I uploaded a mp3 file (1.07 min) to transcribe using the video model, and the transcription is stuck, no errors, just infinite loading
This is the file https://drive.google.com/file/d/1QLucfAwJZXxSAOKSIWoPtIbljSM9pvsl/view?usp=sharing
UPDATE: It failed with “Error running recognize request. Too many retries, giving up.”
May I know the documentation you are using in transcribing your audio?
In the case above I’m sending the file to transcription using the UI in google cloud console
But I did it after noticing the issue by sending the same file using the Node.js api, following the docs here: https://cloud.google.com/speech-to-text/docs
I was able to fetch the transcription in JSON format using gcloud CLI. I used Asynchronous Speech Recognition for transcribing an audio file that is longer than a minute. But I converted your audio file from Mp3 to FLAC since the process can’t be completed(based on my replication) when I used Mp3.
These are the commands that I used:
gcloud ml speech recognize-long-running ‘gs://bucket-name/audio.flac’ --language-code=‘en-US’ --async --audio-channel-count=2 --separate-channel-recognition> > gcloud ml speech operations describe [name]
This is a snippet of my output:
[upload|U3SRL4Nam0rC33oE49ZTuw==]
So the issue is the file is an mp3 and not flac? should I also re-encode the file?
if it just a container change, then it is weird that mp3 fails and flac doesn’t because it is the same encoded data inside
also, I can’t access the output link you shared, it requires a google employee account
@Poala_Tenorio it happens again, I sent a transcription long process and didn’t get any results (should be in a bucket)
this is the process id I received from the api: 7346104001135815711
is there a way I can check the status of this id?
I manged to get the latest status of the process 7346104001135815711
received: code 13, Too many retries, giving up.
any way to see why is this happening?