-
Robust Preference Optimization through Reward Model Distillation
Authors:
Adam Fisch,
Jacob Eisenstein,
Vicky Zayats,
Alekh Agarwal,
Ahmad Beirami,
Chirag Nagpal,
Pete Shaw,
Jonathan Berant
Abstract:
Language model (LM) post-training (or alignment) involves maximizing a reward function that is derived from preference annotations. Direct Preference Optimization (DPO) is a popular offline alignment method that trains a policy directly on preference data without the need to train a reward model or apply reinforcement learning. However, typical preference datasets have only a single, or at most a…
▽ More
Language model (LM) post-training (or alignment) involves maximizing a reward function that is derived from preference annotations. Direct Preference Optimization (DPO) is a popular offline alignment method that trains a policy directly on preference data without the need to train a reward model or apply reinforcement learning. However, typical preference datasets have only a single, or at most a few, annotation per preference pair, which causes DPO to overconfidently assign rewards that trend towards infinite magnitude. This frequently leads to degenerate policies, sometimes causing even the probabilities of the preferred generations to go to zero. In this work, we analyze this phenomenon and propose distillation to get a better proxy for the true preference distribution over generation pairs: we train the LM to produce probabilities that match the distribution induced by a reward model trained on the preference data. Moreover, to account for uncertainty in the reward model we are distilling from, we optimize against a family of reward models that, as a whole, is likely to include at least one reasonable proxy for the preference distribution. Our results show that distilling from such a family of reward models leads to improved robustness to distribution shift in preference annotations, while preserving the simple supervised nature of DPO.
△ Less
Submitted 29 May, 2024;
originally announced May 2024.
-
Zipper: A Multi-Tower Decoder Architecture for Fusing Modalities
Authors:
Vicky Zayats,
Peter Chen,
Melissa Ferrari,
Dirk Padfield
Abstract:
Integrating multiple generative foundation models, especially those trained on different modalities, into something greater than the sum of its parts poses significant challenges. Two key hurdles are the availability of aligned data (concepts that contain similar meaning but is expressed differently in different modalities), and effectively leveraging unimodal representations in cross-domain gener…
▽ More
Integrating multiple generative foundation models, especially those trained on different modalities, into something greater than the sum of its parts poses significant challenges. Two key hurdles are the availability of aligned data (concepts that contain similar meaning but is expressed differently in different modalities), and effectively leveraging unimodal representations in cross-domain generative tasks, without compromising their original unimodal capabilities.
We propose Zipper, a multi-tower decoder architecture that addresses these concerns by using cross-attention to flexibly compose multimodal generative models from independently pre-trained unimodal decoders. In our experiments fusing speech and text modalities, we show the proposed architecture performs very competitively in scenarios with limited aligned text-speech data. We also showcase the flexibility of our model to selectively maintain unimodal (e.g., text-to-text generation) generation performance by freezing the corresponding modal tower (e.g. text). In cross-modal tasks such as automatic speech recognition (ASR) where the output modality is text, we show that freezing the text backbone results in negligible performance degradation. In cross-modal tasks such as text-to-speech generation (TTS) where the output modality is speech, we show that using a pre-trained speech backbone results in superior performance to the baseline.
△ Less
Submitted 31 May, 2024; v1 submitted 28 May, 2024;
originally announced May 2024.
-
AudioPaLM: A Large Language Model That Can Speak and Listen
Authors:
Paul K. Rubenstein,
Chulayuth Asawaroengchai,
Duc Dung Nguyen,
Ankur Bapna,
Zalán Borsos,
Félix de Chaumont Quitry,
Peter Chen,
Dalia El Badawy,
Wei Han,
Eugene Kharitonov,
Hannah Muckenhirn,
Dirk Padfield,
James Qin,
Danny Rozenberg,
Tara Sainath,
Johan Schalkwyk,
Matt Sharifi,
Michelle Tadmor Ramanovich,
Marco Tagliasacchi,
Alexandru Tudor,
Mihajlo Velimirović,
Damien Vincent,
Jiahui Yu,
Yongqiang Wang,
Vicky Zayats
, et al. (5 additional authors not shown)
Abstract:
We introduce AudioPaLM, a large language model for speech understanding and generation. AudioPaLM fuses text-based and speech-based language models, PaLM-2 [Anil et al., 2023] and AudioLM [Borsos et al., 2022], into a unified multimodal architecture that can process and generate text and speech with applications including speech recognition and speech-to-speech translation. AudioPaLM inherits the…
▽ More
We introduce AudioPaLM, a large language model for speech understanding and generation. AudioPaLM fuses text-based and speech-based language models, PaLM-2 [Anil et al., 2023] and AudioLM [Borsos et al., 2022], into a unified multimodal architecture that can process and generate text and speech with applications including speech recognition and speech-to-speech translation. AudioPaLM inherits the capability to preserve paralinguistic information such as speaker identity and intonation from AudioLM and the linguistic knowledge present only in text large language models such as PaLM-2. We demonstrate that initializing AudioPaLM with the weights of a text-only large language model improves speech processing, successfully leveraging the larger quantity of text training data used in pretraining to assist with the speech tasks. The resulting model significantly outperforms existing systems for speech translation tasks and has the ability to perform zero-shot speech-to-text translation for many languages for which input/target language combinations were not seen in training. AudioPaLM also demonstrates features of audio language models, such as transferring a voice across languages based on a short spoken prompt. We release examples of our method at https://google-research.github.io/seanet/audiopalm/examples
△ Less
Submitted 22 June, 2023;
originally announced June 2023.
