jiesutd NCRFpp issues - Githubissues

jiesutd / NCRFpp

NCRF++, a Neural Sequence Labeling Toolkit. Easy use to any sequence labeling tasks (e.g. NER, POS, Segmentation). It includes character LSTM/CNN, word LSTM/CNN and softmax/CRF components.

Apache License 2.0

1.89k stars 446 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Standardizing reading input files with utf-8 encoding using io.open for both python 2 and 3.

#184 kaushikacharya opened 2 years ago
0
Support trained model serialization to ONNX format

#183 omar-araboghli opened 2 years ago
1
Unknown labels in decoding

#182 olzama opened 2 years ago
0
训练一段时间后，f1为-1

#181 Xzeffort closed 3 years ago
2
Resume model training from a checkpoint

#180 jd-coderepos opened 3 years ago
0
Decode error with nbest=0 when not using CRF

#179 jd-coderepos opened 3 years ago
0
About MAX_SENTENCE_LENGTH parameter

#178 jd-coderepos closed 3 years ago
2
Reading fasttext word embeddings

#177 jd-coderepos closed 3 years ago
2
Decode config parameter when not using CRF

#176 jd-coderepos closed 3 years ago
1
Experimenting with Transformer models?

#175 jd-coderepos opened 3 years ago
1
Test Data Format for Chunking

#174 Frost45 opened 3 years ago
0
Different results on CPU versus GPU

#173 jd-coderepos closed 3 years ago
1
Difference between main_parse.py and main.py

#172 jd-coderepos closed 3 years ago
1
About "tcmalloc: large alloc" message and training aborted

#171 jd-coderepos closed 3 years ago
3
大神你好，请教一些关于报错的问题

#170 ljx761795750 opened 3 years ago
1
Can I just use the CRF layer?

#169 JxuHenry closed 3 years ago
1
about requirement

#168 tonytonyjoba closed 3 years ago
1
Bug in forward method when calling _viterbi_decode, mask is not provided

#167 mlej8 opened 3 years ago
0
Provision for Custom Features?

#166 sush-infrrd closed 2 years ago
1
nbest score

#165 myeghaneh closed 3 years ago
1
Conll 使用内存大

#164 wizare closed 2 years ago
2
Model Architecture Question?

#163 RJain12 opened 3 years ago
0
NER paper

#162 lijiayi980130 closed 3 years ago
1
n best decode

#161 lijiayi980130 closed 3 years ago
1
metric on "span" level instead of on token level of iob

#160 myeghaneh closed 3 years ago
1
cutomization

#159 myeghaneh closed 3 years ago
1
ZeroDivisionError: float division by zero

#158 myeghaneh closed 4 years ago
2
I faced with this error, Maybe I should provide the word embedding when I want to train?

#157 myeghaneh closed 4 years ago
1
Edit: Solved

#156 DiegoGorini closed 4 years ago
0
new_tags in _score_sentence

#155 zhujiangang opened 4 years ago
0
关于复现 COLING 2018 paper NER task 任务 WLSTM+CRF+CCNN模型时遇到的一些问题

#154 hzs0828 closed 4 years ago
1
f score < 0.01 for conll2003 data

#153 nnakamura3 closed 4 years ago
2
UserWarning: masked_fill_ (...) please use a mask with dtype torch.bool instead.

#152 espenhk closed 4 years ago
3
About the oracle n-best score？

#151 xiao2mo closed 4 years ago
2
训练标点恢复时的问题（任务被kill，损失爆炸）

#150 nastyZhao closed 4 years ago
2
No pretrained embedding

#149 yeliu918 closed 4 years ago
3
-1 F1 score and 0 precision

#148 po-oya closed 4 years ago
1
WSJ数据集

#147 cswangjiawei closed 4 years ago
1
list index out of range

#146 Rajashan closed 4 years ago
4
Can I use NCRF for punctuation restoration?

#145 behitek closed 4 years ago
1
about bert

#144 ypc-stu closed 1 year ago
5
Find the viterbi score for each tag in output sequence

#143 trannhuthuat96 closed 5 years ago
2
run decode by cpu on model trained by GPU

#142 luoyangen closed 5 years ago
1
Why is test_dir included in the training config?

#141 taasmoe closed 5 years ago
1
about pos tag

#140 mingroad closed 5 years ago
3
Get actual size of data

#139 vasilikivmo closed 5 years ago
1
您好想问一下 word embedding 和char embedding 是都没有冻结吗

#138 xuanzebi closed 5 years ago
1
UserWarning: masked_fill_ received a mask with dtype torch.uint8

#137 soerenetler closed 5 years ago
2
+2 is used for the CRF later which needs start and end padding. The results in demo data doesn’t mean anything as the data is too small.

#136 shenxuhui closed 5 years ago
3
add two more label for downlayer lstm, use original label size for CRF

#135 shenxuhui closed 5 years ago
5