#51 · Primary category： Foundation Models
Chinese-ELECTRA
Pre-trained Chinese ELECTRA（中文ELECTRA预训练模型）
Project last updated：04/19/26
GitHub Stars
1.4K
Forks
165
Contributors
3
License
Apache-2.0
Why we included this project
For teams building Chinese-language NLP systems, this release from the Harbin Institute of Technology and iFlytek joint lab offers a compact encoder worth fine-tuning. It ships several sizes of ELECTRA models, and the small variant reaches BERT-base-level quality with roughly a tenth of the parameters. The models are trained on a large Chinese corpus, with a later 180G edition that adds CommonCrawl content. You can load them through Hugging Face Transformers or PaddleHub and use them as the backbone for text classification, reading comprehension, or sequence labeling. There is also a legal-domain version trained on court documents, handy if your work touches that sector. Baseline results and fine-tuning notes are included, which helps when deciding whether ELECTRA beats the usual BERT-based Chinese encoders for your task.
Articles for this project
No articles for this project yet.
To suggest a topic or contribute an article, contact us.
Related projects in this category
transformers
🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.
CLIP
CLIP (Contrastive Language-Image Pretraining), Predict the most relevant text snippet given an image
MiniCPM-V
A Pocket-Sized MLLM for Ultra-Efficient Image and Video Understanding on Your Phone
generative-models
Generative Models by Stability AI
unilm
Large-scale Self-supervised Pre-training Across Tasks, Languages, and Modalities