Sergio P. Perez

Joined

May 2022

Names

Sergio P. Perez (Preferred)

Emails

****@graphcore.ai (Confirmed)

,

****@imperial.ac.uk (Confirmed)

,

****@gmail.com (Confirmed)

Personal Links

Career & Education History

Researcher

Graphcore (graphcore.ai)

2021 – 2023

PhD student

Imperial College London, Imperial College London (imperial.ac.uk)

2017 – 2021

Advisors, Relations & Conflicts

No relations added

Expertise

Natural Language Processing

2021 – 2023

Quantization

2021 – 2023

Machine Learning Hardware

2021 – 2023

Computational Fluid Dynamics

2017 – 2021

Scientific Computing

2017 – 2021

Publications

Enhancement of damaged-image prediction through Cahn–Hilliard image inpainting
Sergio P. Perez
- 16 Oct 2023
- OpenReview Archive Direct Upload
- Readers: Everyone
Training and inference of large language models using 8-bit floating point
Sergio P. Perez, Yan Zhang, James Briggs, Charlie Blake, Josh Levy-Kramer, Paul Balanca, Carlo Luschi, Stephen Barlow, Andrew W Fitzgibbon
- WANT@NeurIPS 2023 Oral
- Readers: Everyone
Training and inference of large language models using 8-bit floating point
Sergio P. Perez, Yan Zhang, James Briggs, Charlie Blake, Josh Levy-Kramer, Paul Balanca, Carlo Luschi, Stephen Barlow, Andrew W Fitzgibbon
- Submitted to ICLR 2024
- Readers: Everyone
Efficient Sequence Packing without Cross-contamination: Accelerating Large Language Models without Impacting Performance
Mario Michael Krell, Matej Kosec, Sergio P. Perez, Andrew W Fitzgibbon
- Published: 01 Feb 2023, Last Modified: 27 Jun 2024
- Submitted to ICLR 2023
- Readers: Everyone
Efficient Sequence Packing without Cross-contamination: Accelerating Large Language Models without Impacting Performance
Mario Michael Krell, Matej Kosec, Sergio P. Perez, Mrinal Iyer, Andrew W Fitzgibbon
- 16 May 2022 (modified: 27 Jun 2024)
- NeurIPS 2022 Submitted
- Readers: Everyone

Co-Authors