CODEWITHSUNDEEP
azurespeech-to-textazure-cognitive-services
Andrii Dusheba
Andrii Dusheba

Reputation: 11

Azure Custom Speech Model: Use existing model as a base model

I created model base on my datasets. Can I use this model as a base model for creation of another model, if I want to append more datasets to existing ones?

when I try to do so, I get http 400, and message: "The base model isn't valid for this operation"

Upvotes: 0

Views: 312

Answers (1)

Navaneeth Sen
Navaneeth Sen

Reputation: 6476

You don't need to make your custom model the base model. The custom model is something which was created over the base model. All you need is to find the endpointId of the custom model and then execute your request against that model either using the sdk or using the rest api.

The below is how a sample custom endpoint response looks like.
The id which ends in the "f203" is the custom model endpointId and the id which ends in "ddcd" is the base model endpointId.

{
      "self": "https://westcentralus.api.cognitive.microsoft.com/speechtotext/v3.0/models/****1e6c-****-****-****-c25f****f203",
      "baseModel": {
        "self": "https://westcentralus.api.cognitive.microsoft.com/speechtotext/v3.0/models/base/****f802-****-****-****-8f1b****ddcd"
      },
      "datasets": [],
      "links": {
        "manifest": "https://westcentralus.api.cognitive.microsoft.com/speechtotext/v3.0/models/****1e6c-****-****-****-c25f****f203/manifest",
        "copyTo": "https://westcentralus.api.cognitive.microsoft.com/speechtotext/v3.0/models/****1e6c-****-****-****-c25f****f203/copyto"
      },
      "properties": {
        "deprecationDates": {
          "adaptationDateTime": "2021-10-15T00:00:00Z",
          "transcriptionDateTime": "2022-10-15T00:00:00Z"
        }
      },
      "lastActionDateTime": "2020-08-27T07:20:50Z",
      "status": "Succeeded",
      "createdDateTime": "2020-08-27T07:18:36Z",
      "locale": "en-NZ",
      "displayName": "Iteration 1",
      "description": "Baseline Model with Random Splitting of Dataset",
      "customProperties": {
        "PortalAPIVersion": "3"
      }
    }

Please have a look at the below link and play around with the apis.

Speech to Text API v3.0 REST Console

Upvotes: 1

Related Questions