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.

Existing metrics for evaluating the quality of automatically generated questions such as BLEU, ROUGE, BERTScore, and BLEURT compare the reference and predicted questions, providing a high score when there is a considerable lexical overlap or semantic similarity between the candidate and the reference questions. This approach has two major shortcomings. First, we need expensive human-provided reference questions. Second, it penalises valid questions that may not have high lexical or semantic similarity to the reference questions. In this paper, we propose a new metric, RQUGE, based on the answerability of the candidate question given the context. The metric consists of a question-answering and a span scorer module, in which we use pre-trained models from the existing literature, and therefore, our metric can be used without further training. We show that RQUGE has a higher correlation with human judgment without relying on the reference question. RQUGE is shown to be significantly more robust to several adversarial corruptions. Additionally, we illustrate that we can significantly improve the performance of QA models on out-of-domain datasets by fine-tuning on the synthetic data generated by a question generation model and re-ranked by RQUGE.

在全球范围内消除语言障碍的目标的驱动下，机器翻译已巩固自己是当今人工智能研究的关键重点。但是，这样的努力围绕着一小部分语言结合在一起，留下了绝大多数低资源的语言。在确保安全，高质量的结果的同时，在牢记道德考虑的同时，打破200个语言障碍需要什么？没有留下的语言，我们首先通过与母语人士的探索性访谈来解决对低资源语言翻译支持的必要性来应对这一挑战。然后，我们创建了旨在缩小低资源和高资源语言之间的性能差距的数据集和模型。更具体地说，我们开发了一种有条件的计算模型，基于专家的稀疏混合物，该模型经过针对针对低资源语言量身定制的新颖有效的数据挖掘技术培训的。我们提出了多次建筑和培训改进，以抵消数千个任务的培训。至关重要的是，我们使用人类翻译的基准，Flores-200评估了40,000多种不同的翻译方向的性能，并将人类评估与新型毒性基准相结合，涵盖Flores-200的所有语言，以评估翻译安全性。我们的模型相对于先前的最新技术，实现了44％BLEU的改善，为实现通用翻译系统奠定了重要的基础。最后，我们开源此工作中描述的所有贡献，可在https://github.com/facebookresearch/fairseq/tree/nllb上访问。

语言模型预训练的最新进展利用大规模数据集创建多语言模型。但是，这些数据集中大多遗漏了低资源语言。这主要是因为网络上没有很好地表示口语，因此被排除在用于创建数据集的大规模爬网中。此外，这些模型的下游用户仅限于最初选择用于预训练的语言的选择。这项工作调查了如何最佳利用现有的预培训模型来为16种非洲语言创建低资源翻译系统。我们关注两个问题：1）如何将预训练的模型用于初始预培训中未包含的语言？ 2）生成的翻译模型如何有效地转移到新域？为了回答这些问题，我们创建了一个新的非洲新闻语料库，涵盖16种语言，其中8种语言不属于任何现有评估数据集的一部分。我们证明，将两种语言转移到其他语言和其他领域的最有效策略是，以少量的高质量翻译数据微调大型预训练模型。

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.

In open-domain dialogue intelligent agents should exhibit the use of knowledge, however there are few convincing demonstrations of this to date. The most popular sequence to sequence models typically "generate and hope" generic utterances that can be memorized in the weights of the model when mapping from input utterance(s) to output, rather than employing recalled knowledge as context. Use of knowledge has so far proved difficult, in part because of the lack of a supervised learning benchmark task which exhibits knowledgeable open dialogue with clear grounding. To that end we collect and release a large dataset with conversations directly grounded with knowledge retrieved from Wikipedia. We then design architectures capable of retrieving knowledge, reading and conditioning on it, and finally generating natural responses. Our best performing dialogue models are able to conduct knowledgeable discussions on open-domain topics as evaluated by automatic metrics and human evaluations, while our new benchmark allows for measuring further improvements in this important research direction.

We explore story generation: creative systems that can build coherent and fluent passages of text about a topic. We collect a large dataset of 300K human-written stories paired with writing prompts from an online forum. Our dataset enables hierarchical story generation, where the model first generates a premise, and then transforms it into a passage of text. We gain further improvements with a novel form of model fusion that improves the relevance of the story to the prompt, and adding a new gated multi-scale self-attention mechanism to model long-range context. Experiments show large improvements over strong baselines on both automated and human evaluations. Human judges prefer stories generated by our approach to those from a strong non-hierarchical model by a factor of two to one.

The pre-dominant approach to language modeling to date is based on recurrent neural networks. Their success on this task is often linked to their ability to capture unbounded context. In this paper we develop a finite context approach through stacked convolutions, which can be more efficient since they allow parallelization over sequential tokens. We propose a novel simplified gating mechanism that outperforms Oord et al. (2016b) and investigate the impact of key architectural decisions. The proposed approach achieves state-of-the-art on the WikiText-103 benchmark, even though it features longterm dependencies, as well as competitive results on the Google Billion Words benchmark. Our model reduces the latency to score a sentence by an order of magnitude compared to a recurrent baseline. To our knowledge, this is the first time a non-recurrent approach is competitive with strong recurrent models on these large scale language tasks.

Due to the environmental impacts caused by the construction industry, repurposing existing buildings and making them more energy-efficient has become a high-priority issue. However, a legitimate concern of land developers is associated with the buildings' state of conservation. For that reason, infrared thermography has been used as a powerful tool to characterize these buildings' state of conservation by detecting pathologies, such as cracks and humidity. Thermal cameras detect the radiation emitted by any material and translate it into temperature-color-coded images. Abnormal temperature changes may indicate the presence of pathologies, however, reading thermal images might not be quite simple. This research project aims to combine infrared thermography and machine learning (ML) to help stakeholders determine the viability of reusing existing buildings by identifying their pathologies and defects more efficiently and accurately. In this particular phase of this research project, we've used an image classification machine learning model of Convolutional Neural Networks (DCNN) to differentiate three levels of cracks in one particular building. The model's accuracy was compared between the MSX and thermal images acquired from two distinct thermal cameras and fused images (formed through multisource information) to test the influence of the input data and network on the detection results.

Temporal action segmentation tags action labels for every frame in an input untrimmed video containing multiple actions in a sequence. For the task of temporal action segmentation, we propose an encoder-decoder-style architecture named C2F-TCN featuring a "coarse-to-fine" ensemble of decoder outputs. The C2F-TCN framework is enhanced with a novel model agnostic temporal feature augmentation strategy formed by the computationally inexpensive strategy of the stochastic max-pooling of segments. It produces more accurate and well-calibrated supervised results on three benchmark action segmentation datasets. We show that the architecture is flexible for both supervised and representation learning. In line with this, we present a novel unsupervised way to learn frame-wise representation from C2F-TCN. Our unsupervised learning approach hinges on the clustering capabilities of the input features and the formation of multi-resolution features from the decoder's implicit structure. Further, we provide the first semi-supervised temporal action segmentation results by merging representation learning with conventional supervised learning. Our semi-supervised learning scheme, called ``Iterative-Contrastive-Classify (ICC)'', progressively improves in performance with more labeled data. The ICC semi-supervised learning in C2F-TCN, with 40% labeled videos, performs similar to fully supervised counterparts.