基于transformer的中文命名实体识别.zip

[English](#Requirement) [中文](#运行环境) # Flat-Lattice-Transformer code for ACL 2020 paper: FLAT: Chinese NER Using Flat-Lattice Transformer. Models and results can be found at our ACL 2020 paper [FLAT: Chinese NER Using Flat-Lattice Transformer](https://arxiv.org/pdf/2004.11795.pdf). # Requirement: ``` Python: 3.7.3 PyTorch: 1.2.0 FastNLP: 0.5.0 Numpy: 1.16.4 ``` you can go [here](https://fastnlp.readthedocs.io/zh/latest/) to know more about FastNLP. How to run the code? ==== 1. Download the character embeddings and word embeddings. Character and Bigram embeddings (gigaword_chn.all.a2b.{'uni' or 'bi'}.ite50.vec) : [Google Drive](https://drive.google.com/file/d/1_Zlf0OAZKVdydk7loUpkzD2KPEotUE8u/view?usp=sharing) or [Baidu Pan](https://pan.baidu.com/s/1pLO6T9D) Word(Lattice) embeddings: yj, (ctb.50d.vec) : [Google Drive](https://drive.google.com/file/d/1K_lG3FlXTgOOf8aQ4brR9g3R40qi1Chv/view?usp=sharing) or [Baidu Pan](https://pan.baidu.com/s/1pLO6T9D) ls, (sgns.merge.word.bz2) : [Baidu Pan](https://pan.baidu.com/s/1luy-GlTdqqvJ3j-A4FcIOw) 2. Modify the `paths.py` to add the pretrained embedding and the dataset 3. Run following commands ``` python preprocess.py (add '--clip_msra' if you need to train FLAT on MSRA NER dataset) cd V0 (without Bert) / V1 (with Bert) python flat_main.py --dataset <dataset_name> (ontonotes, msra, weibo or resume) ``` If you want to record experiment result, you can use fitlog: ``` pip install fitlog fitlog init V0 cd V0 fitlog log logs ``` then set use_fitlog = True in flat_main.py. you can go [here](https://fitlog.readthedocs.io/zh/latest/) to know more about Fitlog. Cite: ======== [bibtex](https://www.aclweb.org/anthology/2020.acl-main.611.bib) --- # 运行环境: ``` Python: 3.7.3 PyTorch: 1.2.0 FastNLP: 0.5.0 Numpy: 1.16.4 ``` 你可以在 [这里](https://fastnlp.readthedocs.io/zh/latest/) 深入了解 FastNLP 这个库. 如何运行？ ==== 1. 请下载预训练的embedding 从[Google Drive](https://drive.google.com/file/d/1_Zlf0OAZKVdydk7loUpkzD2KPEotUE8u/view?usp=sharing) 或 [Baidu Pan](https://pan.baidu.com/s/1pLO6T9D) 下载字和 Bigram 的 embedding (gigaword_chn.all.a2b.{'uni' or 'bi'}.ite50.vec) 从[Google Drive](https://drive.google.com/file/d/1K_lG3FlXTgOOf8aQ4brR9g3R40qi1Chv/view?usp=sharing) 或 [Baidu Pan](https://pan.baidu.com/s/1pLO6T9D) 下载词的 embedding (ctb.50d.vec)(yj) 从[Baidu Pan](https://pan.baidu.com/s/1luy-GlTdqqvJ3j-A4FcIOw) 下载词的embedding (sgns.merge.bigram.bz2)(ls) 2. 修改 `paths.py` 来添加预训练的 embedding 和你的数据集 3. 运行下面的代码 ``` python preprocess.py (add '--clip_msra' if you need to train FLAT on MSRA NER dataset) cd V0 (without Bert) / V1 (with Bert) python flat_main.py --dataset <dataset_name> (ontonotes, msra, weibo or resume) ``` 如果你想方便地记录和观察实验结果, 你可以使用fitlog: ``` pip install fitlog fitlog init V0 cd V0 fitlog log logs ``` 然后把flat_main.py里的 use_fitlog 设置为 True 就行 你可以在 [这里](https://fitlog.readthedocs.io/zh/latest/) 深入了解 Fitlog 这个工具 引用: ======== [bibtex](https://www.aclweb.org/anthology/2020.acl-main.611.bib) 更新说明： ======== 5.7共提交两个版本，其中V2使用tensor.unique()用于去除相对位置中重复组合（记为Flat_unique），V3使用标量替代了FLAt中的相对位置编码(记为Flat_scalar).详见[FLAT瘦身日记](https://zhuanlan.zhihu.com/p/509248057) 使用这两种方法的显存占用如下表所示 batch_size=10 |seq_len| 50 | 100 | 150 | 200 | 250 | 300 | |:-------|----:|-----:|-----:|-----:|-----:|-----:| |Flat|1096MB | 1668MB |2734MB|4118MB|5938MB|8374MB| |Flat_unique|964MB|1204MB|1610MB|2166MB|2922MB|3940MB| |Flat_scalar|878MB|916MB|1028MB|1062MB|1148MB|1322MB| |Bert+Flat|1605MB|2237MB|3333MB|4725MB|6571MB|9039MB| |Bert+Flat_unique|1495MB|1685MB|2129MB|2697MB|3453MB|4585MB| |Bert+Flat_scalar|1409MB|1481MB|1565MB|1617MB|1705MB|2051MB|