# databrickslabs/dolly

Databricks’ Dolly, a large language model trained on the Databricks Machine Learning Platform

Repository: https://github.com/databrickslabs/dolly
Canonical: https://ross.abutalabs.com/products/dolly
Homepage: https://www.databricks.com/blog/2023/03/24/hello-dolly-democratizing-magic-chatgpt-open-models.html
Language: Python
License: Apache-2.0
License Family: permissive
Topics: databricks, gpt, chatbot, dolly
Last push: 2023-06-30T18:36:16+00:00

## Health v2 (maintenance only)
Score: 30/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 89
- inputs: {"age_days": 1258, "days_push": 1160, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 10805, forks 1139 (observed 2026-08-28T04:10:44.352747+00:00)

## What it is
Dolly is Databricks' instruction-following large language model (dolly-v2-12b), fine-tuned from EleutherAI's Pythia-12b on a ~15k record instruction corpus and licensed for commercial use. The repository provides training code, the dataset, and pointers to the model weights on Hugging Face.

## Use cases
- run an instruction-following LLM with a commercial-use license
- fine-tune an open language model on instruction data
- download dolly-v2-12b weights from Hugging Face
- experiment with instruction tuning on Pythia
- build a chatbot from an open-source LLM
- use the databricks-dolly-15k instruction dataset

## When to choose
- you need an instruction-tuned LLM whose weights are licensed for commercial use
- you want to study or reproduce instruction fine-tuning on an open base model
- you want the databricks-dolly-15k dataset for your own training

## When to avoid
- you need state-of-the-art generation quality or strong coding/math abilities
- you want a maintained model with active development
- you need low hallucination or reliable factual QA
- you want a small model for resource-constrained hardware

## Facets
- artifact type: learning-resource
- maturity: maintenance
- function: llm-training, machine-learning, chatbot
- domain: large-language-models, artificial-intelligence, machine-learning
- platform: python, cloud
- tags: instruction-tuning, pythia, huggingface, commercial-license, databricks, fine-tuning, open-weights, gpu

## Member repositories
- databrickslabs/dolly (main) score 30

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:44.352747+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 2026-08-29T17:17:41.556498+00:00, confidence not recorded.
  - readme: https://github.com/databrickslabs/dolly (fetched 2026-08-28T04:10:44.352747+00:00, sha 9f16da9136b8)
  - homepage: https://www.databricks.com/blog/2023/03/24/hello-dolly-democratizing-magic-chatgpt-open-models.html (fetched 2026-08-29T08:16:39.764531+00:00, sha 44136fa355b3)
- Data as of 2026-08-30T08:39:29.467469+00:00.