A Holistic Assessment of the Carbon Footprint of Noor, a Very Large Arabic Language Model
arXiv:2610.00223 · doi:10.18653/v1/2022.bigscience-1.8
Abstract
As ever larger language models grow more ubiquitous, it is crucial to consider their environmental impact. Characterised by extreme size and resource use, recent generations of models have been criticised for their voracious appetite for compute, and thus significant carbon footprint. Although reporting of carbon impact has grown more common in machine learning papers, this reporting is usually limited to compute resources used strictly for training. In this work, we propose a holistic assessment of the footprint of an extreme-scale language model, Noor. Noor is an ongoing project aiming to develop the largest multi-task Arabic language models -- with up to 13B parameters -- leveraging zero-shot generalisation to enable a wide range of downstream tasks via natural language instructions. We assess the total carbon bill of the entire project: starting with data collection and storage costs, including research and development budgets, pretraining costs, future serving estimates, and other exogenous costs necessary for this international cooperation. Notably, we find that inference costs and exogenous factors can have a significant impact on total budget. Finally, we discuss pathways to reduce the carbon footprint of extreme-scale models.
11 pages, 3 figures, 2 tables. Published in Proceedings of BigScience Episode #5 -- Workshop on Challenges & Perspectives in Creating Large Language Models (ACL 2022)
References in corpus (17)
- On the Opportunities and Risks of Foundation Models
- Scaling Laws for Neural Language Models
- The Pile: An 800GB Dataset of Diverse Text for Language Modeling
- Sustainable AI: Environmental Implications, Challenges and Opportunities
- The Computational Limits of Deep Learning
- Compute Trends Across Three Eras of Machine Learning
- Using DeepSpeed and Megatron to Train Megatron-Turing NLG 530B, A Large-Scale Generative Language Model
- CCNet: Extracting High Quality Monolingual Datasets from Web Crawl Data
- Scaling Language Models: Methods, Analysis & Insights from Training Gopher
- GLaM: Efficient Scaling of Language Models with Mixture-of-Experts
- Survey of Machine Learning Accelerators
- Scaling Laws for Autoregressive Generative Modeling
- AI Accelerator Survey and Trends
- Energy Usage Reports: Environmental awareness as part of algorithmic accountability
- Estimating the carbon footprint of the GRAND Project, a multi-decade astrophysics experiment
- Unified Scaling Laws for Routed Language Models
- The carbon footprint of distributed cloud storage