# imcaspar/gpt2-ml

GPT2 for Multiple Languages, including pretrained models. GPT2 多语言支持, 15亿参数中文预训练模型

Repository: https://github.com/imcaspar/gpt2-ml
Canonical: https://ross.abutalabs.com/products/gpt2-ml
Language: Python
License: Apache-2.0
License Family: permissive
Topics: gpt-2, tpu, bert, pretrained-models, chinese, nlp, tensorflow, text-generation, colab
Last push: 2023-05-22T22:32:13+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 2493, "days_push": 1199, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1701, forks 325 (observed 2026-08-28T04:05:24.309044+00:00)

## What it is
GPT2-ML is a TensorFlow-based GPT-2 training and inference project supporting multiple languages, with a focus on Chinese. It provides 1.5B-parameter pretrained Chinese models trained on ~30G of corpus using Cloud TPUs, plus a Colab demo notebook.

## Use cases
- generate Chinese text with a pretrained GPT-2 model
- fine-tune GPT-2 on Chinese corpora using TPUs
- try a 1.5B Chinese language model in Google Colab
- train a multilingual GPT-2 model with a BERT tokenizer
- download pretrained Chinese GPT-2 checkpoints for NLP research

## When to choose
- you need a large pretrained Chinese GPT-2 model
- you want TPU-based GPT-2 training scripts in TensorFlow
- you want a quick Colab demo of Chinese text generation

## When to avoid
- you need PyTorch-based tooling or modern LLM inference stacks
- you need actively maintained code with recent updates
- you need multilingual coverage beyond Chinese in practice

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, llm-training, nlp
- domain: large-language-models, machine-learning
- platform: python, cloud
- tags: gpt-2, pretrained-models, chinese, text-generation, tensorflow, tpu, bert-tokenizer, colab, natural-language-processing, gpu

## Member repositories
- imcaspar/gpt2-ml (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:24.309044+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 2026-08-30T03:37:42.868838+00:00, confidence not recorded.
  - readme: https://github.com/imcaspar/gpt2-ml (fetched 2026-08-28T04:05:24.309044+00:00, sha 8027a61be15b)
- Data as of 2026-08-30T08:39:29.467469+00:00.