Large Transformers have achieved state-of-the-art performance across many tasks. Most open-source libraries on scaling Transformers focus on improving training or inference with better parallelization. In this work, we present TorchScale, an open-source toolkit that allows researchers and developers to scale up Transformers efficiently and effectively. TorchScale has the implementation of several modeling techniques, which can improve modeling generality and capability, as well as training stability and efficiency. Experimental results on language modeling and neural machine translation demonstrate that TorchScale can successfully scale Transformers to different sizes without tears. The library is available at https://aka.ms/torchscale.

Large sequence to sequence models for tasks such as Neural Machine Translation (NMT) are usually trained over hundreds of millions of samples. However, training is just the origin of a model's life-cycle. Real-world deployments of models require further behavioral adaptations as new requirements emerge or shortcomings become known. Typically, in the space of model behaviors, behavior deletion requests are addressed through model retrainings whereas model finetuning is done to address behavior addition requests, both procedures being instances of data-based model intervention. In this work, we present a preliminary study investigating rank-one editing as a direct intervention method for behavior deletion requests in encoder-decoder transformer models. We propose four editing tasks for NMT and show that the proposed editing algorithm achieves high efficacy, while requiring only a single instance of positive example to fix an erroneous (negative) model behavior.

Training supervised video captioning model requires coupled video-caption pairs. However, for many targeted languages, sufficient paired data are not available. To this end, we introduce the unpaired video captioning task aiming to train models without coupled video-caption pairs in target language. To solve the task, a natural choice is to employ a two-step pipeline system: first utilizing video-to-pivot captioning model to generate captions in pivot language and then utilizing pivot-to-target translation model to translate the pivot captions to the target language. However, in such a pipeline system, 1) visual information cannot reach the translation model, generating visual irrelevant target captions; 2) the errors in the generated pivot captions will be propagated to the translation model, resulting in disfluent target captions. To address these problems, we propose the Unpaired Video Captioning with Visual Injection system (UVC-VI). UVC-VI first introduces the Visual Injection Module (VIM), which aligns source visual and target language domains to inject the source visual information into the target language domain. Meanwhile, VIM directly connects the encoder of the video-to-pivot model and the decoder of the pivot-to-target model, allowing end-to-end inference by completely skipping the generation of pivot captions. To enhance the cross-modality injection of the VIM, UVC-VI further introduces a pluggable video encoder, i.e., Multimodal Collaborative Encoder (MCE). The experiments show that UVC-VI outperforms pipeline systems and exceeds several supervised systems. Furthermore, equipping existing supervised systems with our MCE can achieve 4% and 7% relative margins on the CIDEr scores to current state-of-the-art models on the benchmark MSVD and MSR-VTT datasets, respectively.

Though nearest neighbor Machine Translation ($k$NN-MT) \citep{khandelwal2020nearest} has proved to introduce significant performance boosts over standard neural MT systems, it is prohibitively slow since it uses the entire reference corpus as the datastore for the nearest neighbor search. This means each step for each beam in the beam search has to search over the entire reference corpus. $k$NN-MT is thus two-orders slower than vanilla MT models, making it hard to be applied to real-world applications, especially online services. In this work, we propose Fast $k$NN-MT to address this issue. Fast $k$NN-MT constructs a significantly smaller datastore for the nearest neighbor search: for each word in a source sentence, Fast $k$NN-MT first selects its nearest token-level neighbors, which is limited to tokens that are the same as the query token. Then at each decoding step, in contrast to using the entire corpus as the datastore, the search space is limited to target tokens corresponding to the previously selected reference source tokens. This strategy avoids search through the whole datastore for nearest neighbors and drastically improves decoding efficiency. Without loss of performance, Fast $k$NN-MT is two-orders faster than $k$NN-MT, and is only two times slower than the standard NMT model. Fast $k$NN-MT enables the practical use of $k$NN-MT systems in real-world MT applications. The code is available at \url{https://github.com/ShannonAI/fast-knn-nmt}

