Python Async dumps problem Speech API

[Dellax8]

Hi,
with the transcribe_async.py file i can't get my results back.
The only changes that i made in transcribe_async.py are:

• Language
• Raw file passed by uri.

The error is in print(json.dumps(response['response']['results']))
Python says that the error is the Key 'results'

I try to fix it, but i can't understand what is wrong in this program.
I hope you can help me.

Thanks

[theacodes]

Hi @Dellax8 , can you try changing that line to print(json.dumps(response)) and sending us the output?

[puneith]

@Dellax8 Do you mean rest or grpc transcribe_async.py? Also can you let us know what language you changed it to?

[Dellax8]

@jonparrott
{"done": true, "metadata": {"lastUpdateTime": "2016-10-18T15:46:35.033374Z", "progressPercent": 100, "startTime": "2016-10-18T15:46:28.054832Z", "@type": "type.googleapis.com/google.cloud.speech.v1beta1.AsyncRecognizeMetadata"}, "name": "#NAME#", "response": {"@type": "type.googleapis.com/google.cloud.speech.v1beta1.AsyncRecognizeResponse"}}

@puneith
I Try with REST. Now i Try with grpc but i can't' understand what is wrong.
The language is it-IT but i think that this is not the problem.

[Dellax8]

with grpc transcribe_async.py i can't compile.

there are errors in "
from google.cloud.credentials import get_credentials
from google.cloud.speech.v1beta1 import cloud_speech_pb2
from google.longrunning import operations_grpc_pb2
from grpc.beta import implementations"

the errors are in the import part.
I have installed the requirements by pip, but it's not working yet

[puneith]

@Dellax8 Can you paste the error you are getting?

[Dellax8]

@puneith

in grpc - i fix the import error, but the result is empty...
in REST -
print(json.dumps(response['response']['results']))
KeyError: 'results'

[puneith]

Can you send me the audio file if you don't mind.

[Dellax8]

the encoding was wrong
I create a new file but it doesn't work

This is the audio Raw File

teone2.zip

[jerjou]

Hello - note that if you have encoded the file with different parameters than the sample one, you'll have to change the configuration parameters you send to the API. In this case, it looks like your audio file was encoded with a sample rate near 32000, which means you'll have to change the sampleRate parameter to match.

I was able to get at least a partial result from the API after changing the sampleRate and language parameters, though I wasn't able to verify since I don't speak italian :) Anyway, based on that I believe this should resolve this issue - please reopen this issue if it doesn't!