On the Efficient Representation and Execution of Deep Acoustic Models
Citations Over TimeTop 10% of 2016 papers
Abstract
In this paper we present a simple and computationally efficient quantization scheme that enables us to reduce the resolution of the parameters of a neural network from 32-bit floating point values to 8-bit integer values.The proposed quantization scheme leads to significant memory savings and enables the use of optimized hardware instructions for integer arithmetic, thus significantly reducing the cost of inference.Finally, we propose a 'quantization aware' training process that applies the proposed scheme during network training and find that it allows us to recover most of the loss in accuracy introduced by quantization.We validate the proposed techniques by applying them to a long short-term memory-based acoustic model on an open-ended large vocabulary speech recognition task.
Related Papers
- → An adaptive caching algorithm suitable for time-varying user accesses in VOD systems(2014)10 cited
- Study and Two Types of Typical Usage of DataGrid Web Server Control(2005)
- Using DataGrid Control to Realize DataBase of Querying in VB6.0(2000)
- Susquehanna Chorale Spring Concert "Roots and Wings"(2017)
- → DETERMINING QUALITY REQUIREMENTS AT THE UNIVERSITIES TO IMPROVE THE QUALITY OF EDUCATION(2018)