salesforce awd-lstm-lm issues

salesforce / awd-lstm-lm

LSTM and QRNN Language Model Toolkit for PyTorch

BSD 3-Clause "New" or "Revised" License

1.96k stars 488 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Effect of different seeds and other hyperparameters

#74 octavian-ganea opened 6 years ago
0
Switching criteria with non-monotone interval has difference in paper and code?

#73 yashu-seth opened 6 years ago
3
Variational WeightDrop not disabled at evaluation time

#72 hglaude opened 6 years ago
0
Typo for changing dropout parameter when resuming

#71 gyuwankim opened 6 years ago
0
KeyError: 'ax' in line `prm.data = optimizer.state[prm]['ax'].clone()`

#70 wasiahmad closed 6 years ago
20
Using AdaptiveLogsoftmaxWithLoss

#69 akurniawan opened 6 years ago
0
any pretrained models for Character level Penn Treebank (PTB) with LSTM??

#68 vinayakarannil opened 6 years ago
0
Error while trying to reproduce results for Pytorch 0.3

#67 mricepops opened 6 years ago
0
Fix --cuda argument

#66 wangzhe258369 closed 6 years ago
1
Dictionary - handling OOV tokens

#65 chiphuyen opened 6 years ago
1
DataParallel

#64 djstrong opened 6 years ago
3
Adaptive softmax fixes (#1)

#63 amirsaffari opened 6 years ago
1
splits cross entropy can be further optimized

#62 ReactiveCJ opened 6 years ago
2
Assign index id to word after sorting for adaptive softmax shortlist

#61 ReactiveCJ closed 2 years ago
2
How to use the trained model as a discriminative model?

#60 PromptExpert opened 6 years ago
2
Line 80 is useless in model.py

#59 jind11 closed 6 years ago
1
Why is the decoder using nhid as input size even when tie_weights is set at True ?

#58 FrancoisMentec opened 6 years ago
0
Weights sharing

#57 enod closed 6 years ago
1
Fix setting model dropoutE when resume and remove lr overwrite

#56 ollmer opened 6 years ago
1
Finetune issue

#55 giancds opened 6 years ago
3
In ASGD, what do we use for parameter, is it averaged one or normal SGD one?

#54 SongJeongHyun closed 6 years ago
1
In line 252, val_loss should be val_loss2 isn't it?

#53 SongJeongHyun closed 6 years ago
1
Few questions about main.py

#52 SongJeongHyun closed 6 years ago
1
Detail on WeightDrop class `_setup()` cuDNN RNN weight compacting issue & `register_parameter()`

#51 esvhd opened 6 years ago
7
How can i use Adam optimizer instead of SGD?

#50 SongJeongHyun closed 6 years ago
1
Adaptive softmax question

#49 YontiLevin closed 6 years ago
2
model.decoder is never used?

#48 mojesty opened 6 years ago
3
Triggering condition for ASGD bug

#47 anthonywchen closed 6 years ago
1
generate.py producing bad samples

#46 a-dai opened 6 years ago
8
Question about embedding dropout vs lockeddropout

#45 YontiLevin closed 6 years ago
1
Unable to load model from different directory

#44 aleSuglia opened 6 years ago
1
Modifications to work with pytorch 0.4

#43 shawntan closed 6 years ago
4
What does Finetune do?

#42 PetrochukM opened 6 years ago
3
Unpredictable behavior of adaptive softmax

#41 songyuzhou324 opened 6 years ago
3
Mention requirements and instructions for QRNN in readme/requirements.txt

#40 Iwontbecreative opened 6 years ago
1
Model crashes under pytorch 0.4

#39 zou3519 closed 6 years ago
5
Multiple GPU option

#38 songyuzhou324 closed 6 years ago
4
the script `getdata.sh` creates an empty `enwik8` folder and then finds a python script within the folder

#37 tariq-hasan-zz closed 6 years ago
1
Redundant code in getdata.sh

#36 tariq-hasan-zz closed 6 years ago
2
How-to generate after training word level qrnn?

#35 jhave opened 6 years ago
2
Question about using monotonic AvSGD

#34 Ruowei94 closed 6 years ago
1
Confused regarding motivation of randomized BPTT

#33 PetrochukM closed 6 years ago
7
Bug with val_loss2

#32 PetrochukM closed 6 years ago
2
[getdata.sh] rm duplicate

#31 julien-c closed 6 years ago
2
Issues with SplitCrossEntropyLoss

#30 ccarter-cs closed 6 years ago
8
GPU memory and cap

#29 cerisara closed 6 years ago
4
Fine-tune broken for QRNNs?

#28 daemon opened 6 years ago
2
Add --model arg for PTB/QRNN

#27 adrianbg closed 6 years ago
1
finetune & pointer bugs?

#26 aykutfirat opened 6 years ago
3
Add code and instructions for "An Analysis of Neural Language Modeling at Multiple Scales"

#25 Smerity closed 6 years ago
0

Previous Next