Skip to content

Latest commit

 

History

History
61 lines (46 loc) · 2.7 KB

File metadata and controls

61 lines (46 loc) · 2.7 KB

00 - Overview

Composable Model Graph is a dual-language (TypeScript + Python) library ecosystem for building inspectable transformation graphs. The two implementations are kept at parity; see structure.md, with philosophy.md and development.md for why it exists and how features earn their place.

The shape

Every system built on this ecosystem follows one shape:

input → transform → state/output → evaluation → feedback
  • input - whatever enters the graph.
  • transform - a named, typed function from input to output.
  • state/output - every intermediate output is recorded as a trace step.
  • evaluation - the final output is scored into an EvaluationResult.
  • feedback - the evaluation is mapped to a FeedbackAction.

Why a graph of transforms

A single function hides its intermediate states. A model graph makes them first-class: each step is recorded with its input, output, and timing. This makes runs inspectable, which is the precondition for evaluation and feedback.

What lives where

Package Role
core The primitive layer. Types, factories, and the graph runner (sequential by default, a general DAG where a use case needs it).
math The neural-network proof. Activations, losses, dense layers.
evaluators Generic ways to turn an output into an EvaluationResult.
feedback Generic ways to turn an EvaluationResult into a FeedbackAction.
estimation Decode the best path through sequential candidate states.
terminal Project executable topology and completed traces into deterministic text.
constraints Observe caller-projected relationships and retain lossless findings.

The Python side mirrors these packages with parity (see structure.md). Linearity is a default, not a design limit: the runner is sequential by default and a general DAG when a use case needs it.

Boundaries

This repository ships only generic primitives. It is not a harness, agent framework, workflow engine, or ML framework. The end goal is to bridge toward a real harness lifecycle later - but that harness is private and is built on top of these primitives, not inside this repository.

Reading order

  1. 01 - Core primitive
  2. 02 - Neural-network architecture
  3. 03 - Error, sensitivity, feedback
  4. 04 - Harness bridge
  5. 07 - Terminal graph view
  6. 08 - Constraint findings

See also: philosophy.md, development.md, structure.md.