roformer

Transformer model

An enhanced transformer model with improved relative position embeddings for natural language processing tasks

Rotary Transformer

GitHub

837 stars
8 watching
51 forks
Language: Python
last commit: over 4 years ago

Related projects:

RepositoryDescriptionStars
zhuiyitechnology/roformer-v2An improved version of a transformer-based language model with enhanced speed and accuracy through structural simplification and pre-training148
zhuiyitechnology/roformer-simAn upgraded version of SimBERT with integrated retrieval and generation capabilities441
zhuiyitechnology/gau-alphaAn implementation of a transformer-based NLP model utilizing gated attention units98
tongjilibo/bert4torchAn implementation of transformer models in PyTorch for natural language processing tasks1,257
thudm/chinese-transformer-xlA pre-trained Chinese language model based on the Transformer-XL architecture.218
zhuiyitechnology/wobertA Word-based Chinese BERT model trained on large-scale text data using pre-trained models as a foundation460
lucidrains/reformer-pytorchAn implementation of Reformer, an efficient Transformer model for natural language processing tasks.2,132
fastnlp/cptA pre-trained transformer model for natural language understanding and generation tasks in Chinese482
zhuiyitechnology/t5-pegasusPretrained Chinese text generation model trained on large-scale data558
leviswind/pytorch-transformerImplementation of a transformer-based translation model in PyTorch240
yangjianxin1/ofa-chineseTransforms the OFA-Chinese model to work with the Hugging Face Transformers framework123
chrislemke/sk-transformersProvides a collection of reusable data transformation tools10
german-nlp-group/german-transformer-trainingTrains German transformer models to improve language understanding23
robostack/jupyter-rosProvides ROS support for Jupyter notebooks to enable robotics developers to create interactive and dynamic visualizations of robot behavior.592
rosinality/glow-pytorchA PyTorch implementation of Glow, a generative flow model using invertible 1x1 convolutions.522