# e-p-armstrong/augmentoolkit

Create Custom LLMs

Repository: https://github.com/e-p-armstrong/augmentoolkit
Canonical: https://ross.abutalabs.com/products/augmentoolkit
Language: Python
License: MIT
License Family: permissive
Topics: ai, dataset-generation, finetuning-llms
Last push: 2026-06-27T06:36:32+00:00

## Health v2 (maintenance only)
Score: 60/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 89, release rhythm 16, longevity 71
- inputs: {"age_days": 1006, "days_push": 67, "days_rel": 447, "gap_med": 136.0, "n_releases_24m": 3}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1864, forks 247 (observed 2026-08-28T04:05:46.155493+00:00)

## What it is
Augmentoolkit is a Python application that generates domain-expert fine-tuning datasets from uploaded documents and trains custom LLMs on them, optionally fully offline using local inference. It also produces RAG-ready datasets and can spin up an inference server.

## Use cases
- create a custom LLM trained on my own documents
- generate fine-tuning datasets from text files
- make an AI expert in a specific domain or research field
- build a RAG dataset automatically from my documents
- fine-tune an LLM offline without an API key
- create a lore expert chatbot for a fictional universe

## When to choose
- you want full control over an LLM's knowledge by fine-tuning it on your own sources
- you need synthetic instruction datasets generated locally without external APIs
- you want both a fine-tuned model and a RAG dataset from the same documents

## When to avoid
- you only need prompt-based RAG without fine-tuning
- you lack a capable GPU and don't want to rent training compute
- you need a quick no-code chatbot rather than a training pipeline

## Facets
- artifact type: application
- maturity: active
- function: llm-training, rag, data-generation, llm-inference, etl
- domain: large-language-models, machine-learning, artificial-intelligence
- platform: windows, python, self-hosted
- tags: finetuning, synthetic-data, domain-expert-llm, dataset-generation, local-inference, offline, retrieval-augmented-generation, data-engineering, linux, macos, gpu

## Member repositories
- e-p-armstrong/augmentoolkit (main) score 60

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:46.155493+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:15:36.348781+00:00, confidence not recorded.
  - readme: https://github.com/e-p-armstrong/augmentoolkit (fetched 2026-08-28T04:05:46.155493+00:00, sha 5af20865450d)
- Data as of 2026-08-30T08:39:29.467469+00:00.