test-article

test-article

How to evaluate Generative Model

Evaluating generative model is hard
  • Use Human evaluation to determine quality of generated data
    •
  • Inception score is used to measure quality of generated images. FID is also a measure to determine distance
    •
  • KSD is way to measure distance between 2 probability distribution
    •
  • BLEU score an dPerplexity is used to measure quality of generated text specially in translation.
    •
  • There are diversity score to measure how diver the generated data is
    •


    The future of creativity is generative ai. Here are slides and deep dive for Generative AI

    Another Article

    From the blog
    Build Dataproducts

    How Dataknobs help in building data products

    Enterprises are most successful when they treat data like a product. It enable to use data in multiple use cases. However data product should be designed differently compared to software product.

    Be Data Centric and well governed

    Generative AI is one of approach to build data product

    Generative AI has enabled many transformative scenarios. We combine generative AI, AI, automation, web scraping, ingesting dataset to build new data products. We have expertise in generative AI, but for business benefit we define our goal to build data product in data centric manner.

    Spotlight

    Generative AI slides
    Learn generative AI, LLM, foundation model
  • Learn generative AI - applications, LLM, architecture
  • See best practices for prompt engineering
  • Evaluate whether you should use out of box foundation model, fne tune or use in-context learning
  • Most important - be aware of concerns, issues, challenges, risk of genAI and LLM
  • See vendor comparison - Azure, OpenAI, GCP, Bard, Anthropic. Review framework for cost computation for LLM
    •