|
--- |
|
license: apache-2.0 |
|
datasets: |
|
- bookcorpus |
|
- wikipedia |
|
language: |
|
- en |
|
--- |
|
|
|
# BERT L12-H128 (uncased) |
|
|
|
Mini BERT models from https://arxiv.org/abs/1908.08962 that the HF team didn't convert. The original [conversion script](https://github.com/huggingface/transformers/blob/main/src/transformers/models/bert/convert_bert_original_tf_checkpoint_to_pytorch.py) is used. |
|
|
|
See the original Google repo: [google-research/bert](https://github.com/google-research/bert) |
|
|
|
Note: it's not clear if these checkpoints have undergone knowledge distillation. |
|
|
|
## Model variants |
|
|
|
| |H=128|H=256|H=512|H=768| |
|
|---|:---:|:---:|:---:|:---:| |
|
| **L=2** |[2/128 (BERT-Tiny)][2_128]|[2/256][2_256]|[2/512][2_512]|[2/768][2_768]| |
|
| **L=4** |[4/128][4_128]|[4/256 (BERT-Mini)][4_256]|[4/512 (BERT-Small)][4_512]|[4/768][4_768]| |
|
| **L=6** |[6/128][6_128]|[6/256][6_256]|[6/512][6_512]|[6/768][6_768]| |
|
| **L=8** |[8/128][8_128]|[8/256][8_256]|[8/512 (BERT-Medium)][8_512]|[8/768][8_768]| |
|
| **L=10** |[10/128][10_128]|[10/256][10_256]|[10/512][10_512]|[10/768][10_768]| |
|
| **L=12** |[**12/128**][12_128]|[12/256][12_256]|[12/512][12_512]|[12/768 (BERT-Base, original)][12_768]| |
|
|
|
[2_128]: https://huggingface.co/gaunernst/bert-tiny-uncased |
|
[2_256]: https://huggingface.co/gaunernst/bert-L2-H256-uncased |
|
[2_512]: https://huggingface.co/gaunernst/bert-L2-H512-uncased |
|
[2_768]: https://huggingface.co/gaunernst/bert-L2-H768-uncased |
|
[4_128]: https://huggingface.co/gaunernst/bert-L4-H128-uncased |
|
[4_256]: https://huggingface.co/gaunernst/bert-mini-uncased |
|
[4_512]: https://huggingface.co/gaunernst/bert-small-uncased |
|
[4_768]: https://huggingface.co/gaunernst/bert-L4-H768-uncased |
|
[6_128]: https://huggingface.co/gaunernst/bert-L6-H128-uncased |
|
[6_256]: https://huggingface.co/gaunernst/bert-L6-H256-uncased |
|
[6_512]: https://huggingface.co/gaunernst/bert-L6-H512-uncased |
|
[6_768]: https://huggingface.co/gaunernst/bert-L6-H768-uncased |
|
[8_128]: https://huggingface.co/gaunernst/bert-L8-H128-uncased |
|
[8_256]: https://huggingface.co/gaunernst/bert-L8-H256-uncased |
|
[8_512]: https://huggingface.co/gaunernst/bert-medium-uncased |
|
[8_768]: https://huggingface.co/gaunernst/bert-L8-H768-uncased |
|
[10_128]: https://huggingface.co/gaunernst/bert-L10-H128-uncased |
|
[10_256]: https://huggingface.co/gaunernst/bert-L10-H256-uncased |
|
[10_512]: https://huggingface.co/gaunernst/bert-L10-H512-uncased |
|
[10_768]: https://huggingface.co/gaunernst/bert-L10-H768-uncased |
|
[12_128]: https://huggingface.co/gaunernst/bert-L12-H128-uncased |
|
[12_256]: https://huggingface.co/gaunernst/bert-L12-H256-uncased |
|
[12_512]: https://huggingface.co/gaunernst/bert-L12-H512-uncased |
|
[12_768]: https://huggingface.co/bert-base-uncased |
|
|
|
## Usage |
|
|
|
See other BERT model cards e.g. https://huggingface.co/bert-base-uncased |
|
|
|
## Citation |
|
|
|
```bibtex |
|
@article{turc2019, |
|
title={Well-Read Students Learn Better: On the Importance of Pre-training Compact Models}, |
|
author={Turc, Iulia and Chang, Ming-Wei and Lee, Kenton and Toutanova, Kristina}, |
|
journal={arXiv preprint arXiv:1908.08962v2 }, |
|
year={2019} |
|
} |
|
``` |
