are you using Amazon SageMaker to deploy your model or Inference Endpoint?
Inference.
But i can see your point, if we just set it up with sage maker we can modify anything we want correct?
we want to set it up like this and run it against many commercial tools for STT we use now:
To be like this:
curl https://xnmxj6m0sqyl3z55.us-east-1.aws.endpoints.huggingface.cloud
-X POST
–data {
“encoding”: “wav”,
“language”: “Japanese”,
“transcribe”: true,
“languageDetect”: false,
“model”: “large”,
“audio”: “signed-url-for-google-or-aws-buckets”,
}
-H “Authorization: Bearer xxxx”
-H “Content-Type: json”
question: what is the difference between these two options:
