I created a Google Cloud Speech v2 recognizer via the front end of the web google cloud console. The Recognizer service is located in europe-west4, because I need the chirp model. The problem comes when i try to get the recognizer via the Node.js client library.
client.getRecognizer({ name: `projects/${project}/locations/europe-west4/recognizers/my-first-chirp` })
The error is:
Google speech v2: 'Expected resource location to be global, but found europe-west4 in resource name.',
When i try to create a different recognizer in europe-west4 i get the same error, but when creating in "global" location there is no problem. The issue is that the model and language I need are not supported in "global".
When I tried to find my recognizer in "global" with .getRecognizer
I got an error sayign there isn't such recognizer (as expected, because it is not in global).
I there a way to use google's node library with europe-west4?
When creating the client you need to set the region in the URL:
For Javascript/Typescript:
this.client = new v2.SpeechClient({
apiEndpoint: `europe-west4-speech.googleapis.com`,
});
For Python
client_options_var = client_options.ClientOptions(
api_endpoint="europe-west4-speech.googleapis.com"
)
client = SpeechClient(client_options=client_options_var)
Beware the "-speech" part region + "-speech.googleapis.com"