Update README.md

README.md

@@ -1 +1,86 @@
-模型正在训练中，敬请期待~~
+---
+language: Chinese
+datasets: CLUECorpusSmall
+widget: 
+- text: "内容丰富、版式设计考究、图片华丽、印制精美。[MASK]纸箱内还放了充气袋用于保护。"
+
+
+---
+
+# Chinese Pegasus
+
+## Model description
+
+This model is pre-trained by [UER-py](https://arxiv.org/abs/1909.05658).
+
+## How to use
+
+You can use this model directly with a pipeline for text2text generation :
+
+```python
+>>> from transformers import BertTokenizer, PegasusForConditionalGeneration, Text2TextGenerationPipeline
+>>> tokenizer = BertTokenizer.from_pretrained("uer/pegasus-base-chinese-cluecorpussmall")
+>>> model = PegasusForConditionalGeneration.from_pretrained("uer/pegasus-base-chinese-cluecorpussmall")
+>>> text2text_generator = Text2TextGenerationPipeline(model, tokenizer)  
+>>> text2text_generator("内容丰富、版式设计考究、图片华丽、印制精美。[MASK]纸箱内还放了充气袋用于保护。", max_length=50, do_sample=False)
+    [{'generated_text': '书 的 质 量 很 好 。'}]
+```
+
+## Training data
+
+[CLUECorpusSmall](https://github.com/CLUEbenchmark/CLUECorpus2020/) is used as training data. 
+
+## Training procedure
+
+The model is pre-trained by [UER-py](https://github.com/dbiir/UER-py/) on [Tencent Cloud](https://cloud.tencent.com/). We pre-train 1,000,000 steps with a sequence length of 512.
+
+
+```
+python3 preprocess.py --corpus_path corpora/cluecorpussmall.txt \
+                      --vocab_path models/google_zh_vocab.txt \
+                      --dataset_path cluecorpussmall_bart_seq512_dataset.pt \
+                      --processes_num 32 --seq_length 512 \
+                      --dynamic_masking --target bart 
+```
+
+```
+python3 pretrain.py --dataset_path cluecorpussmall_bart_seq512_dataset.pt \
+                    --vocab_path models/google_zh_vocab.txt \
+                    --config_path models/bart/base_config.json \
+                    --output_model_path models/cluecorpussmall_bart_base_seq512_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 16 \
+                    --span_masking --span_max_length 3 \
+                    --embedding word_pos --tgt_embedding word_pos \
+                    --encoder transformer --mask fully_visible --decoder transformer \
+                    --target bart --tie_weights --has_lmtarget_bias
+```
+
+Finally, we convert the pre-trained model into Huggingface's format:
+
+```
+python3 scripts/convert_bart_from_uer_to_huggingface.py --input_model_path cluecorpussmall_bart_base_seq512_model.bin-250000 \                                                                
+                                                        --output_model_path pytorch_model.bin \                                                                                            
+                                                        --layers_num 6
+```
+
+
+### BibTeX entry and citation info
+
+```
+@article{lewis2019bart,
+  title={Bart: Denoising sequence-to-sequence pre-training for natural language generation, translation, and comprehension},
+  author={Lewis, Mike and Liu, Yinhan and Goyal, Naman and Ghazvininejad, Marjan and Mohamed, Abdelrahman and Levy, Omer and Stoyanov, Ves and Zettlemoyer, Luke},
+  journal={arXiv preprint arXiv:1910.13461},
+  year={2019}
+}
+
+@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}
+}
+```