LangChain Basics Tutorial #1 - LLMs & PromptTemplates with Colab

[@NightFore]

Your channel is a gold mine for someone who has no idea how to use all these frameworks and whatnot. I saw a few videos and they were all crystal clear.

[@TheBluCypher]

Sam, Thank You. I've been binging through all of your videos and your content is absolute GOLD. You have no idea how much this is helping people like myself who want to start using LLMs around practical applications

[@samwitteveenai]

Thanks! Glad they are useful.

[@victorbankz5]

Nice, I love the fact that there are some people who started learning langchain before me, Meaning I can connect with those who know more than me. Do you think we can connect i need your help on some errors related to langchain.

[@ThulasiRaghavendraRao]

You nailed it. This is a very useful video for understanding why we are using Langchain and how it is useful to connect with LLM. This is what I was looking for, and I got the right information. Thank you.

[@ItsRyanStudios]

Your channel is about to explode. I've a software engineer and an avid chatgpt user for a couple of months now. This is the type of content I crave. Thank you for the valuable information.

[@samwitteveenai]

Glad you like it let me know what you want to see in the future.

[@FreestyleTraceur]

100% agreed. There's a lot of mediocre videos out there riding the recent AI/LLM hype. But pretty much all the ones I've seen have been pretty superficial - not getting into the weeds much. Sam's content here is a cut above the rest for sure.

[@samwitteveenai]

This makes me really pleased to hear. The whole reason I made this channel was I was sick of seeing a lot of channels superficially covering this stuff and not explaining things. I have been working in this field for many years and it is great to see people who want to learn and get involved, hopefully I can help out with that.

[@123ankit321]

This tutorial is so much better than most tutorials online , which are either fast paced or way less structured. Looking forward to see this community grow. Thanks a ton , Sam !!

[@MichaelMeighu]

I can't believe how good his explanations are! And its free!

[@resistance_tn]

This is some amazing content right here, another comment mentioned that this channel is gonna be huge and I must say I agree 1000% ! I really like the pace, the clear tone, examples, higher level explanation then deep diving into the code. I'm also a (10+ yoe) software engineer who's aiming to build some product leveraging LLMs (for fun) and being a complete noob in AI/ML this is the type of videos that I've been looking for ! Subscribed and will be waiting for that notification, thank you again Sam !

[@samwitteveenai]

Thanks for the kinds words. Very much appreciated.

[@simonsmashup]

Bro, your video cleared so many concepts for me. I really appreciate it.

[@iliketaiwan]

This tutorial is amazing! I hope I could have found this earlier. Deeply thanks for your excellent work

[@pauledam2174]

Thank you Sam. This is so good. Tempo is perfect. Explanation is clear! 😂

[@LesCalvin3]

Came here to say just that. Thanks, Sam!

[@aliasad8544]

This is by far one of the best videos on langchain. thanks mate for sharing this.

[@hichamzmarrou3762]

Amazing content, learning every day something new from this channel, I hope I can keep up with your video and colab notebooks. Thank you very much.

[@LearningWorldChatGPT]

Another fantastic playlist! Your tutorials are top notch... Thanks a lot, Sam !

[@srikarpamidi1946]

Incredible content-I didn’t really think I had enough coding knowledge to get into LLMs but this really made me excited to start messing around with it!

[@Haz2288]

The concept of using LLMs like this just blew the world of AI wide open for me. Massive thanks 🙏

[@HazemAzim]

Very Smooth and comprehensive Intro to Langchain .. thumbs up Thanks

[@samwitteveenai]

Glad you liked it!

[@noone-jq1xw]

I can see this channel going places. Extremely informative and in-depth. Thanks so much for the content.

[@davidoktavianusmendrofa2631]

One of marvelous video playlist 💯 of 💯

[@davidoktavianusmendrofa2631]

can you provide playlist more specific about openAI azure?

[@santoshbammidi-x3q]

excellent videos, please bring some more videos on lang chains and gen AI

[@ys621]

You dropped this 👑. So useful. Taking it easy and walking through the topic explaining everything, without treating us like idiots.

[@fojcikdominik]

Great video! Lots of knowledge

[@LouvoresPauloRicardo]

@samwitteveenai excelent playlist man! Some "deprecated" functions in code, but excelent! Thank you !

[@samwitteveenai]

Yeah a lot of the vids are well over a year old now. Anything particular you couldn't work out etc. or huge changes? I am planning to do a new updated LC vids soon

[@fra4897]

I love your content man, keep it up

[@samwitteveenai]

Thanks, much appreaciated.

Guys also do not forget this prompt in ChatGPT: I want you to act as a teacher, teaching how to use Langchain and AI to beginners. You will provide code examples using python programming language. You will give me examples and wait for my prompt for additional questions. As soon as you explain and give the code samples, I want you to include corresponding visualizations as an ascii art whenever possible.