-
MultiTurnCleanup: A Benchmark for Multi-Turn Spoken Conversational Transcript Cleanup
Authors:
Hua Shen,
Vicky Zayats,
Johann C. Rocholl,
Daniel D. Walker,
Dirk Padfield
Abstract:
Current disfluency detection models focus on individual utterances each from a single speaker. However, numerous discontinuity phenomena in spoken conversational transcripts occur across multiple turns, hampering human readability and the performance of downstream NLP tasks. This study addresses these phenomena by proposing an innovative Multi-Turn Cleanup task for spoken conversational transcript…
▽ More
Current disfluency detection models focus on individual utterances each from a single speaker. However, numerous discontinuity phenomena in spoken conversational transcripts occur across multiple turns, hampering human readability and the performance of downstream NLP tasks. This study addresses these phenomena by proposing an innovative Multi-Turn Cleanup task for spoken conversational transcripts and collecting a new dataset, MultiTurnCleanup1. We design a data labeling schema to collect the high-quality dataset and provide extensive data analysis. Furthermore, we leverage two modeling approaches for experimental evaluation as benchmarks for future research.
△ Less
Submitted 27 October, 2023; v1 submitted 19 May, 2023;
originally announced May 2023.
-
Teaching BERT to Wait: Balancing Accuracy and Latency for Streaming Disfluency Detection
Authors:
Angelica Chen,
Vicky Zayats,
Daniel D. Walker,
Dirk Padfield
Abstract:
In modern interactive speech-based systems, speech is consumed and transcribed incrementally prior to having disfluencies removed. This post-processing step is crucial for producing clean transcripts and high performance on downstream tasks (e.g. machine translation). However, most current state-of-the-art NLP models such as the Transformer operate non-incrementally, potentially causing unacceptab…
▽ More
In modern interactive speech-based systems, speech is consumed and transcribed incrementally prior to having disfluencies removed. This post-processing step is crucial for producing clean transcripts and high performance on downstream tasks (e.g. machine translation). However, most current state-of-the-art NLP models such as the Transformer operate non-incrementally, potentially causing unacceptable delays. We propose a streaming BERT-based sequence tagging model that, combined with a novel training objective, is capable of detecting disfluencies in real-time while balancing accuracy and latency. This is accomplished by training the model to decide whether to immediately output a prediction for the current input or to wait for further context. Essentially, the model learns to dynamically size its lookahead window. Our results demonstrate that our model produces comparably accurate predictions and does so sooner than our baselines, with lower flicker. Furthermore, the model attains state-of-the-art latency and stability scores when compared with recent work on incremental disfluency detection.
△ Less
Submitted 1 May, 2022;
originally announced May 2022.
-
Residual Adapters for Parameter-Efficient ASR Adaptation to Atypical and Accented Speech
Authors:
Katrin Tomanek,
Vicky Zayats,
Dirk Padfield,
Kara Vaillancourt,
Fadi Biadsy
Abstract:
Automatic Speech Recognition (ASR) systems are often optimized to work best for speakers with canonical speech patterns. Unfortunately, these systems perform poorly when tested on atypical speech and heavily accented speech. It has previously been shown that personalization through model fine-tuning substantially improves performance. However, maintaining such large models per speaker is costly an…
▽ More
Automatic Speech Recognition (ASR) systems are often optimized to work best for speakers with canonical speech patterns. Unfortunately, these systems perform poorly when tested on atypical speech and heavily accented speech. It has previously been shown that personalization through model fine-tuning substantially improves performance. However, maintaining such large models per speaker is costly and difficult to scale. We show that by adding a relatively small number of extra parameters to the encoder layers via so-called residual adapter, we can achieve similar adaptation gains compared to model fine-tuning, while only updating a tiny fraction (less than 0.5%) of the model parameters. We demonstrate this on two speech adaptation tasks (atypical and accented speech) and for two state-of-the-art ASR architectures.
