vrenkens nabu issues - Githubissues

vrenkens / nabu

Code for end-to-end ASR with neural networks, build with TensorFlow

MIT License

108 stars 43 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Performance issue in the definition of marigin_loss, nabu/neuralnetworks/trainers/loss_functions.py(P1)

#60 DLPerf opened 3 years ago
0
Multi-task learning

#59 KunalDhawan opened 5 years ago
0
Test loss = 1 / decoder giving back labels

#58 jtcampbellsoup opened 5 years ago
3
Visualize gradients during training

#57 AzizCode92 closed 5 years ago
0
Would you please give an example about the data formet?

#56 helloWaterM opened 5 years ago
0
tensorflow.python.framework.errors_impl.InvalidArgumentError: assertion failed: [Jeff thought you argued in favor of a centrifuge purchase. <eos>]

#55 nipengmath opened 5 years ago
9
Training a language model for rescoring outputs

#54 k-priyadarshi opened 5 years ago
2
Hi Vincent! Is there any way I can train a language model on my dataset using Nabu for rescoring the outputs? Please reply.

#53 k-priyadarshi closed 5 years ago
1
Adding recurrent batch normalization

#52 AzizCode92 closed 5 years ago
3
CTC decoder results saving

#51 rzcwade closed 5 years ago
4
data overlapped in itself in tensorboard (validation loss)

#50 AzizCode92 closed 5 years ago
3
Interpreting test loss as PER?

#49 k-priyadarshi closed 5 years ago
2
windowing attention

#48 riyijiye closed 5 years ago
2
DBLSTM/librispeech recipe

#47 AzizCode92 closed 5 years ago
1
Update trainer.py

#46 AzizCode92 closed 5 years ago
0
Clarification required for some methods

#45 k-priyadarshi closed 6 years ago
2
Enabling the layer normalization decrease the performance of the model

#44 AzizCode92 opened 6 years ago
3
sync optimizer problem

#43 fanlu opened 6 years ago
15
Beam_search_decoder error of "list index out of range"

#42 AzizCode92 closed 6 years ago
7
fixing typo

#41 AzizCode92 closed 6 years ago
1
Name of folder standart should be renamed to standard

#40 AzizCode92 closed 6 years ago
1
Single Machine Mode with Multi-GPU

#39 AzizCode92 closed 6 years ago
6
_cut_sequence issue

#38 rzcwade closed 6 years ago
4
pBLSTM not implemented for LAS/WSJ recipe?

#37 rzcwade closed 6 years ago
2
encoded and encoded_seq_length in attention.py

#36 rzcwade closed 6 years ago
2
phoneme vs. grapheme based?

#35 mjhanphd closed 6 years ago
2
Training loss over 400?

#34 rzcwade closed 6 years ago
2
LAS/WSJ

#33 rzcwade closed 6 years ago
4
issue with training (LAS TIMIT)

#32 rzcwade closed 6 years ago
12
Distribute training problem

#31 fanlu closed 6 years ago
7
which recipe to suggest for librispeech

#30 riyijiye closed 6 years ago
5
handling <sos> and <eos>

#29 riyijiye closed 6 years ago
5
your local BeamSearchDecoder implementation

#28 riyijiye closed 6 years ago
3
speller decoder construction

#27 riyijiye closed 6 years ago
2
the model converge so slow

#26 betterdam closed 6 years ago
2
pblstm implementation

#25 riyijiye closed 6 years ago
2
Why do ‘<sos>’ and '<eos>' have the same label?

#24 ZhengkunTian closed 6 years ago
2
Distributed computing

#23 riyijiye closed 6 years ago
2
Joint CTC-Attention

#22 ghost closed 6 years ago
10
question about speller

#21 riyijiye closed 6 years ago
2
timit las loss large

#20 wbgxx333 closed 6 years ago
1
kaldi use mfcc feature,but fbank in config/recipes/DNN/WSJ/feature_processor.cfg

#19 fanlu closed 6 years ago
20
error while read data when I use recipe DBLSTM

#18 fanlu closed 6 years ago
2
module can't find

#17 fanlu closed 6 years ago
7
What is the relationship between this project and tfkaldi?

#16 fanlu closed 6 years ago
2
Character Error rate (CER) or Phoneme Error Rate (PER) ?

#15 xavigiro closed 6 years ago
2
fisher information

#14 jannaescur closed 6 years ago
3
results tensorboard

#13 jannaescur closed 6 years ago
3
evaluate results

#12 jannaescur closed 6 years ago
14
LAS paper

#11 jannaescur closed 6 years ago
1