Local LLM

Project Summary

Field Value

PRJ ID

PRJ-SPOKE-014

Owner

Evan Rosado

Priority

P2 (Medium)

Status

Maintenance

Repository

~/atelier/_projects/personal/ollama-local

Antora Component

N/A (non-Antora)

Antora Title

N/A

Category

Development

2026 Commits

3

Site URL

N/A (local service)

Purpose

The ollama-local project provides Docker Compose configurations for running Ollama (local LLM inference) on the Domus workstation. It includes GPU-accelerated and CPU-only compose files, start/stop/status scripts, and API usage examples.

This supports the local model fine-tuning initiative and provides offline AI capability for code assistance and document generation when cloud APIs are unavailable.

Scope

In Scope

  • Docker Compose for Ollama (GPU and CPU variants)

  • Service lifecycle scripts (start, stop, status)

  • Ollama REST API usage examples

  • Model management (pull, run, list)

Out of Scope

  • Fine-tuning workflows (covered by local-model-fine-tuning project)

  • AI-powered CLI tools (covered by Kora project)

  • Production deployment (this is a local development service)

Status

Indicator Detail

Activity Level

Maintenance — 3 commits, functional and stable

Maturity

Stable — Docker Compose and scripts working

Last Activity

2026

Key Milestone

GPU-accelerated Docker Compose with lifecycle scripts

Deployment Status

Running locally on workstation, RTX 5090 GPU

Metadata

Field Value

PRJ ID

PRJ-SPOKE-014

Author

Evan Rosado

Date Created

2026-03-30

Last Updated

2026-03-30

Status

Maintenance

Next Review

2026-06-30