SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2102.02888 · 2021

1-bit Adam: Communication Efficient Large-Scale Training with Adam's Convergence Speed

arXiv · PDF · Open in the Atlas

Code that ran

We have not lifted any functions out of this paper's repositories yet, so there is nothing we have run. The repositories linked to it are listed below.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Scalable training of large models (like BERT and GPT-3) requires careful optimization rooted in model design, architecture, and system capabilities. From a system standpoint, communication has become a major bottleneck, especially on commodity systems with standard TCP interconnects that offer limited network bandwidth. Communication compression is an important technique to reduce training time on such systems. One of the most effective methods is error-compensated compression, which offers robust convergence speed even under 1-bit compression. However, state-of-the-art error compensation techniques only work with basic optimizers like SGD and momentum SGD, which are linearly dependent on the gradients. They do not work with non-linear gradient-based optimizers like Adam, which offer state-of-the-art convergence efficiency and accuracy for models like BERT. In this paper, we propose 1-bit Adam that reduces the communication volume by up to $5\times$, offers much better scalability, and provides the same convergence speed as uncompressed Adam. Our key finding is that Adam's variance (non-linear term) becomes stable (after a warmup phase) and can be used as a fixed precondition for the rest of the training (compression phase). Experiments on up to 256 GPUs show that 1-bit Adam enables up to $3.3\times$ higher throughput for BERT-Large pre-training and up to $2.9\times$ higher throughput for SQuAD fine-tuning. In addition, we provide theoretical analysis for our proposed work.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2102.02888")
get_code_for_paper("2102.02888")
have("2102.02888")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2102.02888.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2102.02888.svg)](https://syntology.ai/paper/2102.02888)

Connect an agent — have() is free.