paper-with-me

Papers

Computational Bottlenecks of Training Small-scale Large Language Models

2024-10-25 · Saleh Ashkboos, Iman Mirzadeh, Keivan Alizadeh, Mohammad Hossein Sekhavat, Moin Nabi, Mehrdad Farajtabar, Fartash Faghri

While large language models (LLMs) dominate the AI landscape, Small-scale large Language Models (SLMs) are gaining attention due to cost and efficiency demands from consumers. However, there is limited research on the training behavior and computational requirements of SLMs. In this study, we explore the computational bottlenecks of training SLMs (up to 2B parameters) by examining the effects of various hyperparameters and configurations, including GPU type, batch size, model size, communication protocol, attention type, and the number of GPUs. We assess these factors on popular cloud services using metrics such as loss per dollar and tokens per second. Our findings aim to support the broader adoption and optimization of language model training for low-resource AI research institutes.

📄 PDF Abstract BibTeX arXiv:2410.19456

Code (0)

등록된 구현이 없습니다.

Tasks

GPULanguage ModelingLanguage Modelling

Methods 이 논문이 사용한 방법론

Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Attention 설명 없음

Similar Papers 제목 키워드 기반

BootSeer: Analyzing and Mitigating Initialization Bottlenecks in Large-Scale LLM Training

2025-07-16 · Rui Li, Xiaoyun Zhi, Jinxin Chi, Menghan Yu 외 arxiv

Large Language Models (LLMs) have become a cornerstone of modern AI, driving breakthroughs in natural language processing and expanding into multimodal jobs involving images, audio, and video. As with most computational …

SNAP-V: A RISC-V SoC with Configurable Neuromorphic Acceleration for Small-Scale Spiking Neural Networks

2026-03-12 · Kanishka Gunawardana, Sanka Peeris, Kavishka Rambukwella, Thamish Wanduragala 외 arxiv

Spiking Neural Networks (SNNs) have gained significant attention in edge computing due to their low power consumption and computational efficiency. However, existing implementations either use conventional System on Chip…

Computational Efficiency

EDiT: A Local-SGD-Based Efficient Distributed Training Method for Large Language Models

2024-12-10 · JiaLiang Cheng, Ning Gao, Yun Yue, Zhiling Ye 외

Distributed training methods are crucial for large language models (LLMs). However, existing distributed training methods often suffer from communication bottlenecks, stragglers, and limited elasticity, particularly in h…

Towards provably efficient quantum algorithms for large-scale machine-learning models

2023-03-06 · Junyu Liu, Minzhao Liu, Jin-Peng Liu, Ziyu Ye 외

Large machine learning models are revolutionary technologies of artificial intelligence whose bottlenecks include huge computational expenses, power, and time used both in the pre-training and fine-tuning process. In thi…

SCALE: Selective Resource Allocation for Overcoming Performance Bottlenecks in Mathematical Test-time Scaling

2025-11-29 · Yang Xiao, Chunpu Xu, Ruifeng Yuan, Jiashuo Wang 외 arxiv

Test-time compute scaling has emerged as a powerful paradigm for enhancing mathematical reasoning in large language models (LLMs) by allocating additional computational resources during inference. However, current method…

Mathematical Reasoning