Piero Molino

Justin Zhao, Timothy Wang, Wael Abid et al.

Low Rank Adaptation (LoRA) has emerged as one of the most widely adopted methods for Parameter Efficient Fine-Tuning (PEFT) of Large Language Models (LLMs). LoRA reduces the number of trainable parameters and memory usage while achieving comparable performance to full fine-tuning. We aim to assess the viability of training and serving LLMs fine-tuned with LoRA in real-world applications. First, we measure the quality of LLMs fine-tuned with quantized low rank adapters across 10 base models and 31 tasks for a total of 310 models. We find that 4-bit LoRA fine-tuned models outperform base models by 34 points and GPT-4 by 10 points on average. Second, we investigate the most effective base models for fine-tuning and assess the correlative and predictive capacities of task complexity heuristics in forecasting the outcomes of fine-tuning. Finally, we evaluate the latency and concurrency capabilities of LoRAX, an open-source Multi-LoRA inference server that facilitates the deployment of multiple LoRA fine-tuned models on a single GPU using shared base model weights and dynamic adapter loading. LoRAX powers LoRA Land, a web application that hosts 25 LoRA fine-tuned Mistral-7B LLMs on a single NVIDIA A100 GPU with 80GB memory. LoRA Land highlights the quality and cost-effectiveness of employing multiple specialized LLMs over a single, general-purpose LLM.

Piero Molino, Jacopo Tagliabue

We examine how much of the contemporary progress in artificial intelligence (and, specifically, in natural language processing), can be, more or less directly, traced back to the seminal work and ideas of the Austrian-British philosopher Ludwig Wittgenstein, with particular focus on his late views. Discussing Wittgenstein's original theses will give us the chance to survey the state of artificial intelligence, and comment on both its strengths and weaknesses. A similar text appeared first in Spanish as a chapter of CENTENARIO DEL SILENCIO (2021), a book celebrating 100 years since the publication of the Tractatus.

Avanika Narayan, Piero Molino, Karan Goel et al.

The rapid proliferation of machine learning models across domains and deployment settings has given rise to various communities (e.g. industry practitioners) which seek to benchmark models across tasks and objectives of personal value. Unfortunately, these users cannot use standard benchmark results to perform such value-driven comparisons as traditional benchmarks evaluate models on a single objective (e.g. average accuracy) and fail to facilitate a standardized training framework that controls for confounding variables (e.g. computational budget), making fair comparisons difficult. To address these challenges, we introduce the open-source Ludwig Benchmarking Toolkit (LBT), a personalized benchmarking toolkit for running end-to-end benchmark studies (from hyperparameter optimization to evaluation) across an easily extensible set of tasks, deep learning models, datasets and evaluation metrics. LBT provides a configurable interface for controlling training and customizing evaluation, a standardized training framework for eliminating confounding variables, and support for multi-objective evaluation. We demonstrate how LBT can be used to create personalized benchmark studies with a large-scale comparative analysis for text classification across 7 models and 9 datasets. We explore the trade-offs between inference latency and performance, relationships between dataset attributes and performance, and the effects of pretraining on convergence and robustness, showing how LBT can be used to satisfy various benchmarking objectives.

Lei Shu, Hu Xu, Bing Liu et al.

Dialogue management (DM) plays a key role in the quality of the interaction with the user in a task-oriented dialogue system. In most existing approaches, the agent predicts only one DM policy action per turn. This significantly limits the expressive power of the conversational agent and introduces unwanted turns of interactions that may challenge users' patience. Longer conversations also lead to more errors and the system needs to be more robust to handle them. In this paper, we compare the performance of several models on the task of predicting multiple acts for each turn. A novel policy model is proposed based on a recurrent cell called gated Continue-Act-Slots (gCAS) that overcomes the limitations of the existing models. Experimental results show that gCAS outperforms other approaches. The code is available at https://leishu02.github.io/

Lei Shu, Piero Molino, Mahdi Namazifar et al.

