On Efficient Constructions of Checkpoints
Chen, Yu, Liu, Zhenming, Ren, Bin, Jin, Xin
Efficient construction of checkpoints/snapshots is a critical tool for training and diagnosing deep learning models. In this paper, we propose a lossy compression scheme for checkpoint constructions (called LC-Checkpoint). LC-Checkpoint simultaneously maximizes the compression rate and optimizes the recovery speed, under the assumption that SGD is used to train the model. LC-Checkpointuses quantization and priority promotion to store the most crucial information for SGD to recover, and then uses a Huffman coding to leverage the non-uniform distribution of the gradient scales. Our extensive experiments show that LC-Checkpoint achieves a compression rate up to $28\times$ and recovery speedup up to $5.77\times$ over a state-of-the-art algorithm (SCAR).
Sep-27-2020
- Country:
- North America > United States
- Virginia > Williamsburg (0.04)
- Maryland > Baltimore (0.04)
- North America > United States
- Genre:
- Research Report (0.82)
- Technology: