$30 off During Our Annual Pro Sale. View Details »

Information Retrieval and Latest AI Trends

Information Retrieval and Latest AI Trends

AbdulMajedRaja RS

April 11, 2023
Tweet

More Decks by AbdulMajedRaja RS

Other Decks in Technology

Transcript

  1. Information Retrieval
    Latest AI Trends
    AbdulMajedRaja RS

    View Slide

  2. Information Retrieval
    01
    The Rise of LLMs
    02
    Semantic Search made easy
    03
    Local AI Models
    04
    AGI & The End
    05

    View Slide

  3. Information Retrieval
    is the science of searching for information in a
    document, searching for documents themselves, and
    also searching for the metadata that describes
    data, and for databases of texts, images or sounds.

    View Slide

  4. Information Retrieval
    Where is it Today
    01
    x
    y

    View Slide

  5. Then Now

    View Slide

  6. Then Now

    View Slide

  7. The Rise of LLMs
    LLM - Large Language Model
    02
    x
    y

    View Slide

  8. 2. More
    Compute
    4 .Open
    Source
    3. More
    Data
    1. Better
    Models

    View Slide

  9. 2. Instruct
    Fine-tuned
    models
    1. LLMs
    3. RLHF

    View Slide

  10. GPT-4
    IR Demo

    View Slide

  11. Semantic Search
    03
    x
    y

    View Slide

  12. Language models give computers
    the ability to search by
    meaning and go beyond
    searching by matching
    keywords.
    This capability is called
    semantic search.

    View Slide

  13. Local AI Models
    Running LLMs- Locally on your PC
    04
    x
    y

    View Slide

  14. ● LLama from Meta
    ● Dolly from DataBricks
    ● GPT4All from Nomic AI
    GPT-3.5 Level Models

    View Slide

  15. ● Stable Diffusion
    AI Image Generation Models

    View Slide

  16. AI Image Generation Models
    from diffusers import StableDiffusionPipeline, DPMSolverMultistepScheduler
    model_id = "stabilityai/stable-diffusion-2-1"
    # Use the DPMSolverMultistepScheduler (DPM-Solver++) scheduler here instead
    pipe = StableDiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.float16)
    pipe.scheduler = DPMSolverMultistepScheduler.from_config(pipe.scheduler.config)
    pipe = pipe.to("cuda")
    prompt = "a photo of an astronaut riding a horse on mars"
    image = pipe(prompt).images[0]
    image.save("astronaut_rides_horse.png")

    View Slide

  17. Hugging Face Models

    View Slide

  18. Hugging Face Spaces

    View Slide

  19. AGI and The End
    AGI- Artificial General Intelligence
    04
    x
    y

    View Slide

  20. Resources
    ● Slide Template - Slidesgo.com
    ● Semantic Search Cohere -
    https://docs.cohere.ai/docs/semantic-search
    ● Huggingface.co

    View Slide