This paper proposes a novel end-to-end architecture for task-oriented dialogue systems. It is based on a simple and practical yet very effective sequence-to-sequence approach, where language understanding and state tracking tasks are modeled jointly with a structured copy-augmented sequential decoder and a multi-label decoder for each slot. The policy engine and language generation tasks are modeled jointly following that. The copy-augmented sequential decoder deals with new or unknown values in the conversation, while the multi-label decoder combined with the sequential decoder ensures the explicit assignment of values to slots. On the generation part, slot binary classifiers are used to improve performance. This architecture is scalable to real-world scenarios and is shown through an empirical evaluation to achieve state-of-the-art performance on both the Cambridge Restaurant dataset and the Stanford in-car assistant dataset\footnote{The code is available at \url{https://github.com/uber-research/FSDM}}

Piero Molino, Christopher Ré

In the last years machine learning (ML) has moved from a academic endeavor to a pervasive technology adopted in almost every aspect of computing. ML-powered products are now embedded in our digital lives: from recommendations of what to watch, to divining our search intent, to powering virtual assistants in consumer and enterprise settings. Recent successes in applying ML in natural sciences revealed that ML can be used to tackle some of the hardest real-world problems humanity faces today. For these reasons ML has become central in the strategy of tech companies and has gathered even more attention from academia than ever before. Despite these successes, what we have witnessed so far is just the beginning. Right now the people training and using ML models are expert developers working within large organizations, but we believe the next wave of ML systems will allow a larger amount of people, potentially without coding skills, to perform the same tasks. These new ML systems will not require users to fully understand all the details of how models are trained and utilized for obtaining predictions. Declarative interfaces are well suited for this goal, by hiding complexity and favouring separation of interests, and can lead to increased productivity. We worked on such abstract interfaces by developing two declarative ML systems, Overton and Ludwig, that require users to declare only their data schema (names and types of inputs) and tasks rather then writing low level ML code. In this article we will describe how ML systems are currently structured, highlight important factors for their success and adoption, what are the issues current ML systems are facing and how the systems we developed addressed them. Finally we will talk about learnings from the development of ML systems throughout the years and how we believe the next generation of ML systems will look like.

Lei Shu, Alexandros Papangelis, Yi-Chia Wang et al.

This work introduces Focused-Variation Network (FVN), a novel model to control language generation. The main problems in previous controlled language generation models range from the difficulty of generating text according to the given attributes, to the lack of diversity of the generated texts. FVN addresses these issues by learning disjoint discrete latent spaces for each attribute inside codebooks, which allows for both controllability and diversity, while at the same time generating fluent text. We evaluate FVN on two text generation datasets with annotated content and style, and show state-of-the-art performance as assessed by automatic and human evaluations.

Yue Weng, Sai Sumanth Miryala, Chandra Khatri et al.

The quality of automatic speech recognition (ASR) is critical to Dialogue Systems as ASR errors propagate to and directly impact downstream tasks such as language understanding (LU). In this paper, we propose multi-task neural approaches to perform contextual language correction on ASR outputs jointly with LU to improve the performance of both tasks simultaneously. To measure the effectiveness of this approach we used a public benchmark, the 2nd Dialogue State Tracking (DSTC2) corpus. As a baseline approach, we trained task-specific Statistical Language Models (SLM) and fine-tuned state-of-the-art Generalized Pre-training (GPT) Language Model to re-rank the n-best ASR hypotheses, followed by a model to identify the dialog act and slots. i) We further trained ranker models using GPT and Hierarchical CNN-RNN models with discriminatory losses to detect the best output given n-best hypotheses. We extended these ranker models to first select the best ASR output and then identify the dialogue act and slots in an end to end fashion. ii) We also proposed a novel joint ASR error correction and LU model, a word confusion pointer network (WCN-Ptr) with multi-head self-attention on top, which consumes the word confusions populated from the n-best. We show that the error rates of off the shelf ASR and following LU systems can be reduced significantly by 14% relative with joint models trained using small amounts of in-domain data.

Andrea Madotto, Mahdi Namazifar, Joost Huizinga et al.

This work presents an exploration and imitation-learning-based agent capable of state-of-the-art performance in playing text-based computer games. Text-based computer games describe their world to the player through natural language and expect the player to interact with the game using text. These games are of interest as they can be seen as a testbed for language understanding, problem-solving, and language generation by artificial agents. Moreover, they provide a learning environment in which these skills can be acquired through interactions with an environment rather than using fixed corpora. One aspect that makes these games particularly challenging for learning agents is the combinatorially large action space. Existing methods for solving text-based games are limited to games that are either very simple or have an action space restricted to a predetermined set of admissible actions. In this work, we propose to use the exploration approach of Go-Explore for solving text-based games. More specifically, in an initial exploration phase, we first extract trajectories with high rewards, after which we train a policy to solve the game by imitating these trajectories. Our experiments show that this approach outperforms existing solutions in solving text-based games, and it is more sample efficient in terms of the number of interactions with the environment. Moreover, we show that the learned policy can generalize better than existing solutions to unseen games without using any restriction on the action space.