We introduce UDAAN, an open-source post-editing tool that can reduce manual editing efforts to quickly produce publishable-standard documents in several Indic languages. UDAAN has an end-to-end Machine Translation (MT) plus post-editing pipeline wherein users can upload a document to obtain raw MT output. Further, users can edit the raw translations using our tool. UDAAN offers several advantages: a) Domain-aware, vocabulary-based lexical constrained MT. b) source-target and target-target lexicon suggestions for users. Replacements are based on the source and target texts lexicon alignment. c) Translation suggestions are based on logs created during user interaction. d) Source-target sentence alignment visualisation that reduces the cognitive load of users during editing. e) Translated outputs from our tool are available in multiple formats: docs, latex, and PDF. We also provide the facility to use around 100 in-domain dictionaries for lexicon-aware machine translation. Although we limit our experiments to English-to-Hindi translation, our tool is independent of the source and target languages. Experimental results based on the usage of the tools and users feedback show that our tool speeds up the translation time by approximately a factor of three compared to the baseline method of translating documents from scratch. Our tool is available for both Windows and Linux platforms. The tool is open-source under MIT license, and the source code can be accessed from our website at https://www.udaanproject.org. Demonstration and tutorial videos for various features of our tool can be accessed at https://www.youtube.com/channel/UClfK7iC8J7b22bj3GwAUaCw. Our MT pipeline can be accessed at https://udaaniitb.aicte-india.org/udaan/translate/.

We present our work on collecting ArzEn-ST, a code-switched Egyptian Arabic - English Speech Translation Corpus. This corpus is an extension of the ArzEn speech corpus, which was collected through informal interviews with bilingual speakers. In this work, we collect translations in both directions, monolingual Egyptian Arabic and monolingual English, forming a three-way speech translation corpus. We make the translation guidelines and corpus publicly available. We also report results for baseline systems for machine translation and speech translation tasks. We believe this is a valuable resource that can motivate and facilitate further research studying the code-switching phenomenon from a linguistic perspective and can be used to train and evaluate NLP systems.

This paper studies the limits of language models' statistical learning in the context of Zipf's law. First, we demonstrate that Zipf-law token distribution emerges irrespective of the chosen tokenization. Second, we show that Zipf distribution is characterized by two distinct groups of tokens that differ both in terms of their frequency and their semantics. Namely, the tokens that have a one-to-one correspondence with one semantic concept have different statistical properties than those with semantic ambiguity. Finally, we demonstrate how these properties interfere with statistical learning procedures motivated by distributional semantics.

I've made some progress since my previous post on machine translation for medical chat, and this is a second checkpoint. I visited a friend in Tokyo for a week and my Japanese proficiency is extremely limited, mainly coming from Duolingo and the little I remember from anime. While I was there, I kept up with my Duolingo practice and relied heavily on Google Lens, which translates text in images. Google Lens was great, and was fast with offline models. It was particularly good for translating signs, such as those in parks or tourist areas.

A supervised text classification model relies on labelled datasets to train the model (Sebastiani, 2002). From an experimental perspective, the design and evaluation of classification models typically rely on data pertaining to fixed periods of time. Recent research demonstrates that such models, while showing competitive performance in their experimental environment, underperform when they need to classify new data that is distant in time from that observed during training (Alkhalifa and Zubiaga, 2022). This deterioration of performance has been demonstrated for different classification tasks, including topic classification (Rocha, Mourão, Pereira, Gonçalves, and Meira, 2008), sentiment classification (Lukes and Søgaard, 2018), hate speech detection (Florio, Basile, Polignano, Basile, and Patti, 2020), stance detection (Alkhalifa, Kochkina, and Zubiaga, 2021) and political ideology detection (Röttger and Pierrehumbert, 2021). This performance drop can happen for multiple reasons, including among others the evolution in language use (Smith, 2004) or the evolution of public opinion (Bonilla and Mo, 2019) and its extent may vary (Alkhalifa et al., 2021). This poses an important challenge and limitation on such models when one plans to continue using the model over a long period of time to classify new, incoming data, as can be the case with a stream of user-generated contents (Cheng, Chen, Lee, and Li, 2021).

Existing work on sign language translation - that is, translation from sign language videos into sentences in a written language - has focused mainly on (1) data collected in a controlled environment or (2) data in a specific domain, which limits the applicability to real-world settings. In this paper, we introduce OpenASL, a large-scale American Sign Language (ASL) - English dataset collected from online video sites (e.g., YouTube). OpenASL contains 288 hours of ASL videos in multiple domains from over 200 signers and is the largest publicly available ASL translation dataset to date. To tackle the challenges of sign language translation in realistic settings and without glosses, we propose a set of techniques including sign search as a pretext task for pre-training and fusion of mouthing and handshape features. The proposed techniques produce consistent and large improvements in translation quality, over baseline models based on prior work. Our data and code are publicly available at https://github.com/chevalierNoir/OpenASL