1 code implementation • ICLR 2020 • Massimo Caccia, Lucas Caccia, William Fedus, Hugo Larochelle, Joelle Pineau, Laurent Charlin
Generating high-quality text with sufficient diversity is essential for a wide range of Natural Language Generation (NLG) tasks.
1 code implementation • 4 Dec 2018 • Lucas Caccia, Herke van Hoof, Aaron Courville, Joelle Pineau
In this work, we show that one can adapt deep generative models for this task by unravelling lidar scans into a 2D point map.
no code implementations • 23 May 2019 • Pierre Thodoroff, Nishanth Anand, Lucas Caccia, Doina Precup, Joelle Pineau
Despite recent successes in Reinforcement Learning, value-based methods often suffer from high variance hindering performance.
1 code implementation • 11 Aug 2019 • Rahaf Aljundi, Lucas Caccia, Eugene Belilovsky, Massimo Caccia, Min Lin, Laurent Charlin, Tinne Tuytelaars
Methods based on replay, either generative or from a stored memory, have been shown to be effective approaches for continual learning, matching or exceeding the state of the art in a number of standard benchmarks.
no code implementations • 25 Sep 2019 • Lucas Caccia, Eugene Belilovsky, Massimo Caccia, Joelle Pineau
We first replace the episodic memory used in Experience Replay with SQM, leading to significant gains on standard continual learning benchmarks using a fixed memory budget.
1 code implementation • ICML 2020 • Lucas Caccia, Eugene Belilovsky, Massimo Caccia, Joelle Pineau
We show how to use discrete auto-encoders to effectively address this challenge and introduce Adaptive Quantization Modules (AQM) to control variation in the compression ability of the module at any given stage of learning.
1 code implementation • NeurIPS 2020 • Massimo Caccia, Pau Rodriguez, Oleksiy Ostapenko, Fabrice Normandin, Min Lin, Lucas Caccia, Issam Laradji, Irina Rish, Alexandre Lacoste, David Vazquez, Laurent Charlin
We propose Continual-MAML, an online extension of the popular MAML algorithm as a strong baseline for this scenario.
3 code implementations • 11 Apr 2021 • Lucas Caccia, Rahaf Aljundi, Nader Asadi, Tinne Tuytelaars, Joelle Pineau, Eugene Belilovsky
In this work, we focus on the change in representations of observed data that arises when previously unobserved classes appear in the incoming data stream, and new classes must be distinguished from previous ones.
no code implementations • 11 Jun 2021 • Eugene Belilovsky, Louis Leconte, Lucas Caccia, Michael Eickenberg, Edouard Oyallon
With the use of a replay buffer we show that this approach can be extended to asynchronous settings, where modules can operate and continue to update with possibly large communication delays.
no code implementations • 16 Jun 2021 • Lucas Caccia, Joelle Pineau
This paper presents SPeCiaL: a method for unsupervised pretraining of representations tailored for continual learning.
1 code implementation • 17 Jun 2021 • Lucas Caccia, Jing Xu, Myle Ott, Marc'Aurelio Ranzato, Ludovic Denoyer
Practitioners have then to decide how to allocate their computational budget in order to obtain the best performance at any point in time.
3 code implementations • ICLR 2022 • Lucas Caccia, Rahaf Aljundi, Nader Asadi, Tinne Tuytelaars, Joelle Pineau, Eugene Belilovsky
In this work, we focus on the change in representations of observed data that arises when previously unobserved classes appear in the incoming data stream, and new classes must be distinguished from previous ones.
1 code implementation • NeurIPS 2023 • Lucas Caccia, Edoardo Ponti, Zhan Su, Matheus Pereira, Nicolas Le Roux, Alessandro Sordoni
We find that routing is most beneficial during multi-task pre-training rather than during few-shot adaptation and propose $\texttt{MHR}$-$\mu$, which discards routing and fine-tunes the average of the pre-trained adapters on each downstream tasks.
1 code implementation • 18 Nov 2022 • Jean-Baptiste Gaya, Thang Doan, Lucas Caccia, Laure Soulier, Ludovic Denoyer, Roberta Raileanu
We introduce Continual Subspace of Policies (CSP), a new approach that incrementally builds a subspace of policies for training a reinforcement learning agent on a sequence of tasks.
no code implementations • 11 Apr 2023 • Gwen Legate, Lucas Caccia, Eugene Belilovsky
In Federated Learning, a global model is learned by aggregating model updates computed at a set of independent client nodes, to reduce communication costs multiple gradient steps are performed at each node prior to aggregation.
no code implementations • 9 Oct 2023 • Xinyi Wang, Lucas Caccia, Oleksiy Ostapenko, Xingdi Yuan, William Yang Wang, Alessandro Sordoni
Large language models (LLMs) have recently attracted considerable interest for their ability to perform complex reasoning tasks, such as chain-of-thought reasoning.