Combine MULTIPLE LLMs to build an AI API! (super simple!!!) Langflow | LangChain | Groq | OpenAI

Подписаться 425 тыс.

Просмотров 106 тыс.

50% 1

Click here to follow along: github.com/lan...
CSV file: docs.google.co...
Found this hard? Learn the fundamentals of JavaScript here: www.codewithan...
🚀 Sign up to www.codewithan... to receive access for the final code.
____
⭐ Check out my IDE here and get 1 month free: jb.gg/get_webs...
⭐ New to code and none of this is making sense? Watch my '12hr+ RU-vid Coding Bootcamp' in which you will learn HTML, CSS and JavaScript Fundamentals completely from scratch. It's on my channel and its 100% free.
⭐ If you would like to buy me a coffee, well thank you very much that is mega kind! : www.buymeacoff...
⭐ Sign up for weekly coding tips from my newsletter partnership: bit.ly/JS-tips
You can also find me on:
Twitter: / ania_kubow
Instagram: / aniakubow

Развлечения

Опубликовано:

5 окт 2024

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист

Посмотреть позже

Комментарии : 80

@gentek1074 3 месяца назад

Second! Ania you’re freaking awesome. I been playing with LangChain and LangFlow looks sick! I’m excited to join you on this session 🎉

@aniakubow 3 месяца назад

Can't wait!

@biorig 3 месяца назад

I have no idea about programming, and I went 'Monkey see, monkey do'. Thanks to Ania, saying, 'this thing here, that thing there', I was able to follow through a lot without actually knowing what it does. This is so cool.

@JagdeepSingh-or5ke 3 месяца назад

amazing tutorial, Ania. I would love to see how we can host this or turn it into a fully functioning SaaS. I'm guessing it involves proper integration of the app logic in vs code along with docker or maybe there's an easier way?

@daffertube 2 месяца назад

Yes. You'd containerize the application for scalability. You can expose a langflow as an API which may also need containerization, depending on if it contains business critical logic.

@JagdeepSingh-or5ke 2 месяца назад

@@daffertube thank you so much. I think I understand 50% of what you said. I believe Docker is famously used for containerising. I’ll figure something out. Thanks for the help!!!!!

@stell4you 3 месяца назад

I'm not sure if you know how many "experts" there are making a lot of money on training courses for this content.

@Sobieski526 2 месяца назад

Would love to see something similar for a Midjourney/Stable Diffusion API

@biorig 3 месяца назад

Thank you so much for making this easy! Greetings from a medical professional.

@thequickstartcreative 3 месяца назад

Once again, great video. I was able to get this to work. My information is a bit complex so I actually had to run the output through a script to format it properly. This does mean I can't stream the output but I noticed the output was only streaming in the playground and not with the API.

@duponce 2 месяца назад

Your content is incredible, the best I've seen so far, I'm impressed. Congratulations

@aniakubow 2 месяца назад

Thanks so much!

@RealLexable 2 месяца назад

Error unsupported message type: ???

@saarza9991 3 месяца назад

Ania is not only an extremely professional software engineer, but also very beautiful. God bless you ❤🙏

@davidef.982 2 месяца назад

All your tutorials are amazing ! Thanks a million

@aniakubow 2 месяца назад

So nice of you to say :)

@davidef.982 2 месяца назад

@@aniakubow once again, thanks to you … 🫶🏻

@thefreakguy1946 2 месяца назад

mam u are the best...thanks to GOD i found this channel

@aniakubow 2 месяца назад

So kind of you !

@RealLexable 3 месяца назад

Would be Ollama possible as well? And great work as usual❤

@rodrigosnader 3 месяца назад

Yep, it is available in Langflow :)

@aniakubow 3 месяца назад

Yessss we use it in the tutorial :)

@denisalbertocosivasquez6907 3 месяца назад

Saludos desde Lima - Perú.

@Rohittarkar175 Месяц назад

Can you please make a video about how to train model to generate inages .

@sirishkumar-m5z Месяц назад

Awesome guide on combining multiple LLMs! Another great tool to consider is Cohere, which offers versatile LLMs that are easy to integrate into your projects. #LLMs #Cohere

@changtimwu 3 месяца назад

21:13. "You can only compare vector embeddings against each other if they are created from the same LLM." This is slightly inaccurate. I would suggest saying, "if both embeddings are created from the same embedding model."

@aniakubow 3 месяца назад

That is a good point - I should have been more accurate. Thanks for pointing this out

@wesleymogaka Месяц назад

Accurate

@nirzortalukder6832 Месяц назад

Hey, Thanks a lot for the awesome video. I have a question. I created a RAG app on Langflow provided inside the Datastax website. But I cannot find the API option, which is on the bottom right corner of your screen. Why is that? Are you using the Datastax website's Langflow or a locally installed one?

@ustaticdb 3 месяца назад

Hi Ania, I’ve found very interesting a video demonstration of API calls using a proxy and Cryptography but since you launched your website that video was removed or something

@khalidelgazzar 3 месяца назад

On my watch list😊

@aniakubow 3 месяца назад

💪💪

@bluewhale0908 2 месяца назад

Awesome video tutorial. Thank you for doing this. One question, is there any way to upload a file to be trained via api to langflow file component? Thanks again.

@wesleymogaka Месяц назад

Try editing the python API behind the component or create your own. Most Components are basically function with inputs (on the right) and outputs (on the left).

