11 questions
0
votes
1
answer
182
views
Whisper model Real-time endpoint container deployment failed on Azure ML
I tried to deploy Whisper on Azure ML.
I am using the Whipser-openAI-v3 model for deployment.
The endpoint creation successes but deployment failed with the error ResouceOperationFailed
and so the ...
0
votes
0
answers
179
views
Whisper Inference
why transcribe stage we remove N_FRAMES from mel and in for loop over the mel_segment it didn't take the last segment if it's less than 3000 frame why? let's suppose that he mel = [80,4100] first mel ...
0
votes
1
answer
149
views
Mozilla deepspeech from deepspeech import Model not working
Im trying to use mozilla deepspeech to transcribe text however im running into issues importing the Model module.
here is my code
from deepspeech.model import model
model_file_path='deepspeech-0.9.3-...
1
vote
0
answers
13
views
Sphinxtrain: Unable to lookup word that exists in the dictionary
I'm adapting a sphinx model for Brazilian portuguese with my own data by following their tutorial and got stuck on the bw command in the "Accumulating observation counts" section. I made ...
- 218
0
votes
1
answer
526
views
react-speech-recognition package not working
It's a simple react package that convert user audio to text. I install the package and try its basic code example but it shows a error "RecognitionManager.js:247 Uncaught ReferenceError: ...
0
votes
1
answer
1k
views
Why is Word Information Lost (WIL) calculated the way it is?
Word Information Lost (WIL) is a measure of the performance of an automated speech recognition (ASR) service (e.g. AWS Transcribe, Google Speech-to-Text, etc.) against a gold standard (usually human-...
- 362
4
votes
0
answers
841
views
Speaker Diarization is disabled even for supported languages in Google Speech-to-Text API V2
I'm trying to use Google's Speech-to-Text v2 API for transcription and speaker diarization.
Per this supported languages page, I should be able to create a Recognizer using the "long" model ...
- 362
0
votes
1
answer
33
views
How does placing the output (word) labels on the initial transitions of the words in an FST lead to effective composition?
I am going through hbka.pdf (WFST paper). https://cs.nyu.edu/~mohri/pub/hbka.pdf
A WFST figure for reference
Here the input label i, the output label o, and weight w of a transition are marked on the ...
1
vote
0
answers
693
views
Fine tunned Whisper-medium always predict "" for all samples
i'm trying to fine tunning whisper-medium for Koreans language.
Here is tutorial that i followed.
And here is my experiment setting
python==3.9.16
transformers==4.27.4
tokenizers==0.13.3
torch==2.0.0
...
- 11
3
votes
3
answers
4k
views
How to get all hugging face models list using python?
Is there any way to get list of models available on Hugging Face? E.g. for Automatic Speech Recognition (ASR).
5
votes
1
answer
8k
views
How to segment and transcribe an audio from a video into timestamped segments?
I want to segment a video transcript into chapters based on the content of each line of speech. The transcript would be used to generate a series of start and end timestamps for each chapter. This is ...
- 741