# openai/finetune-transformer-lm

Code and model for the paper "Improving Language Understanding by Generative Pre-Training"

Repository: https://github.com/openai/finetune-transformer-lm
Canonical: https://ross.abutalabs.com/products/finetune-transformer-lm
Homepage: https://s3-us-west-2.amazonaws.com/openai-assets/research-covers/language-unsupervised/language_understanding_paper.pdf
Language: Python
License: MIT
License Family: permissive
Topics: paper
Archived: true
Last push: 2019-01-25T09:52:16+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3005, "days_push": 2777, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2308, forks 521 (observed 2026-08-28T04:06:35.917458+00:00)

## What it is
Reference code and model for OpenAI's paper 'Improving Language Understanding by Generative Pre-Training' (the original GPT-1), implemented in Python with TensorFlow. It reproduces the ROCStories Cloze test results and is explicitly archived with no further updates.

## Use cases
- reproduce the ROCStories cloze results from the GPT-1 paper
- study the original generative pre-training + fine-tuning approach
- fine-tune a pretrained transformer language model on a story completion task
- understand how GPT-1 was originally implemented

## When to choose
- you need the exact historical code for the GPT-1 paper
- you are doing research on early transformer LM fine-tuning methods

## When to avoid
- you want a maintained library for fine-tuning modern language models
- you need deterministic training results
- you want production-ready NLP tooling or current model architectures

## Facets
- artifact type: library
- maturity: abandoned
- function: machine-learning, deep-learning, nlp, llm-training, llm-inference
- domain: machine-learning, deep-learning, large-language-models
- platform: python
- tags: gpt-1, transformer, generative-pretraining, fine-tuning, research-code, archived, rocstories, paper-code, natural-language-processing

## Member repositories
- openai/finetune-transformer-lm (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:35.917458+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 2026-08-30T02:39:38.057687+00:00, confidence not recorded.
  - readme: https://github.com/openai/finetune-transformer-lm (fetched 2026-08-28T04:06:35.917458+00:00, sha dfb3ca93b417)
- Data as of 2026-08-30T08:39:29.467469+00:00.