strands-decider
View on GitHubA small, fast decision model, or system one model, for agentic workflows. Pick between options or rate on a scale faster than an LLM, with a calibrated confidence on every decision.
A small decision model and Python serving package for calibrated yes/no, choice, and scoring tasks in agent workflows. It replaces text generation with a pointer readout for faster inference and supports text and image inputs.
Use Cases
Route tasks to the right model or teamSelect an agent toolValidate tool-call argumentsClassify safety or policyScore model outputsTriage incoming requestsCheck whether an answer is adequate
Built With
- Language
- Python
- Frameworks
- PyTorch · Transformers · PEFT · FastAPI · Strands Agents SDK · MLX
Tags
decision model · agentic workflows · classification · scoring · confidence calibration · model routing · tool selection · guardrails · low-latency inference · multimodal