(function() { var utmInheritingDomain = "appstore.com", utmRegExp = /(&|\?)utm_[A-Za-z]+=[A-Za-z0-9]+/gi, links = document.getElementsByTagName("a"), utms = [ "utm_medium={{URL – utm_medium}}", "utm_source={{URL – utm_source}}", "utm_campaign={{URL – utm_campaign}}" ]; for (var index = 0; index < links.length; index += 1) { var tempLink = links[index].href, tempParts; if (tempLink.indexOf(utmInheritingDomain) > 0) { tempLink = tempLink.replace(utmRegExp, ""); tempParts = tempLink.split("#"); if (tempParts[0].indexOf("?") < 0 ) { tempParts[0] += "?" + utms.join("&"); } else { tempParts[0] += "&" + utms.join("&"); } tempLink = tempParts.join("#"); } links[index].href = tempLink; } }());

BPipe: Memory-Balanced Pipeline Parallelism
for Training Large Language Models

See full paper

This paper was published by FriendliAI and SNU at ICML ‘23.

Abstract:

Pipeline parallelism is a key technique for training large language models within GPU clusters. However, it often leads to a memory imbalance problem, where certain GPUs face high memory pressure while others underutilize their capacity. This imbalance results in suboptimal training performance, even when the overall GPU memory capacity is sufficient for more efficient setups. To address this inefficiency, we propose BPipe, a novel approach for achieving memory balance in pipeline parallelism. BPipe employs an activation balancing method to transfer intermediate activations between GPUs during training, enabling all GPUs to utilize comparable amounts of memory. With balanced memory utilization, BPipe enhances the training efficiency of large language models like GPT-3 by eliminating redundant recomputations or increasing the micro-batch size. Our evaluation conducted on 48 A100 GPUs across six nodes interconnected with HDR InfiniBand shows that BPipe accelerates the training of GPT-3 96B and GPT-3 134B models by 1.25x-2.17x compared to Megatron-LM, a state-of-the-art framework for training large language models.


Reference & Presentation:

ICML ‘23