noahchalifour rnnt-speech-recognition issues

noahchalifour / rnnt-speech-recognition

End-to-end speech recognition using RNN Transducers in Tensorflow 2.0

MIT License

241 stars 78 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

The execution program propressing_common_voice.py reports the following error

#50 renzeyu-ai opened 2 years ago
0
EPOCH RESULTS: Loss: 0.0000

#49 weiwenying opened 3 years ago
1
tensorflow-datasets==4.2.0

#48 weiwenying opened 3 years ago
0
ImportError: tensorboard.plugins.hparams import api

#47 goreng2 closed 2 years ago
0
fix run_evaluate bug, delete gpus pram

#46 muramasa2 closed 3 years ago
0
Fix tensorflow-datasets version

#45 muramasa2 closed 3 years ago
3
Preprocessing issue when using token_type='character'

#44 ayushgarg70 opened 4 years ago
0
loss issue

#43 banayoyo closed 4 years ago
1
get Segmentation fault when training

#42 etyhh opened 4 years ago
8
Question !

#41 farmboy-dev closed 3 years ago
2
undefined symbol: _ZN10tensorflow12OpDefBuilder4AttrESs

#40 banayoyo closed 4 years ago
4
Error when setup the rnnt loss

#39 wwwwcu closed 4 years ago
1
error setting up rent loss in macOS

#38 Manish-rai21bit opened 4 years ago
1
unable to setup rnnt_loss using scripts/build_rnnt.sh

#37 manish-raivt closed 4 years ago
1
warp-transducer - Is rnnt_loss() "internal state" causing wrong loss computation?

#36 stefan-falk closed 3 years ago
1
Training is not converging. eval_wer sticks at ~95%.

#35 stefan-falk opened 4 years ago
7
Argument name error in preprocess_librispeech.py

#34 jtdutta1 closed 4 years ago
3
Training your own data set (audio and transcription)

#33 tumusudheer opened 4 years ago
1
Inference is giving ValueError: When input_signature is provided, all inputs to the Python function must be convertible to tensors:

#32 tumusudheer opened 4 years ago
5
Building Warp_transducer failed in google colab.

#31 omerasif-itu opened 4 years ago
9
Increasingly negative loss (becoming nan) and "forward backward likelihood mismatch"

#30 stefan-falk closed 4 years ago
1
Getting ValueError: Attempt to convert a value (PerReplica ..) when starting training

#29 stefan-falk opened 4 years ago
6
Segmentation fault when training

#28 dyc3 closed 4 years ago
2
ValueError: Python inputs incompatible with input_signature

#27 dyc3 closed 4 years ago
1
NaN loss

#26 dyc3 closed 4 years ago
4
Training a model with --fp16_run does not work

#25 dyc3 opened 4 years ago
1
Multi-thread Audio Conversion

#24 shaunm closed 4 years ago
4
Inference problem with small dataset

#23 NAM-hj closed 4 years ago
4
parallel GPUs working like serial

#22 ybNo1 closed 4 years ago
1
TF2.1 : AttributeError: module 'tensorflow.python.framework.ops' has no attribute 'RegisterShape'

#21 MoYinYu closed 4 years ago
2
TF2.0: no supported kernel for GPU devices is available

#20 MoYinYu closed 4 years ago
1
Question about gradient calculation.

#19 NAM-hj closed 4 years ago
1
Multi-GPU training is not working

#18 prajwaljpj opened 4 years ago
15
Is is normal with steps time bigger than 26 seconds in 4 Nvidia 1080Ti GPUs?

#17 TeaPoly closed 4 years ago
0
RAM OOM Problem

#16 kjh21212 opened 4 years ago
11
Question about impl

#15 digital10111 closed 4 years ago
1
unhealthy loss

#14 Jervint closed 4 years ago
4
How to calculate WER?

#13 Bowenwu1 closed 4 years ago
2
Get Error when using docker training by GPU

#12 Jervint closed 4 years ago
0
loss is not healthy

#11 ahmedalbahnasawy closed 4 years ago
3
multiple gpu error

#10 kjh21212 closed 4 years ago
2
An op outside of the function building code is being passed a "Graph" tensor. It is possible to have Graph tensors leak out of the function building context by including a tf.init_scope in your function building code.

#9 kjh21212 closed 4 years ago
2
LOSS IS NEGATIVE

#8 tulasiram58827 closed 4 years ago
3
Question about Streaming.

#7 tsxce closed 4 years ago
1
How to use the multi gpu?

#6 kjh21212 closed 4 years ago
3
Perf on device

#5 kezakool opened 4 years ago
2
Segmentation fault

#4 GaryGao99 closed 4 years ago
3
bug in evaluate.py

#3 nichongjia-2007 closed 4 years ago
2
tf-nightly version

#2 nichongjia-2007 closed 4 years ago
1
can you give a trained model for demo？

#1 l2009312042 closed 4 years ago
8