△ Less
Submitted 14 September, 2021;
originally announced September 2021.
-
Disfluency Detection with Unlabeled Data and Small BERT Models
Authors:
Johann C. Rocholl,
Vicky Zayats,
Daniel D. Walker,
Noah B. Murad,
Aaron Schneider,
Daniel J. Liebling
Abstract:
Disfluency detection models now approach high accuracy on English text. However, little exploration has been done in improving the size and inference time of the model. At the same time, automatic speech recognition (ASR) models are moving from server-side inference to local, on-device inference. Supporting models in the transcription pipeline (like disfluency detection) must follow suit. In this…
▽ More
Disfluency detection models now approach high accuracy on English text. However, little exploration has been done in improving the size and inference time of the model. At the same time, automatic speech recognition (ASR) models are moving from server-side inference to local, on-device inference. Supporting models in the transcription pipeline (like disfluency detection) must follow suit. In this work we concentrate on the disfluency detection task, focusing on small, fast, on-device models based on the BERT architecture. We demonstrate it is possible to train disfluency detection models as small as 1.3 MiB, while retaining high performance. We build on previous work that showed the benefit of data augmentation approaches such as self-training. Then, we evaluate the effect of domain mismatch between conversational and written text on model performance. We find that domain adaptation and data augmentation strategies have a more pronounced effect on these smaller models, as compared to conventional BERT models.
△ Less
Submitted 27 July, 2021; v1 submitted 21 April, 2021;
originally announced April 2021.
-
Representations for Question Answering from Documents with Tables and Text
Authors:
Vicky Zayats,
Kristina Toutanova,
Mari Ostendorf
Abstract:
Tables in Web documents are pervasive and can be directly used to answer many of the queries searched on the Web, motivating their integration in question answering. Very often information presented in tables is succinct and hard to interpret with standard language representations. On the other hand, tables often appear within textual context, such as an article describing the table. Using the inf…
▽ More
Tables in Web documents are pervasive and can be directly used to answer many of the queries searched on the Web, motivating their integration in question answering. Very often information presented in tables is succinct and hard to interpret with standard language representations. On the other hand, tables often appear within textual context, such as an article describing the table. Using the information from an article as additional context can potentially enrich table representations. In this work we aim to improve question answering from tables by refining table representations based on information from surrounding text. We also present an effective method to combine text and table-based predictions for question answering from full documents, obtaining significant improvements on the Natural Questions dataset.
△ Less
Submitted 26 January, 2021;
originally announced January 2021.
-
Disfluencies and Human Speech Transcription Errors
Authors:
Vicky Zayats,
Trang Tran,
Richard Wright,
Courtney Mansfield,
Mari Ostendorf
Abstract:
This paper explores contexts associated with errors in transcrip-tion of spontaneous speech, shedding light on human perceptionof disfluencies and other conversational speech phenomena. Anew version of the Switchboard corpus is provided with disfluency annotations for careful speech transcripts, together with results showing the impact of transcription errors on evaluation of automatic disfluency…
▽ More
This paper explores contexts associated with errors in transcrip-tion of spontaneous speech, shedding light on human perceptionof disfluencies and other conversational speech phenomena. Anew version of the Switchboard corpus is provided with disfluency annotations for careful speech transcripts, together with results showing the impact of transcription errors on evaluation of automatic disfluency detection.
△ Less
Submitted 8 April, 2019;
originally announced April 2019.
-
Giving Attention to the Unexpected: Using Prosody Innovations in Disfluency Detection
Authors:
Vicky Zayats,
Mari Ostendorf
Abstract:
Disfluencies in spontaneous speech are known to be associated with prosodic disruptions. However, most algorithms for disfluency detection use only word transcripts. Integrating prosodic cues has proved difficult because of the many sources of variability affecting the acoustic correlates. This paper introduces a new approach to extracting acoustic-prosodic cues using text-based distributional pre…
▽ More
Disfluencies in spontaneous speech are known to be associated with prosodic disruptions. However, most algorithms for disfluency detection use only word transcripts. Integrating prosodic cues has proved difficult because of the many sources of variability affecting the acoustic correlates. This paper introduces a new approach to extracting acoustic-prosodic cues using text-based distributional prediction of acoustic cues to derive vector z-score features (innovations). We explore both early and late fusion techniques for integrating text and prosody, showing gains over a high-accuracy text-only model.
