大型语言模型经常经过数十万个计算天的训练，已经显示出零和少数学习的显着功能。鉴于它们的计算成本，如果没有大量资本，这些模型很难复制。对于通过API可用的少数产品，没有访问完整的模型权重，因此很难学习。我们提供开放训练的预训练变压器（OPT），这是一套仅解码器预训练的变压器，范围从12500万到175b参数，我们旨在与感兴趣的研究人员完全和负责任地分享。我们表明，OPT-175B与GPT-3相当，而仅需要1/7碳足迹才能开发。我们还释放了日志，详细介绍了我们面临的基础架构挑战，以及用于尝试所有发布模型的代码。

专家层（MOES）的混合物通过条件计算实现语言模型的高效缩放。本文提出了一个详细的实证研究，自回归鞋语言模型与广泛的设置中的密集模型相比：在域外语言建模，零和少量射击和全部微调。除了微调外，我们发现Moes基本上更加计算效率。在更适度的培训预算下，MOES可以使用$ \ SIM值4倍的计算，符合密集模型的性能。该差距在比例下变窄，但我们最大的MOE模型（1.1T参数）始终如一地优于计算等效的密集模型（6.7b参数）。总体而言，这种表现差距在任务和域中有很大差异，表明MOE和密集模型以不值得研究的方式概括不同的方式。我们使我们的代码和模型公开可用于研究使用。

GPT-3等大型自回归语言模型是几秒钟的学习者，可以在没有微调的情况下执行各种语言任务。虽然已知这些模型能够共同代表许多不同的语言，但他们的培训数据由英语主导，可能限制了它们的交叉概括。在这项工作中，我们在覆盖多种语言的平衡语料库上培训多语言自回归语言模型，并在广泛的任务中研究他们几乎没有零点的学习能力。我们最大的模型，具有75亿参数，在20多种代表语言中，在几种代表语言中，在几种代表性语言中，在几种代表性语言中，在多语言型号推理中表现出可比大小的GPT-3（在0次设置和0次拍摄设置中的绝对精度改善+ 7.4％ 4-拍摄设置中的9.4％）和自然语言推理（每次拍摄和4次设置中的每一个+ 5.4％）。在Flores-101机器翻译基准测试中，我们的模型优于GPT-3在182个翻译方向上有32个培训例子，同时超过45个方向的官方监督基线。我们介绍了模型成功和失败的位置的详细分析，特别是它尤其显示在某些任务中实现交叉语境的内容学习，而仍然存在改善表面的鲁棒性和适应没有a的任务的余地自然冻结形式。最后，我们评估我们在仇恨语音检测中以五种语言的仇恨语音检测的模型，并发现它具有与可比大小的GPT-3模型类似的限制。

本文探讨了超线性增长趋势的环境影响，从整体角度来看，跨越数据，算法和系统硬件。我们通过在行业规模机器学习用例中检查模型开发周期来表征AI计算的碳足迹，同时考虑系统硬件的生命周期。进一步迈出一步，我们捕获AI计算的操作和制造碳足迹，并为硬件 - 软件设计和尺度优化的结束分析以及如何帮助降低AI的整体碳足迹。根据行业经验和经验教训，我们分享关键挑战，并在AI的许多方面上绘制了重要的发展方向。我们希望本文提出的关键信息和见解能够激发社区以环保的方式推进AI领域。

在预介质期间，预解压器变压器遭受梯度幅度不匹配：早期层处的梯度远远大于更高层的层。我们所提出的常规程序架构可以减轻这些问题，这为每层增加了三个归一化操作：自我注意后的一层规范，自我注意输出的头部明智的缩放，以及第一完全连接层之后的层标。额外的运营产生忽略不计的计算成本（+ 0.4％的参数增加），但是改善了从12500万到27亿个参数的因果和屏蔽语言模型的预先欣赏困惑和下游任务性能。例如，在我们最强的1.3B参数基线顶部添加NARMFORMER可以在相同的计算预算中更快地达到24％的平等困惑，或者更好地收敛0.27困惑。该模型达到GPT3大（1.3B）零拍摄性能速度快60％。对于屏蔽语言建模，Normformer平均将微调胶水性能提高1.9％。 Fairseq HTTPS://github.com/pytorch/faireq/tree/main/examples/normformer提供培训ormalformer模型的代码。

