router

if you are using the mlser server throught command line like:

mlserver start .

then use the following model-settings.json (include batching variable):

{
    "name": "audio",
    "implementation": "models.GeneralAudio",
    "max_batch_size": 5,
    "max_batch_time": 1,
    "parameters": {
        "uri": "./fakeuri"
    }
}

if you are are compling the mlserver then use the follwoing (remove batching variables):

{
    "name": "audio",
    "implementation": "models.GeneralAudio",
    "parameters": {
        "uri": "./fakeuri"
    }
}