What is LangChain? 101 Beginner's Guide Explained with Animations
A fast-paced introduction to LangChain describing its modules: prompts, models, indexes, chains, memory and agents. It is packed with examples and animations to get the main points across as simply as possible. The agent section will be covered in more depth in a following video.
💌 Link to the newsletter
practical-ai-builder.beehiiv....
⏳ Timestamps
00:00 Intro
00:04 What is it?
00:11 Where is it?
00:18 Why is it needed?
00:42 What it provides
01:28 Why connecting LLM to data and making it agentic is useful
01:43 Introducing LangChain modules
01:51 Models - Intro
01:58 Embeddings Models
02:11 Semantic Search
02:18 Open AI Embedding Model
02:35 HuggingFace's Open Source Embedding Model
03:00 Language Models
03:27 Prompts - Intro
03:44 Prompt Templates
04:12 Substitution Engine
04:23 Prompts - Common use cases
04:25 LLM Few shot learning
05:06 LLM Output parsing
06:10 Example Selectors - Motivation
06:24 Example Selectors
06:55 Chat Prompt Template
07:38 Indexes - Intro
07:46 Document Loaders
08:10 Text splitter
08:34 Vector DB PDF Ingestion Example
08:39 Vectorstores
08:58 Retrievers
09:21 Self-querying with Chroma DB
09:34 Recap
09:39 Chains
09:47 Chain with Memory
10:10 Chain use cases
10:17 Chaining Chains together
10:41 Chain
10:45 Agents
10:56 Thank you
🔗 Links
Source code: github.com/edrickdch/langchai...
LangChain: python.langchain.com/en/lates...
Self-Ask Paper: ofir.io/self-ask.pdf
ReAct Paper: arxiv.org/abs/2210.03629
💬 If you're looking for a PDF Chat Tool: pdf.ai/?via=edrick
🙏 Support the channel with a donation: paypal.me/edrickdch
Bruh you got that Fireship vibe to ya. I dig it. Can't wait to see more vids from you!
This video deserves more attentions, so I decide to leave a comment here!
Great video! I enjoyed the simplicity of the explanation
Great and clear explanation!! Congrats!!
Stoked for the agents video 🙌🏾💜
Excellent presentation...thanks
Hi Edrick, great content! Please more more more!!!
very informative and to the point
Great Presentation style mate keep it up, you can make it in the big leagues
I find Langchain to be most useful for vectorizing data and interacting with vectorDBs. I've tried to use the other prompt and chianing features but they add a huge amount of boilerplate and complexity to what is already easily solved by string interpolation. In most cases you really can't even afford to waste tokens on the bloat langchain injects into your prompts
the content and quality of the video is amazing! I liked the fireship style, the amount of information you covered was super! keep going
Keep up the good work you got another subscription from me.
Nice video, great work!
Thanks Jason!
Awesome video, thanks
amazing thank you😍
You're the best 👍😃
Banger video
Indeed
Nice 👍 tutorial
Fluffless and awesome video! Nice work!
Please teach how to translate pdf above 5mb and maintaining the order of text and photos using the langchain library
subscribed
Great
did you make this video tutorial?
bro is a second Fireship
how the holy fuck do I link my MusicGen baseten model into langchain with a chat interface?? Where does the god damn python file go?
Just make a PromptTemplate, which we wont show you an example of for audio. have fun ;)
nice intro, but annoying reading (no breaks between sentences).
What a video 🥹🥹❤️❤️❤️ the meme content 🤌🏻
Great video! I enjoyed the simplicity of the explanation