|
--- |
|
language: Chinese |
|
datasets: CLUECorpusSmall |
|
widget: |
|
- text: "中国的首都是[MASK]" |
|
|
|
|
|
|
|
--- |
|
|
|
# Chinese word-based RoBERTa Miniatures |
|
|
|
## Model description |
|
|
|
This is the set of 5 Chinese word-based RoBERTa models pre-trained by [UER-py](https://arxiv.org/abs/1909.05658). |
|
|
|
[Turc et al.](https://arxiv.org/abs/1908.08962) have shown that the standard BERT recipe is effective on a wide range of model sizes. Following their paper, we released the 5 Chinese word-based RoBERTa models. In order to facilitate users to reproduce the results, we used the publicly available corpus and word segmentation tool, and provided all training details. |
|
|
|
You can download the 5 Chinese RoBERTa miniatures either from the [UER-py Github page](https://github.com/dbiir/UER-py/), or via HuggingFace from the links below: |
|
|
|
| | Link | |
|
| -------- | :-----------------------: | |
|
| **Tiny** | [**2/128 (Tiny)**][2_128] | |
|
| **Mini** | [**4/256 (Mini)**][4_256] | |
|
| **Small** | [**4/512 (Small)**][4_512] | |
|
| **Medium** | [**8/512 (Medium)**][8_512] | |
|
| **Base** | [**12/768 (Base)**][12_768] | |
|
|
|
## How to use |
|
|
|
You can use this model directly with a pipeline for masked language modeling: |
|
|
|
```python |
|
>>> from transformers import pipeline |
|
>>> unmasker = pipeline('fill-mask', model='uer/roberta-base-word-chinese-cluecorpussmall') |
|
>>> unmasker("[MASK]的首都是北京。") |
|
[ |
|
{'sequence': '中国 的首都是北京。', |
|
'score': 0.21525809168815613, |
|
'token': 2873, |
|
'token_str': '中国'}, |
|
{'sequence': '北京 的首都是北京。', |
|
'score': 0.15194718539714813, |
|
'token': 9502, |
|
'token_str': '北京'}, |
|
{'sequence': '我们 的首都是北京。', |
|
'score': 0.08854265511035919, |
|
'token': 4215, |
|
'token_str': '我们'}, |
|
{'sequence': '美国 的首都是北京。', |
|
'score': 0.06808705627918243, |
|
'token': 7810, |
|
'token_str': '美国'}, |
|
{'sequence': '日本 的首都是北京。', |
|
'score': 0.06071401759982109, |
|
'token': 7788, |
|
'token_str': '日本'} |
|
] |
|
``` |
|
|
|
BertTokenizer does not support sentencepiece, so we use AlbertTokenizer here. |
|
|
|
Here is how to use this model to get the features of a given text in PyTorch: |
|
|
|
```python |
|
from transformers import AlbertTokenizer, BertModel |
|
tokenizer = AlbertTokenizer.from_pretrained('uer/roberta-base-word-chinese-cluecorpussmall') |
|
model = BertModel.from_pretrained("uer/roberta-base-word-chinese-cluecorpussmall") |
|
text = "用你喜欢的任何文本替换我。" |
|
encoded_input = tokenizer(text, return_tensors='pt') |
|
output = model(**encoded_input) |
|
``` |
|
|
|
and in TensorFlow: |
|
|
|
```python |
|
from transformers import AlbertTokenizer, TFBertModel |
|
tokenizer = AlbertTokenizer.from_pretrained('uer/roberta-base-word-chinese-cluecorpussmall') |
|
model = TFBertModel.from_pretrained("uer/roberta-base-word-chinese-cluecorpussmall") |
|
text = "用你喜欢的任何文本替换我。" |
|
encoded_input = tokenizer(text, return_tensors='tf') |
|
output = model(encoded_input) |
|
``` |
|
|
|
## Training data |
|
|
|
[CLUECorpusSmall](https://github.com/CLUEbenchmark/CLUECorpus2020/) is used as training data. Google's [sentencepiece](https://github.com/google/sentencepiece) is used for word segmentation. The sentencepiece model is trained on CLUECorpusSmall corpus: |
|
|
|
``` |
|
>>> import sentencepiece as spm |
|
>>> spm.SentencePieceTrainer.train(input='cluecorpussmall.txt', |
|
model_prefix='cluecorpussmall_spm', |
|
vocab_size=100000, |
|
max_sentence_length=1024, |
|
max_sentencepiece_length=6, |
|
user_defined_symbols=['[MASK]','[unused1]','[unused2]', |
|
'[unused3]','[unused4]','[unused5]','[unused6]', |
