Please enable JavaScript.

Coggle requires JavaScript to display documents.

Generative AI, provides fundamental structure (encoder-decoder) and…

- - - - language modality
        
        Large Language Models (LLMs)
      - multi-modal (including
        language, images)
        
        Multi-modal Large Language Models (MLLMs)
  - - - Structured State Space Models (SSMs)
- - - - for developing a
        
        Generative AI Application
        
        employs one or more
  - - - STAGE 3: ADAPT & ALIGN
        
        STAGE 4: ...
        
        Prompt Engineering
        
        Fine-tuning
        
        Align with human feedback
      - options
        
        Pre-train your own model
        
        Choose and existing Foundation Model
        
        criteria
        
        Open source vs. proprietary
        
        LLM parameter size
        
        Small (< 1 Bn) to medium (1-30 Bn)
        
        for single use case
        
        1 more item...
        
        Large (100+ Bn)
        
        for use cases that are comprehensive, need a general world knowledge and need to generative to multiple tasks.
        
        1 more item...