Slapo: A Schedule Language for Progressive Optimization of Large Deep Learning Model Training

• URL: http://arxiv.org/abs/2302.08005v2
• Date: Sat, 23 Dec 2023 03:52:35 GMT
• Title: Slapo: A Schedule Language for Progressive Optimization of Large Deep Learning Model Training
• Authors: Hongzheng Chen, Cody Hao Yu, Shuai Zheng, Zhen Zhang, Zhiru Zhang, Yida Wang
• Abstract summary: Slapo is a schedule language that decouples the execution of a tensor-level operator from its arithmetic definition. We show that Slapo can improve training throughput by up to 2.92x on a single machine with 8 NVIDIA V100 GPUs.
• Score: 17.556432199389615
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Recent years have seen an increase in the development of large deep learning (DL) models, which makes training efficiency crucial. Common practice is struggling with the trade-off between usability and performance. On one hand, DL frameworks such as PyTorch use dynamic graphs to facilitate model developers at a price of sub-optimal model training performance. On the other hand, practitioners propose various approaches to improving the training efficiency by sacrificing some of the flexibility, ranging from making the graph static for more thorough optimization (e.g., XLA) to customizing optimization towards large-scale distributed training (e.g., DeepSpeed and Megatron-LM). In this paper, we aim to address the tension between usability and training efficiency through separation of concerns. Inspired by DL compilers that decouple the platform-specific optimizations of a tensor-level operator from its arithmetic definition, this paper proposes a schedule language, Slapo, to decouple model execution from definition. Specifically, Slapo works on a PyTorch model and uses a set of schedule primitives to convert the model for common model training optimizations such as high-performance kernels, effective 3D parallelism, and efficient activation checkpointing. Compared to existing optimization solutions, Slapo progressively optimizes the model "as-needed" through high-level primitives, and thus preserving programmability and debuggability for users to a large extent. Our evaluation results show that by scheduling the existing hand-crafted optimizations in a systematic way using Slapo, we are able to improve training throughput by up to 2.92x on a single machine with 8 NVIDIA V100 GPUs, and by up to 1.41x on multiple machines with up to 64 GPUs, when compared to the out-of-the-box performance of DeepSpeed and Megatron-LM.

Related papers

• Efficient Deep Learning Board: Training Feedback Is Not All You Need [28.910266386748525]
  We propose EfficientDL, an innovative deep learning board for automatic performance prediction and component recommendation. The magic of no training feedback comes from our proposed comprehensive, multi-dimensional, fine-grained system component dataset. For example, EfficientDL operates seamlessly with mainstream models such as ResNet50, MobileNetV3, EfficientNet-B0, MaxViT-T, Swin-B, and DaViT-T.
  arXiv  Detail & Related papers  (2024-10-17T14:43:34Z)
• CoLLiE: Collaborative Training of Large Language Models in an Efficient Way [59.09824823710863]
  CoLLiE is an efficient library that facilitates collaborative training of large language models. With its modular design and comprehensive functionality, CoLLiE offers a balanced blend of efficiency, ease of use, and customization.
  arXiv  Detail & Related papers  (2023-12-01T08:02:16Z)
• Harnessing Deep Learning and HPC Kernels via High-Level Loop and Tensor Abstractions on CPU Architectures [67.47328776279204]
  This work introduces a framework to develop efficient, portable Deep Learning and High Performance Computing kernels. We decompose the kernel development in two steps: 1) Expressing the computational core using Processing Primitives (TPPs) and 2) Expressing the logical loops around TPPs in a high-level, declarative fashion. We demonstrate the efficacy of our approach using standalone kernels and end-to-end workloads that outperform state-of-the-art implementations on diverse CPU platforms.
  arXiv  Detail & Related papers  (2023-04-25T05:04:44Z)
• RAF: Holistic Compilation for Deep Learning Model Training [17.956035630476173]
  In this paper, we present RAF, a deep learning compiler for training. Unlike existing DLCs, RAF accepts a forward model and in-house generates a training graph. RAF is able to systematically consolidate graph optimizations for performance, memory and distributed training.
  arXiv  Detail & Related papers  (2023-03-08T17:51:13Z)
• VeLO: Training Versatile Learned Optimizers by Scaling Up [67.90237498659397]
  We leverage the same scaling approach behind the success of deep learning to learn versatiles. We train an ingest for deep learning which is itself a small neural network that ingests and outputs parameter updates. We open source our learned, meta-training code, the associated train test data, and an extensive benchmark suite with baselines at velo-code.io.
  arXiv  Detail & Related papers  (2022-11-17T18:39:07Z)
• Joint inference and input optimization in equilibrium networks [68.63726855991052]
  deep equilibrium model is a class of models that foregoes traditional network depth and instead computes the output of a network by finding the fixed point of a single nonlinear layer. We show that there is a natural synergy between these two settings. We demonstrate this strategy on various tasks such as training generative models while optimizing over latent codes, training models for inverse problems like denoising and inpainting, adversarial training and gradient based meta-learning.
  arXiv  Detail & Related papers  (2021-11-25T19:59:33Z)
• Large Language Models Can Be Strong Differentially Private Learners [70.0317718115406]
  Differentially Private (DP) learning has seen limited success for building large deep learning models of text. We show that this performance drop can be mitigated with the use of large pretrained models. We propose a memory saving technique that allows clipping in DP-SGD to run without instantiating per-example gradients.
  arXiv  Detail & Related papers  (2021-10-12T01:45:27Z)
• MetaTune: Meta-Learning Based Cost Model for Fast and Efficient Auto-tuning Frameworks [0.0]
  This paper proposes MetaTune, a meta-learning based cost model that more quickly and accurately predicts the performance of optimized codes with pre-trained model parameters. The framework provides 8 to 13% better inference time on average for four CNN models with comparable or lower optimization time while outperforming transfer learning by 10% in cross-platform cases.
  arXiv  Detail & Related papers  (2021-02-08T13:59:08Z)
• Woodpecker-DL: Accelerating Deep Neural Networks via Hardware-Aware Multifaceted Optimizations [15.659251804042748]
  Woodpecker-DL (WPK) is a hardware-aware deep learning framework. WPK uses graph optimization, automated searches, domain-specific language ( DSL) and system-level exploration to accelerate inference. We show that on a maximum P100 GPU, we can achieve the speedup of 5.40 over cuDNN and 1.63 over TVM on individual operators, and run up to 1.18 times faster than TeslaRT for end-to-end model inference.
  arXiv  Detail & Related papers  (2020-08-11T07:50:34Z)
• Bayesian Optimization for Selecting Efficient Machine Learning Models [53.202224677485525]
  We present a unified Bayesian Optimization framework for jointly optimizing models for both prediction effectiveness and training efficiency. Experiments on model selection for recommendation tasks indicate models selected this way significantly improves model training efficiency.
  arXiv  Detail & Related papers  (2020-08-02T02:56:30Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.