|
'[unused7]','[unused8]','[unused9]','[unused10]'], |
|
pad_id=0, |
|
pad_piece='[PAD]', |
|
unk_id=1, |
|
unk_piece='[UNK]', |
|
bos_id=2, |
|
bos_piece='[CLS]', |
|
eos_id=3, |
|
eos_piece='[SEP]', |
|
train_extremely_large_corpus=True |
|
) |
|
``` |
|
|
|
## Training procedure |
|
|
|
Models are pre-trained by [UER-py](https://github.com/dbiir/UER-py/) on [Tencent Cloud TI-ONE](https://cloud.tencent.com/product/tione/). We pre-train 1,000,000 steps with a sequence length of 128 and then pre-train 250,000 additional steps with a sequence length of 512. We use the same hyper-parameters on different model sizes. |
|
|
|
Taking the case of word-based RoBERTa-Medium |
|
|
|
Stage1: |
|
|
|
``` |
|
python3 preprocess.py --corpus_path corpora/cluecorpussmall.txt \ |
|
--spm_model_path models/cluecorpussmall_spm.model \ |
|
--dataset_path cluecorpussmall_word_seq128_dataset.pt \ |
|
--processes_num 32 --seq_length 128 \ |
|
--dynamic_masking --target mlm |
|
``` |
|
|
|
``` |
|
python3 pretrain.py --dataset_path cluecorpussmall_word_seq128_dataset.pt \ |
|
--spm_model_path models/cluecorpussmall_spm.model \ |
|
--config_path models/bert/medium_config.json \ |
|
--output_model_path models/cluecorpussmall_word_roberta_medium_seq128_model.bin \ |
|
--world_size 8 --gpu_ranks 0 1 2 3 4 5 6 7 \ |
|
--total_steps 1000000 --save_checkpoint_steps 100000 --report_steps 50000 \ |
|
--learning_rate 1e-4 --batch_size 64 \ |
|
--embedding word_pos_seg --encoder transformer --mask fully_visible --target mlm --tie_weights |
|
``` |
|
|
|
Stage2: |
|
|
|
``` |
|
python3 preprocess.py --corpus_path corpora/cluecorpussmall.txt \ |
|
--spm_model_path models/cluecorpussmall_spm.model \ |
|
--dataset_path cluecorpussmall_word_seq512_dataset.pt \ |
|
--processes_num 32 --seq_length 512 \ |
|
--dynamic_masking --target mlm |
|
``` |
|
|
|
``` |
|
python3 pretrain.py --dataset_path cluecorpussmall_word_seq512_dataset.pt \ |
|
--pretrained_model_path models/cluecorpussmall_word_roberta_medium_seq128_model.bin-1000000 \ |
|
--spm_model_path models/cluecorpussmall_spm.model \ |
|
--config_path models/bert/medium_config.json \ |
|
--output_model_path models/cluecorpussmall_word_roberta_medium_seq512_model.bin \ |
|
--world_size 8 --gpu_ranks 0 1 2 3 4 5 6 7 \ |
|
--total_steps 250000 --save_checkpoint_steps 50000 --report_steps 10000 \ |
|
--learning_rate 5e-5 --batch_size 16 \ |
|
--embedding word_pos_seg --encoder transformer --mask fully_visible --target mlm --tie_weights |
|
``` |
|
|
|
Finally, we convert the pre-trained model into Huggingface's format: |
|
|
|
``` |
|
python3 scripts/convert_bert_from_uer_to_huggingface.py --input_model_path models/cluecorpussmall_word_roberta_medium_seq128_model.bin-250000 \ |
|
--output_model_path pytorch_model.bin \ |
|
--layers_num 12 --target mlm |
|
``` |
|
|
|
### BibTeX entry and citation info |
|
|
|
``` |
|
@article{zhao2019uer, |
|
title={UER: An Open-Source Toolkit for Pre-training Models}, |
|
author={Zhao, Zhe and Chen, Hui and Zhang, Jinbin and Zhao, Xin and Liu, Tao and Lu, Wei and Chen, Xi and Deng, Haotang and Ju, Qi and Du, Xiaoyong}, |
|
journal={EMNLP-IJCNLP 2019}, |
|
pages={241}, |
|
year={2019} |
|
} |
|
``` |
|
|
|
[2_128]:https://huggingface.co/uer/roberta-tiny-word-chinese-cluecorpussmall |
|
[4_256]:https://huggingface.co/uer/roberta-mini-word-chinese-cluecorpussmall |
|
[4_512]:https://huggingface.co/uer/roberta-small-word-chinese-cluecorpussmall |
|
[8_512]:https://huggingface.co/uer/roberta-medium-word-chinese-cluecorpussmall |
|
[12_768]:https://huggingface.co/uer/roberta-base-word-chinese-cluecorpussmall |