△ Less
Submitted 8 April, 2019;
originally announced April 2019.
-
Robust cross-domain disfluency detection with pattern match networks
Authors:
Vicky Zayats,
Mari Ostendorf
Abstract:
In this paper we introduce a novel pattern match neural network architecture that uses neighbor similarity scores as features, eliminating the need for feature engineering in a disfluency detection task. We evaluate the approach in disfluency detection for four different speech genres, showing that the approach is as effective as hand-engineered pattern match features when used on in-domain data a…
▽ More
In this paper we introduce a novel pattern match neural network architecture that uses neighbor similarity scores as features, eliminating the need for feature engineering in a disfluency detection task. We evaluate the approach in disfluency detection for four different speech genres, showing that the approach is as effective as hand-engineered pattern match features when used on in-domain data and achieves superior performance in cross-domain scenarios.
△ Less
Submitted 17 November, 2018;
originally announced November 2018.
-
Conversation Modeling on Reddit using a Graph-Structured LSTM
Authors:
Vicky Zayats,
Mari Ostendorf
Abstract:
This paper presents a novel approach for modeling threaded discussions on social media using a graph-structured bidirectional LSTM which represents both hierarchical and temporal conversation structure. In experiments with a task of predicting popularity of comments in Reddit discussions, the proposed model outperforms a node-independent architecture for different sets of input features. Analyses…
▽ More
This paper presents a novel approach for modeling threaded discussions on social media using a graph-structured bidirectional LSTM which represents both hierarchical and temporal conversation structure. In experiments with a task of predicting popularity of comments in Reddit discussions, the proposed model outperforms a node-independent architecture for different sets of input features. Analyses show a benefit to the model over the full course of the discussion, improving detection in both early and late stages. Further, the use of language cues with the bidirectional tree state updates helps with identifying controversial comments.
△ Less
Submitted 6 April, 2017;
originally announced April 2017.
-
Disfluency Detection using a Bidirectional LSTM
Authors:
Vicky Zayats,
Mari Ostendorf,
Hannaneh Hajishirzi
Abstract:
We introduce a new approach for disfluency detection using a Bidirectional Long-Short Term Memory neural network (BLSTM). In addition to the word sequence, the model takes as input pattern match features that were developed to reduce sensitivity to vocabulary size in training, which lead to improved performance over the word sequence alone. The BLSTM takes advantage of explicit repair states in ad…
▽ More
We introduce a new approach for disfluency detection using a Bidirectional Long-Short Term Memory neural network (BLSTM). In addition to the word sequence, the model takes as input pattern match features that were developed to reduce sensitivity to vocabulary size in training, which lead to improved performance over the word sequence alone. The BLSTM takes advantage of explicit repair states in addition to the standard reparandum states. The final output leverages integer linear programming to incorporate constraints of disfluency structure. In experiments on the Switchboard corpus, the model achieves state-of-the-art performance for both the standard disfluency detection task and the correction detection task. Analysis shows that the model has better detection of non-repetition disfluencies, which tend to be much harder to detect.
△ Less
Submitted 11 April, 2016;
originally announced April 2016.
-
Talking to the crowd: What do people react to in online discussions?
Authors:
Aaron Jaech,
Victoria Zayats,
Hao Fang,
Mari Ostendorf,
Hannaneh Hajishirzi
Abstract:
This paper addresses the question of how language use affects community reaction to comments in online discussion forums, and the relative importance of the message vs. the messenger. A new comment ranking task is proposed based on community annotated karma in Reddit discussions, which controls for topic and timing of comments. Experimental work with discussion threads from six subreddits shows th…
▽ More
This paper addresses the question of how language use affects community reaction to comments in online discussion forums, and the relative importance of the message vs. the messenger. A new comment ranking task is proposed based on community annotated karma in Reddit discussions, which controls for topic and timing of comments. Experimental work with discussion threads from six subreddits shows that the importance of different types of language features varies with the community of interest.
△ Less
Submitted 16 August, 2015; v1 submitted 8 July, 2015;
originally announced July 2015.
