Build a private R.A.G. Assistant with CrewAI, Ollama, & AgentOps

Поділитися
Вставка
  • Опубліковано 31 гру 2024

КОМЕНТАРІ • 7

  • @arjuns5374
    @arjuns5374 13 годин тому +1

    damn brooo thabks thaaaaaaaankssss a loottt

    • @cmd_labs
      @cmd_labs  13 годин тому

      Appreciate it!
      Great to meet you @arjuns5374

  • @ajmgdaj
    @ajmgdaj День тому +1

    7 mins in, and you haven't said what r.a.g actually means...

    • @cmd_labs
      @cmd_labs  День тому

      Great to meet you @ajmgdaj!
      RAG = Retrieval Augmented Generation
      It's the general technique of searching for pieces of data that are related to an LLM prompt...
      By including relevant pieces of data, you influence the LLM to generate outputs that are more relevant to a prompt...
      In this example we are asking for a summary of an ebook that was NOT included in the training data of the LLM being used
      So we need to extract the most important takeaways from the ebook before including them in a prompt to the LLM asking for a synthesis of them
      Is this video the first time you are hearing the term R.A.G.?

    • @cmd_labs
      @cmd_labs  День тому +1

      @ajmgdaj
      Also checkout this video on the channel explaining the concept of RAG...
      ua-cam.com/video/iXuvkzQqwqk/v-deo.html

    • @ajmgdaj
      @ajmgdaj День тому +1

      @@cmd_labs I know the concept. That it is called rag was new to me. I was able to google it, but with that abbreviation that could have been frustrating. I think with these new terms of art, one should opt for clarity over brevity. Could have meant anyhting really. Thx for the info! Nice that this is so easy now!

    • @Earzone63
      @Earzone63 10 годин тому

      ​@@ajmgdaj nar once i read into the concept captured my imagination personally i am specificallly looking for implementations that refer to the concept as RAG and not say 'knowledge management'