pycorrector
The pycorrector toolkit is a Python-based solution for correcting Chinese text errors using models like Kenlm, ConvSeq2Seq, BERT, MacBERT, ELECTRA, ERNIE, and GPT. It effectively manages errors such as homophones, similar-looking characters, grammatical mistakes, and proper nouns. Recent updates have incorporated Qwen2.5-based models to handle complex corrections, providing versatile solutions applicable to various industries needing precise language handling. The toolkit features fast processing, parallel training, and high error correction accuracy, all of which can be accessed conveniently through its PyPI package.
