namisan/mt-dnn
A PyTorch-based multi-task deep learning framework that fine-tunes pre-trained language models like BERT across diverse NLP tasks.
Not currently ranked — collecting fresh signals.
star history
The MT-DNN repository implements a multi-task learning approach for natural language understanding using deep neural networks. It builds on pre-trained language models (BERT) and jointly trains across multiple NLP tasks including machine reading comprehension, named entity recognition, and ranking. The project includes implementations of adversarial training, knowledge distillation, and robust fine-tuning techniques for pre-trained models.
Frequently asked
- What is namisan/mt-dnn?
- A PyTorch-based multi-task deep learning framework that fine-tunes pre-trained language models like BERT across diverse NLP tasks.
- Is mt-dnn open source?
- Yes — namisan/mt-dnn is open source, released under the MIT license.
- What language is mt-dnn written in?
- namisan/mt-dnn is primarily written in Python.
- How popular is mt-dnn?
- namisan/mt-dnn has 2.3k stars on GitHub.
- Where can I find mt-dnn?
- namisan/mt-dnn is on GitHub at https://github.com/namisan/mt-dnn.