[@MyEyeOnAi]

I love you

[@lkd982]

Thank you for the great content! Could you say how to get long form response content from langchain, or is it not possible ? I have used a script which gets langchain to learn a number of uploaded txts and use them to answer, but the answers seem to get cut off at around 100 words, even with num_outputs and max_tokens set at high levels? The question are only about 20 words long, so not running up against 4K limit? Looks like the text is getting cut off in mid-sentence sometimes

[@salouamahir8429]

thank you

[@ManiSaintVictor]

Thanks for this tutorial. Just curious for others in the chat: is the tiktoken and cohere dependency new or have I just not been using the api long enough to have noticed. It also seems like I've been getting a new "text_extensions" warning. They're not hard to fix but they feel like a new, extra step.

[@10points56]

Hey Sam, I loved your tutorials, but is there anything that No-code developer like me use langchain in no-code?

[@samwitteveenai]

Unfortunately everything am showing uses code, but there are some No Code options a i think for building with LLMs and I am sure people are making more currently.

[@adriansrfr]

What are the minimum hardware requirements?

[@Newzzysupport]

brilliant

[@hemachandhers]

can langchain support mamba LLM ?

[@ThomasGregory1789]

Is this langchain v0.1...? Thanks.

[@Jasonknash101]

Anoght great video decided to go back to the start and take a look through your videos... GREAT i build my hos fsp template. but i am getting and error around Only one of 'examples' and 'example_selector' should be provided (type=value_error) if you have any ideas as what it might be would be most welcome for the pointer... GTP does not seem to be able o help on this one...

[@landerosedgard]

cool channel. you hav ea link to notebook 1

[@lorenzoleongutierrez7927]

thanks for sharing !

[@myselfandpesit]

Hello Sam. When I type in the code in jupyter notebook I always get a ratelimiterror even though I've never used langchain before. Why is this happening?

[@samwitteveenai]

Do you have all the packages installed? Try running the Colab

[@bleo4485]

Hi Sam, thanks for this video. Is it possible to use LangChain for locally hosted LLM like Alpaca or Vicuna?

[@samwitteveenai]

Yes but the results are not as good. Check out my video on how run LangChain with HuggingFace models.

[@aischool0912]

Can we build a hotel booking bot using langchain framework with LLM?

[@samwitteveenai]

Yes what features would you want it to have?

[@marilynlucas5128]

Good Job!

[@samwitteveenai]

Thanks!

[@marilynlucas5128]

@@samwitteveenai I have the perfect business scheme to turn what you are doing to billions of dollars. Would you like to hear it?

[@ratral]

Thanks.

[@marcossantacruzv]

Hi Sam! I really like your content man! Question... Are you using a pro account in huggingface? Because in my case, I'm trying to use the same model (flan-t5-x1) with my free api-key and it give me a error due to time out. If that is the case, do you have a video with this setup in hugginface, pls?

[@samwitteveenai]

No I am not using a HF pro account. My guess is they could have removed it from the free tier.

[@noone-jq1xw]

Yeah, I'm getting the same timeout issue.

[@oktayalpkabac]

I think the models gave the same answer because the temperature was high.

[@wesleymogaka]

Gettint this error : openai.error.RateLimitError: The server had an error while processing your request. Sorry about that!

[@samwitteveenai]

This is an OpenAI error give it an hour etc and try again, you are possibly getting rate limited.

[@720_TIYASHAKUNDU]

pls pls, do new videos as most of the codes show deprecated. These videos are amazing.

[@samwitteveenai]

yeah this is close on 18 months old now. I will try to record an update soon

[@JohnMcclaned]

10:00 davinci3 isn't fun at parties

[@chaotic_music]

Help me, I'm a newbie. Do I need HuggingFace hub paid subscription for that?

[@UnicornLaunching]

1.5x speed

[@al.gharaee]

Tanx a lot. I wanted to say that the background scenes can be somehow distracting.

[@TaiNguyen-ee2bt]

the AI generated background so annoying

[@kashishvarshney2225]

openai model is not running please someone help error is still showing this error AttributeError Traceback (most recent call last) in () 1 text="what is the capital of india?" ----> 2 print(llm(text)) 6 frames /usr/local/lib/python3.10/dist-packages/langchain/llms/openai.py in _create_retry_decorator(llm, run_manager) 94 95 errors = [ ---> 96 openai.error.Timeout, 97 openai.error.APIError, 98 openai.error.APIConnectionError, AttributeError: module 'openai' has no attribute 'error'

[@koningsbruggen]

Hmm it worked at first, but it mentioned I need to call langchain with like this with the addition of chat: langchain.chat_models import ChatOpenAI but after i did that it stopped working :S Maybe you can update the colab to call it correctly? Or shall I just ignore the issue and continue calling it without 'chat'.

[@koningsbruggen]

I resolved the issue by importing a schema: from langchain.chat_models import ChatOpenAI from langchain.schema import ( AIMessage, HumanMessage, SystemMessage ) And then specifying a system message as well as an input message ( I guess the chat models don't work with just a text input but require a system message too) messages = [ SystemMessage(content="You are a helpful assistant that translates English to French."), HumanMessage(content="Translate this sentence from Fench to Dutch. e suis désolé, mais je ne peux pas traduire cette phrase car elle contient un langage offensant et inapproprié.") ]