OpEvo: An Evolutionary Method for Tensor Operator Optimization
This work addresses the challenge of automatically optimizing tensor operators for deep neural networks, which is crucial for improving training and inference efficiency on hardware platforms, representing a novel method for a known bottleneck.
The authors tackled the problem of inefficient tensor operator optimization due to combinatorial search spaces by proposing OpEvo, an evolutionary method that uses topology-aware mutation, resulting in finding the best configuration with the lowest variance and least trials and wall-clock time compared to SOTA methods.
Training and inference efficiency of deep neural networks highly rely on the performance of tensor operators on hardware platforms. Manually optimizing tensor operators has limitations in terms of supporting new operators or hardware platforms. Therefore, automatically optimizing device code configurations of tensor operators is getting increasingly attractive. However, current methods for tensor operator optimization usually suffer from poor sample-efficiency due to the combinatorial search space. In this work, we propose a novel evolutionary method, OpEvo, which efficiently explores the search spaces of tensor operators by introducing a topology-aware mutation operation based on q-random walk to leverage the topological structures over the search spaces. Our comprehensive experiment results show that compared with state-of-the-art (SOTA) methods OpEvo can find the best configuration with the lowest variance and least efforts in the number of trials and wall-clock time. All code of this work is available online.