Alexandros Papangelis, Mahdi Namazifar, Chandra Khatri et al.

As the field of Spoken Dialogue Systems and Conversational AI grows, so does the need for tools and environments that abstract away implementation details in order to expedite the development process, lower the barrier of entry to the field, and offer a common test-bed for new ideas. In this paper, we present Plato, a flexible Conversational AI platform written in Python that supports any kind of conversational agent architecture, from standard architectures to architectures with jointly-trained components, single- or multi-party interactions, and offline or online training of any conversational agent component. Plato has been designed to be easy to understand and debug and is agnostic to the underlying learning frameworks that train each component.

Avishek Joey Bose, Ankit Jain, Piero Molino et al.

We consider the task of few shot link prediction on graphs. The goal is to learn from a distribution over graphs so that a model is able to quickly infer missing edges in a new graph after a small amount of training. We show that current link prediction methods are generally ill-equipped to handle this task. They cannot effectively transfer learned knowledge from one graph to another and are unable to effectively learn from sparse samples of edges. To address this challenge, we introduce a new gradient-based meta learning framework, Meta-Graph. Our framework leverages higher-order gradients along with a learned graph signature function that conditionally generates a graph neural network initialization. Using a novel set of few shot link prediction benchmarks, we show that Meta-Graph can learn to quickly adapt to a new graph using only a small sample of true edges, enabling not only fast adaptation but also improved results at convergence.

Sumanth Dathathri, Andrea Madotto, Janice Lan et al.

Large transformer-based language models (LMs) trained on huge text corpora have shown unparalleled generation capabilities. However, controlling attributes of the generated language (e.g. switching topic or sentiment) is difficult without modifying the model architecture or fine-tuning on attribute-specific data and entailing the significant cost of retraining. We propose a simple alternative: the Plug and Play Language Model (PPLM) for controllable language generation, which combines a pretrained LM with one or more simple attribute classifiers that guide text generation without any further training of the LM. In the canonical scenario we present, the attribute models are simple classifiers consisting of a user-specified bag of words or a single learned layer with 100,000 times fewer parameters than the LM. Sampling entails a forward and backward pass in which gradients from the attribute model push the LM's hidden activations and thus guide the generation. Model samples demonstrate control over a range of topics and sentiment styles, and extensive automated and human annotated evaluations show attribute alignment and fluency. PPLMs are flexible in that any combination of differentiable attribute models may be used to steer text generation, which will allow for diverse and creative applications beyond the examples given in this paper.

Piero Molino, Yaroslav Dudin, Sai Sumanth Miryala

In this work we present Ludwig, a flexible, extensible and easy to use toolbox which allows users to train deep learning models and use them for obtaining predictions without writing code. Ludwig implements a novel approach to deep learning model building based on two main abstractions: data types and declarative configuration files. The data type abstraction allows for easier code and sub-model reuse, and the standardized interfaces imposed by this abstraction allow for encapsulation and make the code easy to extend. Declarative model definition configuration files enable inexperienced users to obtain effective models and increase the productivity of expert users. Alongside these two innovations, Ludwig introduces a general modularized deep learning architecture called Encoder-Combiner-Decoder that can be instantiated to perform a vast amount of machine learning tasks. These innovations make it possible for engineers, scientists from other fields and, in general, a much broader audience to adopt deep learning models for their tasks, concretely helping in its democratization.

Alexandros Papangelis, Yi-Chia Wang, Piero Molino et al.

We present the first complete attempt at concurrently training conversational agents that communicate only via self-generated language. Using DSTC2 as seed data, we trained natural language understanding (NLU) and generation (NLG) networks for each agent and let the agents interact online. We model the interaction as a stochastic collaborative game where each agent (player) has a role ("assistant", "tourist", "eater", etc.) and their own objectives, and can only interact via natural language they generate. Each agent, therefore, needs to learn to operate optimally in an environment with multiple sources of uncertainty (its own NLU and NLG, the other agent's NLU, Policy, and NLG). In our evaluation, we show that the stochastic-game agents outperform deep learning based supervised baselines.

