ML Engineering And Infra on tab-foundry Docs

Artifacts & Inference

Mon, 01 Jan 0001 00:00:00 +0000

Use this page when the question is about the files and contracts this repo owns.

tab-foundry turns manifests, corpora, and training configs into run directories, checkpoints, benchmark results, and export bundles. This repo stops at producing validated handoff artifacts; downstream runtime systems own long-lived serving.

Start With

ML Engineering: the fastest operational route through repo ownership, artifacts, and verification flows
Workflows: command syntax and artifact expectations
Inference Contract: export-bundle schema and runtime handoff boundary
Getting Started: repo orientation and shared terminology

Key Questions

What does a training run emit?
What is the difference between a manifest, a checkpoint, and an export bundle?
How do I inspect one run or validate one bundle?
What does this repo own, and what belongs in downstream runtime systems?

Workflows

Mon, 01 Jan 0001 00:00:00 +0000

Overview

Use this runbook when you need command syntax, artifact expectations, or the smallest safe verification slice.

Use these alongside this runbook:

README.md for repo overview and quickstart
CONTRIBUTING.md for review expectations
Codebase Navigation for ownership
Dataset Curation for license policy
Inference Contract for export/runtime details
program.md for sweep policy

Environment And Verification

Python 3.14 is pinned in .python-version.
Use ./scripts/dev for bootstrap, doctor, verification, and Iris smoke.
Use the packaged CLI for everything else.

Bootstrap the repo-local environment:

Inference Contract

Mon, 01 Jan 0001 00:00:00 +0000

Use this contract when you need to know what this repo hands to downstream runtime systems and what must remain stable.

Use these alongside this contract:

quickstart: README.md
workflow runbooks: docs/workflows.md
design decisions and repo structure: docs/development/design-decisions.md
architecture reference: docs/development/model-architecture.md
model config reference: docs/development/model-config.md
canonical roadmap: docs/development/roadmap.md

Planning and architecture rationale live under docs/development/, while this page stays focused on the exported bundle and the validation boundary that downstream consumers rely on.