Issue when im running my own french model solved

[JRMeyer]

>>> sebastienlayer
[April 25, 2019, 12:50pm]

hello,

I just try to train my own model based on the french common voice
corpus. No problem on this side for the moment

I trained the moment on a machine with GPU. As I have a problem of CUDA
dependencies when i try to run my model, i build a second VM just for
execution the model without contraint

On this second machine, i think all is ok because when i try to execute
the english pre-trained model, i have no problem. When i do the same
thing with my own model, i have a basic error. I do really not
understand where it came from.

is there somebody who can help me investigating this issue. Thx

One more thing. I train my model with tensorflow 1.13.1 and
deepspeech 0.4.1 as explain on deepspeech github. however the deepseech
runtime i install via pip is based on TensorFlow
v1.12.0-10-ge232881, DeepSpeech: v0.4.1-0-g0e40db6.

Do you think it can be a problem. Perhaps i need to build my own native
client based on tensorflow 1.13 ?

Thanks in advance for your advice.

(deepspeech-venv) slash [layer slash deepspeech slash ~ slash ] slash $ deepspeech slash --model
modelsfr/speech2text_fr_v1.pbmm slash --alphabet modelsfr/alphabet.txt
slash --audio modelsfr/test_audio.wav slash
Loading model from file modelsfr/speech2text_fr_v1.pbmm slash
TensorFlow: v1.12.0-10-ge232881 slash
DeepSpeech: v0.4.1-0-g0e40db6 slash
2019-04-25 11:13:39.159754: I
tensorflow/core/platform/cpu_feature_guard.cc:141 slash ] Your CPU supports
instructions that this TensorFlow binary was not compiled to use: AVX2
FMA slash
Not found: Op type not registered 'AudioSpectrogram' in binary running
on deepspeech. Make sure the Op and Kernel are registered in the binary
running in this process. Note that if you are loading a saved graph
which used ops from tf.contrib, accessing (e.g.) tf.contrib.resampler
should be done before importing the graph, as contrib ops are lazily
registered when the module is first accessed. slash
Traceback (most recent call last): slash
File '/home/layer/deepspeech-venv/bin/deepspeech', line 11, in slash
sys.exit(main()) slash
File
'/home/layer/deepspeech-venv/lib64/python3.6/site-packages/deepspeech/client.py',
line 80, in main slash
ds = Model(args.model, N_FEATURES, N_CONTEXT, args.alphabet,
BEAM_WIDTH) slash
File
'/home/layer/deepspeech-venv/lib64/python3.6/site-packages/deepspeech/init.py',
line 14, in init slash
raise RuntimeError('CreateModel failed with error code
{}'.format(status)) slash
RuntimeError: CreateModel failed with error code 5

[This is an archived TTS discussion thread from discourse.mozilla.org/t/issue-when-im-running-my-own-french-model-solved]

[JRMeyer]

>>> sebastienlayer
[April 25, 2019, 12:50pm]

I solve my problem. I download the native 0.5 alpha 6 client. And it
works. But there is some warning. As the inference output is empty, i
don't know if it's due to model training or alpha version of deepspeech.
So i decide to re-train my model with 0.4.1 version.

[Archived Post]

[JRMeyer]

>>> lissyx
[April 25, 2019, 1:25pm]

Could you join efforts ?
https://github.com/Common-Voice/commonvoice-fr

[Archived Post]

[JRMeyer]

>>> sebastienlayer
[April 25, 2019, 1:27pm]

possible, what kind of help to you need ?

[Archived Post]

[JRMeyer]

>>> lissyx
[April 25, 2019, 1:31pm]

All help is welcome, I'll read your message after

[Archived Post]

[JRMeyer]

>>> lissyx
[April 25, 2019, 1:35pm]



> As the inference output is empty, i don't know if it's due to model
> training or alpha version of deepspeech.

It's likely because of training, not enough epochs. Can you share your
training parameters?

[Archived Post]

[JRMeyer]

>>> lissyx
[April 25, 2019, 1:39pm]



> One more thing. I train my model with tensorflow 1.13.1 and
> deepspeech 0.4.1 as explain on deepspeech github. however the
> deepseech runtime i install via pip is based on TensorFlow
> v1.12.0-10 -ge232881, DeepSpeech: v0.4.1-0-g0e40db6.
>
> Do you think it can be a problem. Perhaps i need to build my own
> native client based on tensorflow 1.13 ?

if you train with current master your model will not be compatible with
the binaries of version 0.4.1

[Archived Post]

[JRMeyer]

>>> sebastienlayer
[April 25, 2019, 1:48pm]

As i'm a beginner, i try with something simple. If you can help me to
determinate what can be a first set of parameters to start, it will be
cool
![:slight_smile:](

python -u DeepSpeech.py slash slash slash

[Archived Post]

[JRMeyer]

>>> lissyx
[April 25, 2019, 1:50pm]

Maybe you can try to get inspired from
https://github.com/lissyx/DeepSpeech/blob/master/bin/train_fr.sh

it'd be great we move on
common-voice/commonvoice-fr#24 and
associated common voice import / cleanup tooling to avoid re-inventing
the wheel everytime.

If you want to work on this docker pipeline feel free

[Archived Post]

[JRMeyer]

>>> lissyx
[April 25, 2019, 1:55pm]



> python -u DeepSpeech.py slash
> --train_files /mnt/disks/ssd/common-voice-fr/clips/train.csv slash
> --dev_files /mnt/disks/ssd/common-voice-fr/clips/dev.csv slash
> --test_files /mnt/disks/ssd/common-voice-fr/clips/test.csv slash
> --alphabet_config_path /mnt/disks/ssd/common-voice-fr/alphabet.txt slash
> --epochs 7 slash
> --checkpoint_dir /mnt/disks/ssd/checkpoints/ slash
> --export_dir /mnt/disks/ssd/model_export/

7 epochs is likely not enough slash
How did you generate your alphabet ? slash
You would also be using english language model since you don't specify
one

All that feels consistent with 'empty' output.

[Archived Post]

[JRMeyer]

>>> sebastienlayer
[April 25, 2019, 2:25pm]

I built the alphabet using the command check_characters.py

Actually, i'm looking how to build a french language model. But I can
use the English one if it does not interfere with learning

I will try the parameters you propose here
https://github.com/lissyx/DeepSpeech/blob/master/bin/train_fr.sh

Thanks

[Archived Post]