of about 1500 photos) in each tweet/post?
great video Greg! thanks for this i want to know that to retrieve the mentions of the twitter account does the Account requires to be upgraded or the free plan account can work for this ?
@@DataIndependent thank you for the response and yes it is not available for the free version as per documentation you are only allowed to create or delete post.
Hey Greg, I've been following your LangChain tutorials for a while. I've started playing with the library a bit on my free time. Do you think you could make a video on how to use LangChain to extract information from large JSON files (that probably don't fit in a single prompt) without loosing too much context so then we can execute natural language queries on them. Also if you're up for it I was thinking of doing a similar thing for HTML where the chain would extract the most relevant information from a HTML file and create some kind of structured document or documents that could be queried using natural language queries. Great videos by the way, keep it up!
Hey @@DataIndependent, sorry for answering so late didn't get notified from your response. For the case of HTML files I was thinking of creating an AI using LangChain that would do web search by crawling into websites and retreiving the relevant information to fulfill a request from the user. So to do that I thought it would be useful to parse & store the HTML info in natural language form so we could query it using Vector Databases for example. This way we could scrape 10 or even 100 sites and then query the DB to retrieve only the relevant information. But I believe that if we store the sites as HTML the Vector DB is not going to work very well. Does this make sense?
@@DataIndependent I was also wondering if you could do a video where you explain how to use other LLMs and how to properly select the right LLM (for example from HuggingFace) instead of just using OpenAI which can be very expensive.
That Josh Hart tweet caught me off guard 🤣. Great video!
Another well constructed, concise and massively informative video. I'm brand new to both AI and Python and I find your videos extremely useful. Thank you!
Thanks for always making practical tutorials. It helps a lot
Thank you so much, this is one of the best AI channels I've found, It has really helped me a lot in my own projects.
Nice! Glad to hear it thank you
this video is masterfully made, great work
good to see it
Thank you for sharing!
love this !!!
awesome project :)
So, it is possible that some Twitter profiles that are commenting and responding on people's tweets are also AI bots?
Thanks for sharing. This video has been super informative and helpful! I was easily able to follow along but I ran into issues with implementing Airtable, so I omitted that step and the bot worked like a charm.
Question: Is there a way that I could edit the script so the bot would include a random photo that I have in a folder (of about 1500 photos) in each tweet/post?
great video Greg! thanks for this i want to know that to retrieve the mentions of the twitter account does the Account requires to be upgraded or the free plan account can work for this ?
I believe the functionality here starts to dip into the paid side
@@DataIndependent thank you for the response and yes it is not available for the free version as per documentation you are only allowed to create or delete post.
Hey Greg, I've been following your LangChain tutorials for a while. I've started playing with the library a bit on my free time.
Do you think you could make a video on how to use LangChain to extract information from large JSON files (that probably don't fit in a single prompt) without loosing too much context so then we can execute natural language queries on them. Also if you're up for it I was thinking of doing a similar thing for HTML where the chain would extract the most relevant information from a HTML file and create some kind of structured document or documents that could be queried using natural language queries.
Great videos by the way, keep it up!
Nice! I can think about it!
What’s the actual problem or use case you’re trying to solve for?
Hey @@DataIndependent, sorry for answering so late didn't get notified from your response. For the case of HTML files I was thinking of creating an AI using LangChain that would do web search by crawling into websites and retreiving the relevant information to fulfill a request from the user. So to do that I thought it would be useful to parse & store the HTML info in natural language form so we could query it using Vector Databases for example. This way we could scrape 10 or even 100 sites and then query the DB to retrieve only the relevant information. But I believe that if we store the sites as HTML the Vector DB is not going to work very well. Does this make sense?
@@DataIndependent I was also wondering if you could do a video where you explain how to use other LLMs and how to properly select the right LLM (for example from HuggingFace) instead of just using OpenAI which can be very expensive.
I did not know airtable nor railway so that is a new area for me for deployment.
Nice - ya railway is a pretty easy set up. I've had fun with it so far
What is prompt injection?
Greg thanks. Could you create a tutorial with a complete deployment of one of your llm models?
Thanks Runy - what do you mean complete deployment? Like self-hosting a model on your own hardware in the cloud?
Yes. I mean deploying it in the cloud on Railway. Step by step. So it can be used.
I don't trust a coder short on spaces.