Hire Weiran D. - Fine-Tuning Engineer

Design and optimization of efficient machine learning models and algorithms for various products.

* Developed an optimization and compression toolkit for accelerating neural networks with applications in computer vision, game AI NPC components, and sequential models deployed at servers and mobile platforms. * Distributed neural network optimization training for reinforcement learning. * Supported multiple projects on neural network optimization and compression.

* Deep Learning R&D * Manages neural processor software and algorithm team * Optimization of deep neural networks for high performance training and inference * Automated design and optimizations of neural network by learning (self-learning, reinforcement learning etc) * Design of recurrent neural network models for time-series predictions for mobile health applications * Efficient use of computing resources through dynamic distribution of training batches * R&D on imaging processing and depth imaging technologies on embedded Linux platforms targeted to consumer electronics * Patent * Apparatus and method for design of efficient convolution for convolution neural networks * A method for optimizing neural networks through stochastic perturbation * An automated self-pruning method for deep neural network compression * A method to balance sparsity for efficient inference of deep neural networks * A normalization method for training deep neural networks * A method for optimizing neural networks