issues
search
hipster-philology
/
pandora
A Tagger-Lemmatizer for Natural Languages
MIT License
9
stars
4
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Performance Drop between Keras and PyTorch on Medieval French
#56
PonteIneptique
closed
7 years ago
9
Epoch's evaluation takes an unusual long time
#55
PonteIneptique
opened
7 years ago
22
Test scores are printed twice (once with Logger, once alone)
#54
PonteIneptique
closed
7 years ago
12
Tests are currently not being passed
#53
emanjavacas
closed
7 years ago
0
main.py, unseen.py, train.py
#52
emanjavacas
closed
7 years ago
1
predict() got an unexpected keyword argument 'batch_size'
#51
Jean-Baptiste-Camps
closed
7 years ago
8
Auto-compute max_lemma_len
#50
PonteIneptique
closed
7 years ago
3
Fix requirements on numpy
#49
PonteIneptique
closed
7 years ago
0
Cannot run PyTorch
#48
PonteIneptique
closed
7 years ago
3
How to handle gpu
#47
emanjavacas
closed
7 years ago
5
Add back Logger to Tagger
#46
PonteIneptique
closed
7 years ago
1
Pytorch
#45
emanjavacas
closed
7 years ago
22
Monitoring of scores
#44
emanjavacas
closed
7 years ago
6
Document model output
#43
emanjavacas
opened
7 years ago
3
Different embedding matrices for token and lemma characters
#42
emanjavacas
opened
7 years ago
2
Padding, end and beginning of sequence symbols
#41
emanjavacas
closed
7 years ago
5
RIP Theano
#40
mikekestemont
closed
7 years ago
2
Logo
#39
mikekestemont
opened
7 years ago
14
Merging the pytorch branch
#38
emanjavacas
closed
7 years ago
7
Add a link to articles in the README.md
#37
PonteIneptique
closed
7 years ago
0
Updating to Keras 2 API ?
#36
Jean-Baptiste-Camps
closed
7 years ago
1
plot related corrections
#35
Jean-Baptiste-Camps
closed
7 years ago
0
Figures not being closed
#34
Jean-Baptiste-Camps
closed
7 years ago
1
solving mandatory dev bug in CLI
#33
Jean-Baptiste-Camps
closed
7 years ago
0
Specifying --dev files mandatory in CLI
#32
Jean-Baptiste-Camps
closed
7 years ago
0
Failed to open TrueType font
#31
Jean-Baptiste-Camps
closed
7 years ago
4
Architecture improvement
#30
PonteIneptique
closed
7 years ago
3
Issue #25 : Adding tests to ensure loading and saving params works
#29
PonteIneptique
closed
7 years ago
6
Printing or exporting score dict should be done outside of evaluation and test
#28
PonteIneptique
closed
7 years ago
4
Changing configuration saving, to allow loading at a given epoch
#27
Jean-Baptiste-Camps
closed
7 years ago
0
Testing every nth epochs to save time
#26
PonteIneptique
closed
7 years ago
5
Config loading should make sure to import right types
#25
PonteIneptique
closed
7 years ago
3
Made an opt-in for evaluation printing during training.
#24
PonteIneptique
closed
7 years ago
1
Config parameters not being saved from actual parameters
#23
Jean-Baptiste-Camps
closed
7 years ago
8
Clean up of the CLI. Fixed #21
#22
PonteIneptique
closed
7 years ago
1
Clean-up CLI system
#21
PonteIneptique
closed
7 years ago
0
Opt-in epoch step results evaluation with CLI
#20
PonteIneptique
closed
7 years ago
9
Issue 18
#19
PonteIneptique
closed
7 years ago
0
Unseen.py on directories lacks new line
#18
PonteIneptique
closed
7 years ago
2
#12 : Fixed nb_classes unknown argument
#17
PonteIneptique
closed
7 years ago
0
Difference in length between predictions and input tokens
#16
Jean-Baptiste-Camps
closed
7 years ago
5
lemma_char_idx
#15
emanjavacas
opened
7 years ago
18
One-hot format for output data
#14
emanjavacas
closed
7 years ago
2
Change output type of keras model
#13
emanjavacas
closed
7 years ago
1
TypeError: to_categorical() got an unexpected keyword argument 'nb_classes'
#12
Jean-Baptiste-Camps
closed
7 years ago
4
Abstract over model (pytorch vs keras)
#11
mikekestemont
closed
7 years ago
7
Learning rate
#10
Jean-Baptiste-Camps
opened
7 years ago
5
Changes in training and prediction commands
#9
Jean-Baptiste-Camps
closed
7 years ago
2
Main.py and tagger.train()
#8
Jean-Baptiste-Camps
closed
7 years ago
1
Command-line interface for training
#7
Jean-Baptiste-Camps
closed
7 years ago
1
Previous
Next