# shibing624/pycorrector

pycorrector is a toolkit for text error correction. 文本纠错，实现了Kenlm，T5，MacBERT，ChatGLM3，Qwen2.5等模型应用在纠错场景，开箱即用。

Repository: https://github.com/shibing624/pycorrector
Canonical: https://ross.abutalabs.com/products/pycorrector
Homepage: https://www.mulanai.com/product/corrector/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: spelling-errors, error-correction, pycorrector, kenlm, error-detection, csc, macbert4csc, t5
Last push: 2026-07-25T14:42:02+00:00

## Health v2 (maintenance only)
Score: 84/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 94, release rhythm 63, longevity 100
- inputs: {"age_days": 3107, "days_push": 39, "days_rel": 39, "gap_med": 206, "n_releases_24m": 4}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 6515, forks 1161 (observed 2026-08-28T04:09:44.847595+00:00)

## What it is
pycorrector is a Python toolkit for Chinese text error correction, supporting phonetic, visual, and grammar error types with models including Kenlm, MacBERT, T5, ChatGLM3, and Qwen2.5. It offers ready-to-use correctors and pretrained models for spelling and grammar correction tasks.

## Use cases
- correct spelling errors in chinese text
- fix grammar mistakes in chinese sentences
- proofread text from pinyin input or speech recognition
- correct ocr output with similar-looking character errors
- correct search engine queries
- fine-tune llm models for chinese text correction

## When to choose
- you need chinese-specific spelling and grammar correction out of the box
- you want to compare or swap multiple correction models (kenlm, bert-based, llm-based)
- you need pretrained chinese correction models from huggingface

## When to avoid
- you need error correction for non-chinese languages
- you need a lightweight spell checker without ml model dependencies
- you cannot run gpu inference for the llm-based correctors

## Facets
- artifact type: library
- maturity: active
- function: nlp, machine-learning, llm-inference, parser
- domain: large-language-models, machine-learning
- platform: python, cross-platform
- tags: text-correction, chinese-nlp, spell-check, grammar-correction, macbert, kenlm, t5, qwen, natural-language-processing

## Member repositories
- shibing624/pycorrector (main) score 84

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:44.847595+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 2026-08-29T17:44:26.654907+00:00, confidence not recorded.
  - readme: https://github.com/shibing624/pycorrector (fetched 2026-08-28T04:09:44.847595+00:00, sha e0adc8ad6811)
  - homepage: https://www.mulanai.com/product/corrector/ (fetched 2026-08-29T08:40:49.982160+00:00, sha c3cdaacad24d)
  - registry_pypi: https://pypi.org/pypi/pycorrector/json (fetched 2026-08-29T08:40:49.991515+00:00, sha 7aefe2c24e6f)
- Data as of 2026-08-30T08:39:29.467469+00:00.