Qualcomm Incorporated

Marinus Willem Van Baalen

Tijmen Pieter Frederik Blankevoort

Markus Nagel

Christos Louizos

Rana Ali Amjad

Andrey Kuzmin

Andrii Skliar

Brian Kahne

A processor-implemented method includes bit shifting a binary representation of a neural network parameter. The neural network parameter has fewer bits, b, than a number of hardware bits, B, supported by hardware that processes the neural network parameter. The bit shifting effectively multiplies the neural network parameter by 2. The method also includes dividing a quantization scale by 2to obtain an updated quantization scale. The method further includes quantizing the bit shifted binary representation with the updated quantization scale to obtain a value for the neural network parameter.

Simulated low bit-width quantization using bit shifted neural network parameters

Various embodiments include methods and devices for joint mixed-precision quantization and structured pruning. Embodiments may include determining whether a plurality of gates of quantization and pruning gates are selected for combination, and in response to determining that the plurality of gates are selected for combination, iteratively for each successive gate of the plurality of gates selected for combination quantizing a residual error of a quantized tensor to a scale of a next bit-width producing a residual error quantized tensor in which the next bit-width increases for each successive iteration, and adding the quantized tensor and the residual error quantized tensor producing a next quantized tensor in which the next quantized tensor has the next bit-width, and in which the next quantized tensor is the quantized tensor for a successive iteration.

Bayesian bits joint mixed-precision quantization and structured pruning using decomposed quantization and Bayesian gates

Various embodiments include methods and neural network computing devices implementing the methods for performing quantization in neural networks. Various embodiments may include equalizing ranges of weight tensors or output channel weights within a first layer of the neural network by scaling each of the output channel weights of the first layer by a corresponding scaling factor, and scaling each of a second adjacent layer's corresponding input channel weights by applying an inverse of the corresponding scaling factor to the input channel weights. The corresponding scaling factor may be determined using a black-box optimizer on a quantization error metric or based on heuristics, equalization of dynamic ranges, equalization of range extrema (minima or maxima), differential learning using straight through estimator (STE) methods and a local or global loss, or using an error metric for the quantization error and a black-box optimizer that minimizes the error metric with respect to the scaling.

Systems and methods of cross layer rescaling for improved quantization performance

Various embodiments include methods and neural network computing devices implementing the methods, for generating an approximation neural network. Various embodiments may include performing approximation operations on a weights tensor associated with a layer of a neural network to generate an approximation weights tensor, determining an expected output error of the layer in the neural network due to the approximation weights tensor, subtracting the expected output error from a bias parameter of the layer to determine an adjusted bias parameter and substituting the adjusted bias parameter for the bias parameter in the layer. Such operations may be performed for one or more layers in a neural network to produce an approximation version of the neural network for execution on a resource limited processor.

Growing community of inventors

Amsterdam, Netherlands

Marinus Willem Van Baalen