Piero Molino, Yang Wang, Jiawei Zhang

Embeddings are a fundamental component of many modern machine learning and natural language processing models. Understanding them and visualizing them is essential for gathering insights about the information they capture and the behavior of the models. State of the art in analyzing embeddings consists in projecting them in two-dimensional planes without any interpretable semantics associated to the axes of the projection, which makes detailed analyses and comparison among multiple sets of embeddings challenging. In this work, we propose to use explicit axes defined as algebraic formulae over embeddings to project them into a lower dimensional, but semantically meaningful subspace, as a simple yet effective analysis and visualization methodology. This methodology assigns an interpretable semantics to the measures of variability and the axes of visualizations, allowing for both comparisons among different sets of embeddings and fine-grained inspection of the embedding spaces. We demonstrate the power of the proposed methodology through a series of case studies that make use of visualizations constructed around the underlying methodology and through a user study. The results show how the methodology is effective at providing more profound insights than classical projection methods and how it is widely applicable to many other use cases.

Jiawei Zhang, Yang Wang, Piero Molino et al.

Interpretation and diagnosis of machine learning models have gained renewed interest in recent years with breakthroughs in new approaches. We present Manifold, a framework that utilizes visual analysis techniques to support interpretation, debugging, and comparison of machine learning models in a more transparent and interactive manner. Conventional techniques usually focus on visualizing the internal logic of a specific model type (i.e., deep neural networks), lacking the ability to extend to a more complex scenario where different model types are integrated. To this end, Manifold is designed as a generic framework that does not rely on or access the internal logic of the model and solely observes the input (i.e., instances or features) and the output (i.e., the predicted result and probability distribution). We describe the workflow of Manifold as an iterative process consisting of three major phases that are commonly involved in the model development and diagnosis process: inspection (hypothesis), explanation (reasoning), and refinement (verification). The visual components supporting these tasks include a scatterplot-based visual summary that overviews the models' outcome and a customizable tabular view that reveals feature discrimination. We demonstrate current applications of the framework on the classification and regression tasks and discuss other potential machine learning use scenarios where Manifold can be applied.

Rosanne Liu, Joel Lehman, Piero Molino et al.

Few ideas have enjoyed as large an impact on deep learning as convolution. For any problem involving pixels or spatial representations, common intuition holds that convolutional neural networks may be appropriate. In this paper we show a striking counterexample to this intuition via the seemingly trivial coordinate transform problem, which simply requires learning a mapping between coordinates in (x,y) Cartesian space and one-hot pixel space. Although convolutional networks would seem appropriate for this task, we show that they fail spectacularly. We demonstrate and carefully analyze the failure first on a toy problem, at which point a simple fix becomes obvious. We call this solution CoordConv, which works by giving convolution access to its own input coordinates through the use of extra coordinate channels. Without sacrificing the computational and parametric efficiency of ordinary convolution, CoordConv allows networks to learn either complete translation invariance or varying degrees of translation dependence, as required by the end task. CoordConv solves the coordinate transform problem with perfect generalization and 150 times faster with 10--100 times fewer parameters than convolution. This stark contrast raises the question: to what extent has this inability of convolution persisted insidiously inside other tasks, subtly hampering performance from within? A complete answer to this question will require further investigation, but we show preliminary evidence that swapping convolution for CoordConv can improve models on a diverse set of tasks. Using CoordConv in a GAN produced less mode collapse as the transform between high-level spatial latents and pixels becomes easier to learn. A Faster R-CNN detection model trained on MNIST showed 24% better IOU when using CoordConv, and in the RL domain agents playing Atari games benefit significantly from the use of CoordConv layers.

Piero Molino, Huaixiu Zheng, Yi-Chia Wang

For a company looking to provide delightful user experiences, it is of paramount importance to take care of any customer issues. This paper proposes COTA, a system to improve speed and reliability of customer support for end users through automated ticket classification and answers selection for support representatives. Two machine learning and natural language processing techniques are demonstrated: one relying on feature engineering (COTA v1) and the other exploiting raw signals through deep learning architectures (COTA v2). COTA v1 employs a new approach that converts the multi-classification task into a ranking problem, demonstrating significantly better performance in the case of thousands of classes. For COTA v2, we propose an Encoder-Combiner-Decoder, a novel deep learning architecture that allows for heterogeneous input and output feature types and injection of prior knowledge through network architecture choices. This paper compares these models and their variants on the task of ticket classification and answer selection, showing model COTA v2 outperforms COTA v1, and analyzes their inner workings and shortcomings. Finally, an A/B test is conducted in a production setting validating the real-world impact of COTA in reducing issue resolution time by 10 percent without reducing customer satisfaction.