News

Our paper "Accelerating Transformer Inference for Translation via Parallel Decoding" was accepted to **ACL 2023** main conference!
Our paper “Accelerating Transformer Inference for Translation via Parallel Decoding” was accepted to ACL 2023 main conference!
Last updated on May 18, 2023
We release Camoscio!
We release Camoscio: An Italian instruction-tuned LLaMA! Checkout the details here (Italian) or here (English).
Last updated on Apr 4, 2023
Our paper "Relative representations enable zero-shot latent space communication" was accepted to ICLR 2023 as an Oral (Notable Top 5%) presentation!
Our paper “Relative representations enable zero-shot latent space communication” was accepted to ICLR 2023 as an Oral (Notable Top 5%) presentation!
Last updated on Feb 10, 2023
The new website is online!
The new website is online!
Last updated on Feb 10, 2023