BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//programme.europython.eu//europython-2024//speaker//ZKDDF
 P
BEGIN:VTIMEZONE
TZID:CET
BEGIN:STANDARD
DTSTART:20001029T040000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000326T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-europython-2024-KDH3J3@programme.europython.eu
DTSTART;TZID=CET:20240712T140000
DTEND;TZID=CET:20240712T143000
DESCRIPTION:Every second spent waiting for initializations and obscure dela
 ys hindering high-frequency logging\, further limited by what you can trac
 k\, an experiment dies. Wouldn’t loading and starting tracking in nearly
  zero time be nice? What if we could track more and faster\, even handling
  arbitrarily large\, complex Python objects with ease?\n\nIn this talk\, I
  will present the results of comparative benchmarks covering Weights & Bia
 ses\, MLflow\, FastTrackML\, Neptune\, Aim\, Comet\, and MLtraq. You will 
 learn their strengths and weaknesses\, what makes them slow and fast\, and
  what sets MLtraq apart\, making it 100x faster and capable of handling te
 ns of thousands of experiments.\n\nThis presentation will not only be enli
 ghtening for those involved in AI/ML experimentation but will also be inva
 luable for anyone interested in the efficient and safe serialization of Py
 thon objects.
DTSTAMP:20260520T133949Z
LOCATION:Terrace 2A
SUMMARY:MLtraq: Track your ML/AI experiments at hyperspeed - Michele Dallac
 hiesa
URL:https://programme.europython.eu/europython-2024/talk/KDH3J3/
END:VEVENT
END:VCALENDAR