Amazon Technologies, Inc.

Runfei Luo

Vineet Khare

Gurumurthy Swaminathan

Xiong Zhou

Ragav Venkatesan

Saurabh Gupta

Bharathan Balaji

Yijie Zhuang

Siddhartha Agarwal

Compression profiles may be searched for trained neural networks. An iterative compression profile search may be performed response to a search request. Different prospective compression profiles may be generated for trained neural networks according to a search policy. Performance of compressed versions of the trained neural networks according to the compression profiles may be tracked. The search policy may be updated according to an evaluation of the performance of the compression profiles for the compressed versions of the trained neural networks using compression performance criteria. When a search criteria is satisfied, a result for the compression profile search may be provided.

Searching compression profiles for trained neural networks

Neural networks with similar architectures may be compressed using shared compression profiles. A request to compress a trained neural network may be received and an architecture of the neural network identified. The identified architecture may be compared with the different network architectures mapped to compression profiles to select a compression profile for the neural network. The compression profile may be applied to remove features of the neural network to generate a compressed version of the neural network.

Applying compression profiles across similar neural network architectures

Techniques for iterative model training and deployment for automated learning systems are described. A method of iterative model training and deployment for automated learning systems comprises generating training data based on inference data, provided by a first version of a model hosted at an endpoint of a machine learning service, and feedback data, received from a client application, using an identifier associated with the inference data and the feedback data, generating a second version of the model using the training data, and deploying the model to the endpoint of the machine learning service.

Iterative model training and deployment for automated learning systems

A compression policy to produce compression profiles for compressing trained machine learning models may be trained using reinforcement learning. An iterative reinforcement learning may be performed response to a search request. Different prospective compression profiles may be generated for received machine learning models according to a compression policy being trained. Performance of compressed versions of the trained neural networks according to the compression profiles may be caused using data sets used to train the machine learning models. The compression policy may be updated according to reward signal determined from an application of a reward function for performance criteria to performance results of the different versions of the machine learning models. When a search criteria is satisfied, the trained compression policy may be provided.

Growing community of inventors

Kirkland, WA, United States of America

Runfei Luo