Thank you for sending your enquiry! One of our team members will contact you shortly.
Thank you for sending your booking! One of our team members will contact you shortly.
Duration 7 hours
Course Outline
Transitioning from Text Generation to Typed Decisions
- Understanding the specific outputs of decision models
- Distinguishing system-one decisions from reasoning LLMs
- Analyzing the cost and latency benefits of bounded decision-making
The Landscape of Open-Weight Decision Models
- AWS Labs’ Strands Decider 2B
- The JEV and GEV open-weight model families
- Other frameworks including Pplx-Decider, Liquid D1, and Clef
- Comparing licensing terms and deployment considerations
Architectural Deep Dive: Inside a Decision Model
- Utilizing an LLM 'torso' without a traditional language modeling head
- Scoring candidate options via the pointer head mechanism
- Deriving confidence scores from model outputs
Practical Application: Running Your First Decisions
- Local setup on CPU and consumer-grade GPU hardware
- Crafting effective questions and defining option sets
- Processing multiple queries within a single pass
- Configuring thresholds and fallback behaviors
Evaluating Decision Quality
- Overview of JevBench and other decision-specific benchmarks
- Balancing accuracy with calibration
- Using the Brier score to gauge confidence reliability
- Recognizing scenarios where decision models are not the optimal tool
Requirements
- Foundational knowledge of Large Language Models and prompt-based applications.
- Proficiency with command-line interfaces and Python scripting.
- No prior experience in machine learning training is necessary.
Target Audience
- AI engineers assessing decision models for agent-based workflows.
- Software developers looking to replace broad LLM calls with constrained, bounded decisions.
- Solution and platform architects designing advanced agent architectures.
- Technical product managers evaluating the suitability of decision model options.
