Problem : Error with Model() function

DeepSpeech
#1

I am having an issue with implementing the pre-trained model using jupyter (I am using this because I am more familiar with it’s interface). Currently on Tensorflow 2.0, Windows 7.

The error is as listed:

TypeError Traceback (most recent call last)
in
25
26 # Load the acoustic model
—> 27 ds = Model(MODEL_GRAPH, N_FEATURES, N_CONTEXT, LM_ALPHABET, BEAM_WIDTH)
28 print(‘Loaded model’)
29

~\Documents\Anaconda\lib\site-packages\deepspeech_init_.py in init(self, *args, **kwargs)
41 self._impl = None
42
—> 43 status, impl = deepspeech.impl.CreateModel(*args, **kwargs)
44 if status != 0:
45 raise RuntimeError(“CreateModel failed with error code {}”.format(status))

TypeError: CreateModel() takes at most 3 arguments (5 given)

The goal of this is to make a workable inference application with a GUI.

I am still new to all of this so if there is a better way of implementing deepspeech on raspberry pi please free feel to tell me.

#2

Just to rephrase, I wanted to have a model in a file format that i can interact with using Python(in another program). Preferably, in h5 file format.

#3

First, please stick to documented versions. Second, for pure inference, you don’t need TensorFlow.

What is not clear in this error ? You are giving too much parameters.

#4

Ok, I will try that first

Because every post I see about implementing deepspeech uses 5 arg.

def init(self, *args, **kwargs):
# make sure the attribute is there if CreateModel fails
self._impl = None

    status, impl = deepspeech.impl.CreateModel(*args, **kwargs)
    if status != 0:
        raise RuntimeError("CreateModel failed with error code {}".format(status))
    self._impl = impl

This is the code I saw on github. Would you be so kind in telling me what is *args & **kwargs? Because I do not know what param it need for it to work.

#5

Where ?

We have examples in native_client/python/client.py as well as under examples/ and API doc in https://deepspeech.readthedocs.io/en/latest/Python-API.html which is linked from the README.

Please, tell me what is unclear in all those.

#6

Omg thank you so much for this. I did not even saw this link anywhere on the Github page. (Must be my lack of research) Anyway, I will read through this and hopefully this link contains the answers I needed. Thank you so much for your help.

From here.
https://github.com/mozilla/DeepSpeech/blob/master/native_client/python/init.py

1 Like
#8

I don’t understand what to do for removing this error : TypeError: CreateModel() takes at most 2 arguments (5 given)

Thank u in advance :slightly_smiling_face:

#9

Read the documentation and pass the correct amount of parameters?

#10

https://deepspeech.readthedocs.io/en/latest/Python-API.html#native_client.python.Model

#11

Command : python recognize.py models/output_graph.pb ae_recorded.wav models/alphabet.txt models/lm.binary models/trie

I have read the documentation and parameter are all correct according to documentation

#12

Obviously, not.

This is your own code ? Do you expect me to help you without even knowing the code ?

You reply to someone else’s old thread, without any context at all. There is nothing we can do.

#13

no this is not my code
recognize.py code :

i am getting error:
Loading model from file models/output_graph.pb
Traceback (most recent call last):
File “recognize.py”, line 96, in
main()
File “recognize.py”, line 63, in main
ds = Model(args.model, N_FEATURES, N_CONTEXT, args.alphabet, BEAM_WIDTH)
File “/home/robust_audio_ae/env/lib/python3.6/site-packages/deepspeech/init.py”, line 40, in init
status, impl = deepspeech.impl.CreateModel(*args, **kwargs)
TypeError: CreateModel() takes at most 2 arguments (5 given)

#14

It is not our code either, so that still holds.

Ok:

    ds = Model(args.model, N_FEATURES, N_CONTEXT, args.alphabet, BEAM_WIDTH)

Have you read the documentation ? Obviously not.

This code is old. Either use the matching older version of DeepSpeech, or update its use and remove the un-needed parameters. The link I gave you already contains the proper information.

#15

Okk
I will go through the link
Thank u