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 487 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

initial input dropout mask is not considered.

#124 shuncask opened 4 years ago
0
The might be a bug in splitcross.py

#123 cswhjiang opened 4 years ago
0
The model behaves normally during training, but during prediction, the weightdrop mechanism cannot be stopped and an error is reported

#122 zhourui-xihu closed 4 years ago
0
val_loss suddenly increases to a larger number(from ~5 to more than 10000)

#121 xuuuluuu closed 4 years ago
1
A UsingWarning about call flatten_parameters()

#120 wxhiff opened 4 years ago
0
AttributeError: 'LSTM' object has no attribute 'weight_hh_l0'

#119 owaski closed 4 years ago
1
Fixing cuda flag

#118 anandtrex closed 4 years ago
1
torch.cuda.FloatTensor

#117 lmxhappy closed 4 years ago
0
Hidden state init of LSTM layers (after the first)

#116 mourga closed 4 years ago
3
Fixing typos and useless commands in main

#115 RikoNyberg opened 4 years ago
1
TypeError: cannot assign 'torch.cuda.FloatTensor' as parameter 'weight_hh_l0' (torch.nn.Parameter or None expected)

#114 S-Abdelnabi closed 4 years ago
1
Dropconnect layer implementation.

#113 swaroop1904 opened 4 years ago
0
why is hidden layer initialized only on beginning of an epoch?

#112 realiti4 closed 4 years ago
2
AssertionError

#111 HCideal opened 4 years ago
1
AttributeError: 'Program' object has no attribute '_program'

#110 acriptis opened 4 years ago
1
Score a test sentence

#109 ykl7 closed 4 years ago
0
Increasingly occupied GPU memory

#108 LeePleased opened 5 years ago
6
Language model infer

#107 flassTer opened 5 years ago
0
Problems with the next word predictions

#106 ishpiki opened 5 years ago
0
possible error on the neural cache model, pointer? Why use the groundtruth data?

#105 Sunnydreamrain closed 5 years ago
1
Make pytorch1.0 happy

#104 Stonesjtu opened 5 years ago
1
Hyper parameter settings for PyTorch 1.0?

#103 manuvn opened 5 years ago
2
save model

#102 amirakazadeh opened 5 years ago
0
how can I train this model with my own dataset??

#101 amirakazadeh opened 5 years ago
0
The output in finetune.py is not processed by model.decoder

#100 by2101 opened 5 years ago
0
Loading a pre-trained model with Weight Drop

#99 dhananjaisharma10 closed 5 years ago
1
High Test Perplexity

#98 Sketchizer closed 5 years ago
0
Why there are two times log_softmax in splitcross.py

#97 marcwww opened 5 years ago
2
how to output sentence's probability?

#96 OswaldoBornemann opened 5 years ago
12
Input batch size doesn't match hidden batch size

#95 homelifes closed 5 years ago
2
ONNX LSTM-LM Export Fails

#94 abrarmatin opened 5 years ago
0
RuntimeError: expects a 2D tensor,but self is 1D

#93 gai1995 closed 5 years ago
0
cPickle.PicklingError: Can't pickle <type 'instancemethod'>: attribute lookup __builtin__.instancemethod failed

#92 ghost opened 5 years ago
1
Is repackage_hidden(hidden) required?

#91 felixfuu closed 5 years ago
1
Training loop cleaning and commenting for better understanding and further improvements

#90 ngarneau opened 5 years ago
1
KeyError: 'ax' in line `prm.data = optimizer.state[prm]['ax'].clone()

#89 shirishr closed 5 years ago
2
Support for pytorch 1.0 scheduled?

#88 shirishr opened 5 years ago
1
On requiring `pytorch-qrnn`

#87 guillaume-chevalier opened 5 years ago
0
RuntimeError: shape '[5290000, 1]' is invalid for input of size 4600

#86 beneyal opened 5 years ago
15
AttributeError: 'torch.FloatTensor' object has no attribute 'detach'

#85 lihy-1224 opened 5 years ago
1
Fix cuda flag.

#84 egrefen opened 5 years ago
1
does bptt length affect test perplexity?

#83 tsauri opened 5 years ago
0
loss computation in splitcross.py is false

#82 adjidieng opened 5 years ago
0
Uses "--no-cuda" and implements device-agnostic code

#81 nkcr opened 5 years ago
0
THCudaCheck FAIL file=/tmp/luarocks_cutorch-scm-1-3884/cutorch/lib/THC/THCGeneral.c line=70 error=35 : CUDA driver version is insufficient for CUDA runtime version

#80 AnuragRoy82 opened 5 years ago
1
cannot assign 'torch.cuda.FloatTensor' as parameter 'weight'

#79 shirishr opened 5 years ago
6
Evaluation without training

#78 ChristineBasta opened 5 years ago
0
Update parameter with ASGD

#77 akurniawan opened 5 years ago
0
Add Hebbian Softmax training

#76 geniki opened 5 years ago
1
Including pretrained embeddings

#75 ChristineBasta opened 6 years ago
0