LangChain Crash Course For Beginners 2025 | LangChain V0.3 LATEST!

Поділитися
Вставка
  • Опубліковано 4 лют 2025

КОМЕНТАРІ • 44

  • @sanyamkarnavat9745
    @sanyamkarnavat9745 День тому

    Amazing dude ... although I implemented various things using Langchain the explaination for every topic was informative and I learnt something new.

  • @vaishnashanmugam5385
    @vaishnashanmugam5385 5 днів тому +1

    Underrated masterpiece with an insane teaching style!

  • @JeetKhanpara
    @JeetKhanpara 2 дні тому

    great work!

  • @shortflicks83
    @shortflicks83 4 дні тому

    I am right now just watch half of the video and this video is like a gold. What an explanation anyone who is a begineer can easily become master and able to know how to use Langchain after that.

  • @Halal_guy21
    @Halal_guy21 Місяць тому +3

    Bro this is underrated as hell, Awesome, just subscribed.

  • @Aitool-r3q
    @Aitool-r3q Місяць тому +7

    00:04 - Introduction to LangChain for building AI applications.
    02:11 - LangChain utilizes templates, chains, RAGs, and agents for dynamic AI workflows.
    06:27 - LangChain enables effortless integration of LLMs with real-world applications.
    08:25 - Creating an API key and setting up a Python development environment.
    12:46 - LangChain provides a unified interface for communicating with various LLM APIs.
    14:51 - LangChain enhances context management and scalability for AI workflows.
    19:13 - Configuring API keys in your environment for OpenAI models.
    21:17 - Managing OpenAI account balance for LangChain usage.
    25:02 - Using LangChain to create concise responses for engaging Instagram posts.
    26:53 - Understanding message types is crucial for building dynamic AI applications.
    30:33 - Real-time terminal conversation with an LLM mimicking ChatGPT locally.
    32:26 - Implementing dynamic chat history for AI interactions.
    36:03 - Integration of AI with Firestore for message storage and retrieval.
    37:58 - Setting up Firebase Firestore for application development.
    41:43 - Initializing Firestore client for chat message storage.
    43:29 - Understanding LangChain's prompt templates enhances application efficiency.
    47:24 - LangChain manages prompt templates for better LLM interaction.
    49:20 - LangChain enables dynamic message handling and task chaining.
    53:00 - Introduction to parallel and conditional chaining in task execution.
    54:50 - Exploring LangChain's efficient task chaining with prompt templates.
    58:34 - Chains streamline coding by reducing complexity for task management.
    1:00:24 - Creating a workflow with runnable Lambdas in LangChain.
    1:04:05 - Understanding Runnable Sequence Class for task chaining in LangChain.
    1:05:54 - Chaining in LangChain simplifies data handling through various methods.
    1:09:33 - Understanding sequential chaining in LangChain with practical examples.
    1:11:21 - Using LangChain to analyze movie critiques through parallel tasks.
    1:15:03 - Combining results from parallel chains for analysis.
    1:16:48 - Implementing parallel and conditional chaining for API interactions.
    1:20:31 - Understanding feedback classification and branching logic in LangChain.
    1:22:29 - Handling user feedback with appropriate response strategies.
    1:26:08 - RAG improves document accessibility for LLMs while addressing context limits.
    1:27:57 - RAG enhances LLMs by efficiently retrieving relevant document sections.
    1:31:43 - Context windows limit token processing in large documents.
    1:33:43 - Chunking enables efficient querying of relevant information based on user prompts.
    1:37:30 - Vector embeddings represent relationships between words in multi-dimensional space.
    1:39:26 - Understanding embeddings and vector databases for text processing.
    1:43:10 - Implementing document loading and embedding in LangChain.
    1:44:56 - Setting up local vector stores for document embeddings.
    1:48:49 - Chunking text with overlap improves context understanding.
    1:50:44 - Embedding and storing texts in a vector database.
    1:54:35 - LangChain retrieves top relevant chunks based on similarity scores.
    1:56:29 - Understanding the importance of parameter adjustments in search results.
    2:00:19 - Storing multiple book documents in a Chroma DB with metadata.
    2:02:17 - Load embedded data and query the database for relevant information.
    2:06:00 - Integrating user questions with relevant document chunks for accurate LLM responses.
    2:07:55 - Understanding Dracula's Castle and its connection to LLMs.
    2:11:53 - Agents enhance AI decision-making by autonomously selecting tools for tasks.
    2:14:05 - AI agents utilize the REACT pattern for problem-solving.
    2:18:10 - Creating an agent enhances LLM capabilities with external tools.
    2:20:25 - Creating tools and agents for LangChain applications.
    2:24:27 - Creating a Python function to return the current system time.
    2:26:46 - Understanding tool integration in LangChain for effective prompt handling.
    2:31:10 - React agents efficiently handle complex queries through reasoning and multiple cycles.
    2:33:10 - Providing tools enhances LLM efficiency and safety in operations.

  • @mohanb02
    @mohanb02 24 дні тому +1

    absolutely fantastic! For a beginner in AI as well as Agents framework, you have fed the essence of it! Looking forward to the upcoming videos too. Thanks a ton Harish.

