stanleylsx/entity_extractor_by_ner
基于Tensorflow2.3开发的NER模型,都是CRF范式,包含Bilstm(IDCNN)-CRF、Bert-Bilstm(IDCNN)-CRF、Bert-CRF,可微调预训练模型,可对抗学习,用于命名实体识别,配置后可直接运行。
Python
Issues
- 0
- 0
- 3
无法生成lab2id和token2id两个文件
#72 opened by bigenheng - 1
start evaluate engines...之后报错
#73 opened by M-YoungBWeN - 3
如何训练自己的数据集
#68 opened by CLMPaul - 0
代码是否可以用在英文数据集上
#71 opened by niufendou - 0
equests.exceptions.SSLError:
#70 opened by bnjhzsdcgvd - 1
Connection error, and we cannot find the requested files in the cached path. Please try again or make sure your Internet connection is on.
#65 opened by 582856 - 0
TypeError: 'NoneType' object is not callable
#67 opened by 582856 - 1
ValueError: Connection error, and we cannot find the requested files in the cached path. Please try again or make sure your Internet connection is on.
#66 opened by 582856 - 1
Idcnn+crf 无法训练
#63 opened by leizhengtao520 - 2
将自己的数据集分为四类替换进去并且修改了suffix里面的标签为自己的标签 ,出现如下错误
#57 opened by 2506791459 - 5
- 2
- 1
将bert微调后的模型进行迁移,迁移后预测就结果就会错误,而在原来的机器上是正常的
#62 opened by guo-xiaomeng - 2
idcnn训练时出现准确率和召回率全0
#60 opened by ChemEng227 - 1
使用微调的时候,训练老是0,为什么
#56 opened by zhu1130 - 3
更换预训练模型,比如albert
#52 opened by hanxun0121 - 1
为什么每次predict的结果会不同呢
#54 opened by zhaoyu123213 - 1
- 1
tensorflow.python.framework.errors_impl.InvalidArgumentError: indices[24,95] = -7 is not in [0, 100)出现这个问题应该怎么办
#58 opened by 2506791459 - 3
文章出现\t 和空格时会报错
#30 opened by tyn513 - 2
ValueError: Tensor conversion requested dtype int32 for Tensor with dtype int64: <tf.Tensor 'targets:0' shape=(32, 300) dtype=int64>
#11 opened by dongkcs - 2
train_dataset, val_dataset = data_manager.get_training_set()两个返回TensorSliceDataset无法用len函数
#34 opened by formytest2 - 2
python版本
#50 opened by NanshineLoong - 1
替换为自己的数据后,出现错误
#46 opened by Wuhu0 - 7
导入自己标注的文本,显示数据集为0的问题
#43 opened by LakersUpAma - 2
- 2
用了自己的标注数据集 结果报错
#39 opened by N1ckyyy - 1
- 1
tensorflow版本与tensorflow-addons版本冲突
#42 opened by Wuhu0 - 1
- 5
训练为啥前几个 准确率啥的都是-1呢?是程序的设定吗?
#29 opened by tyn513 - 3
- 1
predict接口 cpu下推理速度比GPU下快
#32 opened by STHSF - 5
ValueError: Tensor's shape (300, 800) is not compatible with supplied shape (768, 800)
#35 opened by LiAI-tech - 1
label和token是基于train数据学出来的,test出现新的则无法使用
#31 opened by tyn513 - 1
内存消耗较大,出现内存交换
#28 opened by tyn513 - 2
你好,请问你改成tf.data之后,GPU的利用率以及训练效率有提升么
#18 opened by hanggun - 3
- 2
- 1
bert
#24 opened by chenjiawei22 - 1
Finetune-Bert-CRF
#25 opened by chenjiawei22 - 1
无法下载模型,怎么操作呢?
#26 opened by tyn513 - 1
如果不用bert,用的是什么做embd呢?
#27 opened by tyn513 - 0
自己准备好训练集之后出现这种情况,请问怎么解决
#21 opened by End1i - 1
关于数据集标注
#19 opened by iiixiscientia - 1
转换类型有问题
#16 opened by zhu13818202655 - 1
- 3
代码有点问题
#13 opened by CoderBinGe