@WeAreTheMakers0709 4 дня назад

a small doubt this chatbot will work on different websites i m having !! becoz all of them don't have any websites , at that time what we need to do ?can u give brief about this too ?? if possible , By the way i loved your tutorial !! All heartS from INDIA ...

@aniakubow 3 дня назад

Ah so kind of you to say!!! I am not sure if I understand the question :) - but LangFlow might be the best company to turn to on this 😃😃😃

@cliffwidener7700 2 месяца назад

Around the 20min mark: Why can’t I just put the unabridged dictionary in as the Astra database, then the computer will understand which words are more similar to each other and at the same time which words are dissimilar?

@RealLexable 2 месяца назад

also nice body language when all your asking in return while grabing your ringfinger :D..Sorry but sadly i am not that rich enough for you yet :), but highly love and appreciate your content

@nastastic 3 месяца назад

How much would this cost to run on the api side? total cost roughly when deployed to production.

@aniakubow 3 месяца назад

You can use open source LLM to cut cost, and then it just becomes a database cost that depends on your website traffic :) Datastax has a generous free tier

@iainhmunro 2 месяца назад

Followed everything you did and get e great little chatbot out of it and it didnt take long to recreate. Just a couple of questions. I took the code for embedding into a website. Most chatbots have the button at the bottom and the chatbot goes up - langflow goes down, so that is not ideal. Also, how could you add in a logo onto the chatbot ? Many thanks

@ewk4688 2 месяца назад

do we need to host langflow in our server for scalability?

@wesleymogaka Месяц назад

Yes you can, but there is a hosted version

@iainhmunro 2 месяца назад

I did get the web chatbot to load and run locally, but what is the best way to deploy on the internet ?

@kazekageoftheeast6130 3 месяца назад

would you happen to know how to integrated the vertex ai embedding? i plan to use google generative AI too. thanks ania!

@jasper2virtual 2 месяца назад

The world is so unfair, you are gorgeous and smart. But I admire you are so hardworking .

@uckurus2100 2 месяца назад

i take same error when using to mongodb atlas. i never used any data before. how my plans did finish???

@aniakubow 2 месяца назад

Please use the free server regions if you are not :)

@uckurus2100 2 месяца назад

I've never used astradb before but I still get a quota limit warning. Isn't there a free way to do this? $1000 is too much.

@aniakubow 2 месяца назад

I use the free tier. Did you follow the same steps? 🤔 some servers/regions are paid so make sure to use a free one when creating a database

@ArekMateusiak 2 месяца назад

Hi, did you try playing with Vertex AI and langflow? I seem to have issue with it :-/

@aniakubow 2 месяца назад

Oh really? The LangFlow team are pretty responsive - definitely worth getting in touch with them about this

@ArekMateusiak 2 месяца назад

@@aniakubow I will touch base with them. Maybe they will be able to help me :-) Thank you for the hint!

@sihabaakil7806 2 месяца назад

Astra db component has a another option Vector Store what do we do now

@aniakubow 2 месяца назад

Oh it has changed? Yes you would need to use the vector store for vector embeddings :)

@ewk4688 2 месяца назад

do we need Visual C++ 14.0 to run this langflow ??

@orthodox_gentleman 2 месяца назад

What? No! Langflow is python based. Very few Ai apps are built using Visual C++

@oysterboulevard6623 3 месяца назад

Hi i saw a chinese website thats use the same thing, i think they are out of a job now.

@kokoabassplayer 3 месяца назад

thank you

@markofcain1443 2 месяца назад

hi, when i try to connect everything to Astra DB, i get an error saying: 'You exceeded your current quota, please check your plan and billing details." How can i fix it?

@aniakubow 2 месяца назад

It seems you have hit your free quota :) are you using the free tier?

@markofcain1443 2 месяца назад

@@aniakubow yes :(

@markofcain1443 2 месяца назад

@@aniakubow I fixed the issue by replacing the openai embedding with cohere. But now I get an error saying: ValueError: Error initializing AstraDBVectorStore: status_code: 401, body: {'message': 'invalid api token'} What do I do?

@aniakubow 2 месяца назад

@@markofcain1443 check the api token for your Astra db token is the same that you are using for the Astra component

@markofcain1443 2 месяца назад

@@aniakubow I used the generated token for Astra components. The vs code terminal gives this error at the same time: Retrying langchain_community.embeddings.cohere.CohereEmbeddings.embed_with_retry.._embed_with_retry in 4.0 seconds as it raised UnauthorizedError: status_code: 401, body: {'message': 'invalid api token'}.

@orthodox_gentleman 2 месяца назад

This isn’t multiple LLM’s

@aniakubow 2 месяца назад

Use use OpenAI and llama and I show you how to add more :)

@its_code 3 месяца назад

❤❤❤

@WannaWatchMeCode 2 месяца назад

Delete that open api key!

@aniakubow 2 месяца назад

It’s disabled :)

@UCcdTp7XpCkVLkaRCsDcifFg 3 месяца назад

kubow realy hot

@falcongold2024 3 месяца назад

first

@jakobkoprivnik6372 3 месяца назад

You are getting more attractive in each new video, how is that possible?

@nasr-rsln 2 месяца назад

the only down side of your videos is that its very hard to focus on the coding and ignore your beautiful gorgeous eyes.

@ZaraTanaka 3 месяца назад

is it really possible to learn when you have such a bloddy droped from haven beauty infront of you. i came to learn llm better but couldnt focus, all went up my head cause i was so danm lost in her eyes. :)