在许多实际应用中，机器学习数据随着时间的流逝依次到达大块。然后，从业者必须决定如何分配其计算预算，以便在任何时间点获得最佳性能。凸优化的在线学习理论表明，最佳策略是在到达时立即使用数据。但是，这可能不是使用深度非线性网络时的最佳策略，尤其是当这些网络对每个数据进行多个数据进行多次通过时，呈现整体分布而非i.i.d ..在本文中，我们在最简单的情况下将此学习环境正式化。每个数据块都是从相同的基础分布中得出的，并首次尝试从经验回答以下问题：学习者在培训新来的块之前应该等待多长时间？学习者应该采用哪些架构？随着观察到更多的数据，学习者是否应该随着时间的推移增加能力吗？我们使用经典计算机视觉基准测试的卷积神经网络以及在大规模语言建模任务中训练的大型变压器模型进行探讨。代码可在\ url {www.github.com/facebookresearch/alma}中获得。

FAIRSEQ is an open-source sequence modeling toolkit that allows researchers and developers to train custom models for translation, summarization, language modeling, and other text generation tasks. The toolkit is based on PyTorch and supports distributed training across multiple GPUs and machines. We also support fast mixed-precision training and inference on modern GPUs. A demo video can be found here: https://www.youtube. com/watch?v=OtgDdWtHvto.

The analysis of network structure is essential to many scientific areas, ranging from biology to sociology. As the computational task of clustering these networks into partitions, i.e., solving the community detection problem, is generally NP-hard, heuristic solutions are indispensable. The exploration of expedient heuristics has led to the development of particularly promising approaches in the emerging technology of quantum computing. Motivated by the substantial hardware demands for all established quantum community detection approaches, we introduce a novel QUBO based approach that only needs number-of-nodes many qubits and is represented by a QUBO-matrix as sparse as the input graph's adjacency matrix. The substantial improvement on the sparsity of the QUBO-matrix, which is typically very dense in related work, is achieved through the novel concept of separation-nodes. Instead of assigning every node to a community directly, this approach relies on the identification of a separation-node set, which -- upon its removal from the graph -- yields a set of connected components, representing the core components of the communities. Employing a greedy heuristic to assign the nodes from the separation-node sets to the identified community cores, subsequent experimental results yield a proof of concept. This work hence displays a promising approach to NISQ ready quantum community detection, catalyzing the application of quantum computers for the network structure analysis of large scale, real world problem instances.

Recent work has shown that machine learning (ML) models can be trained to accurately forecast the dynamics of unknown chaotic dynamical systems. Such ML models can be used to produce both short-term predictions of the state evolution and long-term predictions of the statistical patterns of the dynamics (``climate''). Both of these tasks can be accomplished by employing a feedback loop, whereby the model is trained to predict forward one time step, then the trained model is iterated for multiple time steps with its output used as the input. In the absence of mitigating techniques, however, this technique can result in artificially rapid error growth, leading to inaccurate predictions and/or climate instability. In this article, we systematically examine the technique of adding noise to the ML model input during training as a means to promote stability and improve prediction accuracy. Furthermore, we introduce Linearized Multi-Noise Training (LMNT), a regularization technique that deterministically approximates the effect of many small, independent noise realizations added to the model input during training. Our case study uses reservoir computing, a machine-learning method using recurrent neural networks, to predict the spatiotemporal chaotic Kuramoto-Sivashinsky equation. We find that reservoir computers trained with noise or with LMNT produce climate predictions that appear to be indefinitely stable and have a climate very similar to the true system, while reservoir computers trained without regularization are unstable. Compared with other types of regularization that yield stability in some cases, we find that both short-term and climate predictions from reservoir computers trained with noise or with LMNT are substantially more accurate. Finally, we show that the deterministic aspect of our LMNT regularization facilitates fast hyperparameter tuning when compared to training with noise.

Large language models (LLMs) have been shown to be able to perform new tasks based on a few demonstrations or natural language instructions. While these capabilities have led to widespread adoption, most LLMs are developed by resource-rich organizations and are frequently kept from the public. As a step towards democratizing this powerful technology, we present BLOOM, a 176B-parameter open-access language model designed and built thanks to a collaboration of hundreds of researchers. BLOOM is a decoder-only Transformer language model that was trained on the ROOTS corpus, a dataset comprising hundreds of sources in 46 natural and 13 programming languages (59 in total). We find that BLOOM achieves competitive performance on a wide variety of benchmarks, with stronger results after undergoing multitask prompted finetuning. To facilitate future research and applications using LLMs, we publicly release our models and code under the Responsible AI License.