The baseline model is the starting point for training. If you're not sure which scenario to choose, select General. In the Scenario and Baseline model list, select the scenario that best fits your domain.Give your training a Name and Description.Go to Speech-to-text > Custom Speech > Training.If you plan to train a model with audio + human-labeled transcription datasets, pick a Speech subscription in a region with dedicated hardware for training. After you upload training data, follow these instructions to start training your model: See Prepare and test your data for step-by-step instructions to prepare human-labeled transcriptions and related text (utterances and pronunciations). The first step to train a model is to upload training data. Improve recognition accuracy on speaking styles, accents, or specific background noises Pronunciation data or phonetic pronunciation in structured text Improve recognition accuracy on industry-specific vocabulary and grammar, like medical terminology or IT jargonĭefine the phonetic and displayed form of a word or term that has nonstandard pronunciation, like product names or acronyms Use this table to determine which dataset to use to address your problems: Use case If you're encountering recognition problems with a base model, you can use human-labeled transcripts and related data to train a custom model and help improve accuracy. Use training to resolve accuracy problems These datasets, along with previously uploaded audio data, are used to refine and train the speech-to-text model. You use human-labeled transcriptions and related text to train a model. Training a speech-to-text model can improve recognition accuracy for the Microsoft baseline model. In this article, you'll learn how to train and deploy Custom Speech models.
0 kommentar(er)
