MiniLM: Small and Fast Pre-trained Models for Language Understanding and Generation

MiniLM is a distilled model from the paper " MiniLM: Deep Self-Attention Distillation for Task-Agnostic Compression of Pre-Trained Transformers ".

Please find the information about preprocessing, training and full details of the MiniLM in the original MiniLM repository .

Please note: This checkpoint can be an inplace substitution for BERT and it needs to be fine-tuned before use!

English Pre-trained Models

We release the uncased 12 -layer model with 384 hidden size distilled from an in-house pre-trained UniLM v2 model in BERT-Base size.

• MiniLMv1-L12-H384-uncased: 12-layer, 384-hidden, 12-heads, 33M parameters, 2.7x faster than BERT-Base

We present the dev results on SQuAD 2.0 and several GLUE benchmark tasks.

Citation

If you find MiniLM useful in your research, please cite the following paper:

@misc{wang2020minilm,
    title={MiniLM: Deep Self-Attention Distillation for Task-Agnostic Compression of Pre-Trained Transformers},
    author={Wenhui Wang and Furu Wei and Li Dong and Hangbo Bao and Nan Yang and Ming Zhou},
    year={2020},
    eprint={2002.10957},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}