BPPSA: Scaling Back-propagation by Parallel Scan Algorithm
This addresses scalability issues in deep learning training for practitioners using model parallelism, though it is incremental as it builds on existing scan algorithms.
The paper tackles the sequential dependency in back-propagation that hinders scalability on parallel systems by reformulating it into a scan operation, achieving up to 2.75x speedup in overall training time and 108x speedup in the backward pass.
In an era when the performance of a single compute device plateaus, software must be designed to scale on massively parallel systems for better runtime performance. However, in the context of training deep learning models, the popular back-propagation (BP) algorithm imposes a strong sequential dependency in the process of gradient computation. Under model parallelism, BP takes $Θ(n)$ steps to complete which hinders its scalability on parallel systems ($n$ represents the number of compute devices into which a model is partitioned). In this work, in order to improve the scalability of BP, we reformulate BP into a scan operation which is a primitive that performs an in-order aggregation on a sequence of values and returns the partial result at each step. We can then scale such reformulation of BP on parallel systems by our modified version of the Blelloch scan algorithm which theoretically takes $Θ(\log n)$ steps. We evaluate our approach on a vanilla Recurrent Neural Network (RNN) training with synthetic datasets and a RNN with Gated Recurrent Units (GRU) training with the IRMAS dataset, and demonstrate up to $2.75\times$ speedup on the overall training time and $108\times$ speedup on the backward pass. We also demonstrate that the retraining of pruned networks can be a practical use case of our method.