BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.com//euroscipy-2026//speaker//88DYMB
BEGIN:VTIMEZONE
TZID:Poland
BEGIN:DAYLIGHT
DTSTART:20250722T000000
TZNAME:CEST
TZOFFSETFROM:+0200
TZOFFSETTO:+0200
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20251026T030000
RDATE:20261025T030000
TZNAME:CET
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20260329T030000
RDATE:20270328T030000
TZNAME:CEST
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
SUMMARY:Same Recipe\, Different Results: Fine-Tuning Models Across Modalit
 ies - Ramon Perez
DTSTART;TZID=Poland:20260722T160000
DTEND;TZID=Poland:20260722T173000
DTSTAMP:20260909T063427Z
UID:pretalx-euroscipy-2026-Y7YM3G@pretalx.com
DESCRIPTION:The intuitions you build fine-tuning text models are surprisin
 gly bad guides for other modalities. Training configurations that work wel
 l for language will silently degrade an image model. Dataset sizes that fe
 el tiny for text are more than enough for adapting a visual style. And aud
 io\, despite seeming like its own world\, follows an image pipeline once y
 ou transform sound into spectrograms\, making what counts as a "token" str
 anger and more interesting than most people expect. The modalities share a
  vocabulary (fine-tuning\, adapters\, checkpoints) but not a playbook\, an
 d the gaps between them are where the most useful lessons live.\n\nThis ta
 lk is a practical\, comparative tour of fine-tuning across four modalities
 : text\, images\, audio\, and video. Rather than focusing on one\, we will
  look at what changes as you move between them\, how you prepare different
  data\, which training strategies transfer and which don't\, where the got
 chas hide\, and what model merging can do for you once training is done. A
 ll examples use Python and the HuggingFace ecosystem with publicly availab
 le models and datasets. Whether you are a practitioner looking to branch o
 ut beyond NLP or someone curious about what multi-modal fine-tuning looks 
 like in practice\, you will leave with a mental map of the landscape and e
 nough pointers to start exploring on your own.
LOCATION:Room 1.19 (Ground Floor\, Shannon)
URL:https://pretalx.com/euroscipy-2026/talk/Y7YM3G/
END:VEVENT
END:VCALENDAR