Arabic model stuck

Archives DeepSpeech
#22

can you please share that as plain text here, instead of a google doc link?

#23

sure

Epoch training loss Validation loss
0 70.788821 54.815269
1 43.40599 45.615384
2 36.944965 40.852923
3 33.187017 37.502073
4 30.515962 34.640956
5 28.921748 32.462484
6 27.693192 30.43777
7 26.646094 26.959204
8 25.71389 24.63013
9 25.01252 23.623141
10 24.456176 22.395472
11 23.689649 21.885186
12 23.181059 21.200026
13 22.769138 20.290973
14 22.604338 19.716581
15 22.366818 19.167822
16 22.205989 21.117617
17 21.955193 18.876346
18 21.86773 18.313352
1 Like
#24

I’m wondering if your training might not have been stopped a bit too early, looking at the figures.

#25

i removed the no early stop this time

#26

The training finished but result is worse than the training with early stop
Any suggestions is highly appreciated

#27

Unfortunately, as I stated earlier, it’s likely expected given the data you have.

#28

I also would say the above training did stop too early.

In my trainings I always used early stopping, with es_epochs=7 and es_min_delta=0.1, as well as plateau reduction with epochs=3 I think.

Set training epochs to your estimated maximal epoch number, something like 50?, but not to high, because of the augmentation ranges.

1 Like
#29

What is your training corpus size in terms of Hours
mine is about 1200 hours i guess

#30

I did use about 1500h for german and 700h for Spanish.
So 1200h should be something you can work with, especially if you can later reduce the domain size with a specialized language model.
I would estimate you can reach something between 15-20% WER, depending how complex understanding arabic is.

-------- Original-Nachricht --------

#31

Thank you daneial
Can you share your training script variables
No of epoch
Learned no rate etc
Thank you

#32

You can find everything here:

The exact variables are included in the published checkpoint files (links are at the bottom of the readme)

1 Like
#33

No if I want to go to higher number of epochs say 30
Can I continue training from where it is stopped
Or I should start fresh training with him number of epoch
Some one told me that Adam optimizer changes as the number of epoch increase

#34

You can continue it, because optimizer state should be saved in the checkpoints.

-------- Original-Nachricht --------

#35

thank you
i will set max numnber of epoch to 30 and see what happened

#36

Hello,
Can you share the changes you made to DS to work with Arabic? I am working in Urdu.
Thanks.

#37

nothing, it works directly just generate new alphabet.txt file for Arabic and that is it
in your case you can use transcript in Urdu and regenerate alphabet.txt for Urdu and point to its location

#38

Even when I put number of epoch =30
Abd remove no early stop
The problem still exist on deepspeech android tflite
To refresh your memory the problem is this
We implement this code


On English model it works fine
When we port our own model it recognized the very first utterance and stop recognizing any thing
When I go to debug on machine I notice that the decoder use the first audio for recognition and does not use latest audio pronounced
#39

Hello, alex. I want to know It is possible provide the Arabic Deepspeech model you are using ?
It is public model or did you trained it yourself?
If its public then where I can download it ? If its not public I want to know it is possible you can provide a alphabet.txt of Arabic.

I am Chinese but my uncle was an Islam so he want to extract the Arabic text from video which speaking Arabic. I know how to train but I am stucked at providing the Arabic.txt file to Deepspeech.
Thanks.