Alexandre Allauzen

[Alexandre Allauzen @AlexAllauzen]

Tuesday October First, we are happy to have a talk by Antoine Gonon (ENS-Lyon) on Symmetries in Neural Networks.

Where: @PariSanteCampus room 0.3, ground floor.

[Physics In History @PhysInHistory]

Physics is always fun to learn.

[Gabriel Peyré @gabrielpeyre]

Classification and Regression Trees (CART) define structured recursive classification and regression functions. O(n*log(n)) time global optimization (despite the exponential number of models) is achieved by dynamic programming. https://t.co/uAqRUe2HVm

[Alexandre Allauzen @AlexAllauzen]

Bravo Erwan !!

[MILESDauphine @DauphineMiles]

Comming soon on Arxiv 🪩
our Paper accepted at EMNLP 🌴🌴
On "Chain and Causal Attention for Efficient Entity Tracking", aka Chacal !
Congratz to Erwan Fagnou and all his co-authors 🍾🍾

[Alexandre Allauzen @AlexAllauzen]

@laurent_besacie Je ne savais pas que tu travaillais à Orsay😀

[laurent besacier @laurent_besacie]

Wow our NeurIPS2021 paper for the track 'datasets & benchmarks' was accepted🍾. We propose a multi-task benchmark to evaluate large speech foundation models based on self-supervised learning. Our benchmark is targeting French language for the moment - see https://t.co/VssezSXPUv

[Alexandre Allauzen @AlexAllauzen]

La Maire de Paris: Sauvons l'Institut de Phonétique ! - Signez la pétition ! https://t.co/48ATbfLK8e via @ChangeFrance

[timorous bestie 😷 @vnfrombucharest]

⚠I am recruiting for a PhD position in ML for NLP⚠

Apply by Nov 19 here: https://t.co/en8idArRa9

Extra context on my website: https://t.co/eeTF9UYk01. DM or email if you have any questions.

[Mathieu Blondel @mblondel_ml]

@fchollet At this rate, the Paris-Saclay metro will fall in that category :)

[Alexandre Allauzen @AlexAllauzen]

@hbredin @Wirewalkeuse @SAMOVAIRIT Bonne installation à Toulouse ! à bientôt.

[CNRS Sciences informatiques @CNRSinformatics]

#FlauBERT est un nouveau modèle de langue en 🇲🇫 appris avec le
supercalculateur Jean Zay du @CNRS #IA #deeplearning ! Il permet une recherche contextualisée.
➡️https://t.co/RctGmLfkWO
➡️https://t.co/0MABxC4cQc
🤝 @laurent_besacie @didier_schwab et @AlexAllauze @Genci_fr

[didier_schwab @didier_schwab]

Aujourd'hui, 198ème anniversaire de Gustave Flaubert. En hommage, nous publions FlauBERT, un modèle de langue pré-entraîné pour le français grâce à @Genci_fr (Travail commun @UGrenobleAlpes, @cnrs, @LIGLab -- @ParisDiderot -- @ESPCI_Paris @cnrs @psl_univ)
https://t.co/vaETrOQNNQ

[laurent besacier @laurent_besacie]

Here is FlauBERT: a French LM learnt (with #CNRS J-Zay supercomputer) on a large and heterogeneous corpus. Along with it comes FLUE (evaluation setup for French NLP). FlauBERT was successfully applied to complex tasks (NLI, WSD, Parsing). More on https://t.co/8e6F6UEdM9