Asynchronous optimization for sequence training of neural networks
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for obtaining, by a first sequence-training speech model, a first batch of training frames that represent speech features of first training utterances; obtaining, by the first sequence-training speech model, one or more first neural network parameters; determining, by the first sequence-training speech model, one or more optimized first neural network parameters based on (i) the first batch of training frames and (ii) the one or more first neural network parameters; obtaining, by a second sequence-training speech model, a second batch of training frames that represent speech features of second training utterances; obtaining one or more second neural network parameters; and determining, by the second sequence-training speech model, one or more optimized second neural network parameters based on (i) the second batch of training frames and (ii) the one or more second neural network parameters.
1. A computer-implemented method executed on data processing hardware that causes the data processing hardware to perform operations comprising:
receiving a current set of parameter values for a neural network model executable on the data processing hardware;
obtaining training utterances each comprising one or more predetermined words spoken by a speaker;
processing the training utterances using the current set of parameter values for the neural network model;
determining updated parameter values for the neural network model based on the processing of the training utterances; and
performing stochastic gradient descent optimization to determine corresponding model parameter gradients for the neural network model based on differences between the current set of parameter values and the updated parameter values for the neural network model.
2. The computer-implemented method of claim 1 , wherein the operations further comprise sending the corresponding model parameter gradients for the neural network model to a centralized server.
3. The computer-implemented method of claim 2 , wherein sending the corresponding model parameter gradients for the neural network model comprises sending the corresponding model parameter gradients for the neural network model to the centralized network without sending the batch of training utterances to the centralized server.
4. The computer-implemented method of claim 1 , wherein the current set of parameter values for the neural network model comprise weights and biases of hidden layers of the neural network model.
5. The computer-implemented method of claim 4 , wherein the updated parameter values for the neural network model comprise updated weights for the neural network model.
6. The computer-implemented method of claim 1 , wherein the neural network model is trained to indicate likelihoods that acoustic feature vectors represent different phonetic units.
7. The computer-implemented method of claim 1 , wherein the data processing hardware resides on a respective computing device associated with the speaker that spoke each of the training utterances.
8. The computer-implemented method of claim 7 , wherein each of the training utterances is recorded by the respective computing device.
9. The computer-implemented method of claim 7 , wherein the respective computing device is in communication with a centralized server via a wide area network.
10. The computer-implemented method of claim 7 , wherein the respective computing device comprises a mobile audio player.
11. A system comprising:
data processing hardware; and
memory hardware in communication with the data processing hardware and storing instructions that when executed on the data processing hardware cause the data processing hardware to perform operations comprising:
receiving a current set of parameter values for a neural network model executable on the data processing hardware;
obtaining training utterances each comprising one or more predetermined words spoken by a speaker;
processing the training utterances using the current set of parameter values for the neural network model;
determining updated parameter values for the neural network model based on the processing of the training utterances; and
performing stochastic gradient descent optimization to determine corresponding model parameter gradients for the neural network model based on differences between the current set of parameter values and the updated parameter values for the neural network model.
12. The system of claim 11 , wherein the operations further comprise sending the corresponding model parameter gradients for the neural network model to a centralized server.
13. The system of claim 12 , wherein sending the corresponding model parameter gradients for the neural network model comprises sending the corresponding model parameter gradients for the neural network model to the centralized network without sending the batch of training utterances to the centralized server.
14. The system of claim 11 , wherein the current set of parameter values for the neural network model comprise weights and biases of hidden layers of the neural network model.
15. The system of claim 14 , wherein the updated parameter values for the neural network model comprise updated weights for the neural network model.
16. The system of claim 11 , wherein the neural network model is trained to indicate likelihoods that acoustic feature vectors represent different phonetic units.
17. The system of claim 11 , wherein the data processing hardware resides on a respective computing device associated with the speaker that spoke each of the training utterances.
18. The system of claim 17 , wherein each of the training utterances is recorded by the respective computing device.
19. The system of claim 17 , wherein the respective computing device is in communication with a centralized server via a wide area network.
20. The system of claim 17 , wherein the respective computing device comprises a mobile audio player.