BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Fast ML UK tutorial
DTSTART:20260327T093000Z
DTEND:20260327T163000Z
DTSTAMP:20260414T133600Z
UID:indico-event-1641573@indico.cern.ch
DESCRIPTION:Speakers: Lauri Antti Olavi Laatu (Imperial College (GB))\, Jo
 nathon Mark Langford (Imperial College (GB))\n\nWe are hosting a one-day F
 ast Machine Learning (Fast ML) tutorial at Imperial College London on 27th
  March\, aimed at UK students and early-career researchers (ECRs) interest
 ed in ML for real-time scientific applications. In this tutorial\, partici
 pants will learn how to convert standard ML algorithms for FPGA synthesis 
 using the hls4ml library\, and explore the trade-offs between model accura
 cy and FPGA resource usage/latency. We will then introduce quantization-aw
 are training techniques that help maintain model performance while minimiz
 ing resource consumption.\nIn the afternoon session\, we will explore more
  complex architectures (e.g. transformers) and demonstrate the practical s
 teps required for FPGA synthesis with the aim to showcase the build proces
 s on a real FPGA device. Participation is in-person only\, and registratio
 n closes on 19th March so that we can finalise numbers for refreshments.\n
 \nhttps://indico.cern.ch/event/1641573/
LOCATION:LT2 (Imperial)
URL:https://indico.cern.ch/event/1641573/
END:VEVENT
END:VCALENDAR