language: en
tags:
- deberta
- deberta-v3
thumbnail: https://huggingface.co/front/thumbnails/microsoft.png
license: mit
DeBERTa: Decoding-enhanced BERT with Disentangled Attention
DeBERTa improves the BERT and RoBERTa models using disentangled attention and enhanced mask decoder. With those two improvements, DeBERTa out perform RoBERTa on a majority of NLU tasks with 80GB training data.
Please check the official repository for more details and updates.
In DeBERTa V3, we replaced the MLM objective with the RTD(Replaced Token Detection) objective introduced by ELECTRA for pre-training, as well as some innovations to be introduced in our upcoming paper. Compared to DeBERTa-V2, our V3 version significantly improves the model performance in downstream tasks. You can find a simple introduction about the model from the appendix A11 in our original paper, but we will provide more details in a separate write-up.
The DeBERTa V3 large model comes with 24 layers and a hidden size of 1024 . Its total parameter number is 418M since we use a vocabulary containing 128K tokens which introduce 131M parameters in the Embedding layer. This model was trained using the 160GB data as DeBERTa V2.
Fine-tuning on NLU tasks
We present the dev results on SQuAD 1.1/2.0 and MNLI tasks.
| Model | SQuAD 1.1 | SQuAD 2.0 | MNLI-m |
|---|---|---|---|
| RoBERTa-large | 94.6/88.9 | 89.4/86.5 | 90.2 |
| XLNet-large | 95.1/89.7 | 90.6/87.9 | 90.8 |
| DeBERTa-large | -/- | 90.7/88.0 | 91.3 |
| DeBERTa-v3-large | -/- | 91.5/89.0 | 92.0 |
Fine-tuning with HF transformers
#!/bin/bash
cd transformers/examples/pytorch/text-classification/
pip install datasets
export TASK_NAME=mnli
output_dir="ds_results"
num_gpus=8
batch_size=8
python -m torch.distributed.launch --nproc_per_node=${num_gpus} \
run_glue.py \
--model_name_or_path microsoft/deberta-v3-large \
--task_name $TASK_NAME \
--do_train \
--do_eval \
--evaluation_strategy steps \
--max_seq_length 256 \
--warmup_steps 50 \
--per_device_train_batch_size ${batch_size} \
--learning_rate 6e-6 \
--num_train_epochs 2 \
--output_dir $output_dir \
--overwrite_output_dir \
--logging_steps 1000 \
--logging_dir $output_dir
Citation
If you find DeBERTa useful for your work, please cite the following paper:
@misc{he2021debertav3,
title={DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharing},
author={Pengcheng He and Jianfeng Gao and Weizhu Chen},
year={2021},
eprint={2111.09543},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
@inproceedings{
he2021deberta,
title={DEBERTA: DECODING-ENHANCED BERT WITH DISENTANGLED ATTENTION},
author={Pengcheng He and Xiaodong Liu and Jianfeng Gao and Weizhu Chen},
booktitle={International Conference on Learning Representations},
year={2021},
url={https://openreview.net/forum?id=XPZIaotutsD}
}