  • @shahoodsajid9593
    @shahoodsajid9593 27 днів тому +1

    This tutorial is very helpful for someone who is just starting out as an AI engineer i.e ME. It is very concise and most importantly up-to-date. Thank you Harish.

  • @adinathkhamkar283
    @adinathkhamkar283 20 днів тому +1

    now i understood rag application workflow,thanks

  • @JayaBhukta
    @JayaBhukta Місяць тому +1

    That's exactly what I was looking for 😊

  • @harish_neel
    @harish_neel  Місяць тому +5

    Pulled together some really great open-source content from Chris, Brandon and other AI agent creators, plus added my own explanations/analogies from my 2+ years of building over 25+ LangChain apps to make it easily digestible for beginners!
    Broke down everything from scratch, plus included real projects you can build.
    Stuck on something? Drop a comment - I'm here to help!

  • @DevD-q9c
    @DevD-q9c Місяць тому

    cover 15 mins and you totally crush it my man love the you explaining keep going and we need some other like pydanticai for agents with pineconedb or any other vector db

  • @neelimaavula-e3e
    @neelimaavula-e3e 17 днів тому

    Wonderful explanation. Found it really helpful. Thank you so much!

  • @hobbycoding7056
    @hobbycoding7056 28 днів тому

    This is the tuitorial I was searching recently. I loved the explanation and subscirbed for more.

    • @harish_neel
      @harish_neel  28 днів тому

      @@hobbycoding7056 glad it helped you!

  • @tusharpatil1957
    @tusharpatil1957 20 днів тому

    great content, neatly explained with examples, very obscure, Thanks!!!
    Please do a similar courses on upcoming tech stuffs!!!

  • @vasudhamusicdiary2802
    @vasudhamusicdiary2802 24 дні тому

    Amazing content and articulation/details. So useful. Thanks a lot Harish. Wishing you the best

  • @deepak5074
    @deepak5074 Місяць тому

    Thanks sir keep continue such video series

  • @ferramatis
    @ferramatis 22 дні тому

    A very nice video, thank you!

  • @akashkunwar
    @akashkunwar 27 днів тому

    Great video thanks for creating awsome videos.

  • @DevD-q9c
    @DevD-q9c 12 днів тому

    sir need part 2 with full production grage agents using langchain. Thanks you sir

    • @harish_neel
      @harish_neel  12 днів тому

      Will upload next few videos by end of month 🙂

    • @DevD-q9c
      @DevD-q9c 11 днів тому

      @@harish_neel Thanks sir

  • @resuleliyev186
    @resuleliyev186 День тому

    So i m begginer and i have been learning C# 3 month and i have some issue and i have like this project, who can help to me solve this problem

  • @akg_ak
    @akg_ak 22 дні тому

    how to train rag model i have pdfs and want to train rag model, i have already made it from scratch using langchain gpt model and gradio interface and my pdfs converted into vector store fAISS which produce output while querying related to pdf how to know accuracy percentage how good is model. Will you please help me?

  • @hooman9698
    @hooman9698 11 днів тому

    That bg

  • @waheedweins
    @waheedweins 26 днів тому

  • @aniketdey800
    @aniketdey800 Місяць тому +1

    don't we need pay for using openai's keys. i mean we can generate them, sure, but isn't billing details necessary if we wan't some response

    • @harish_neel
      @harish_neel  Місяць тому

      @@aniketdey800 yes, when I make the first api call to the model, I’ll explain how to set up billing too during the first chat models section

    • @manishgaurav5129
      @manishgaurav5129 Місяць тому

      We can use google gemini model as well, langchain supports that too, so no need to pay for openai embedding or openai gpt models, you can use google embedding and good gemini model as well

  • @sagardeka2158
    @sagardeka2158 18 днів тому

    I don't know why not use simple formatted string of python instead of prompt template. There is no extra advantage.

    • @harish_neel
      @harish_neel  18 днів тому

      True. When building workflows, it's always some prompt repeatedly used, so I'm used to writing prompt templates. But not necessary always

  • @SafdarAliShah_TimeTraveler
    @SafdarAliShah_TimeTraveler Місяць тому

    🥇

  • @StudyStudioIAS
    @StudyStudioIAS 12 днів тому

    HARSH REALITY OF IT AND TECH JOBS TODAY:
    It will take you 6 months to learn all this new tech and master it, and by the time you are confident in the stack and start searching the jobs, THE TECHNOLOGY AND THE ENTIRE STACK WILL BE OBSOLETE WITH NO MORE JOBS OR HIRING.
    I know its a good technology what you are teaching, BUT WHATS THE GUARANTEE IT WILL NOT LEAD TO A LAY OFF? The answer is there is no guarantee..
    People dont wont skill anymore, people want a stable job and a constant income. That’s the skill that will put food on your plate.

  • @BilalAnwar-pk5pf
    @BilalAnwar-pk5pf Місяць тому

    Bhai AI agent create Karna sekhado

    • @harish_neel
      @harish_neel  Місяць тому +1

      @@BilalAnwar-pk5pf yes, I’ll be covering that towards the end of the course, Bilal

  • @veinzo1jsx
    @veinzo1jsx Місяць тому

    Bro this kiss is for you 💋

  • @contactteam-t6y
    @contactteam-t6y 24 дні тому

    Dear Harish Neel, Your content is extremely good, i highly appriciate it , thanks for such a valuable content.