BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.com//pyconde-pydata-2026//talk//AZ7GD3
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-pyconde-pydata-2026-AZ7GD3@pretalx.com
DTSTART;TZID=CET:20260415T142000
DTEND;TZID=CET:20260415T145000
DESCRIPTION:Large language models have been widely used in tool-calling wor
 kflows thanks to their strong performance in generating appropriate functi
 on calls. However\, due to their size and cost\, they are inaccessible to 
 small-scale builders\, and server-side computing makes data privacy challe
 nging. Small language models (SLMs) are a promising\, affordable alternati
 ve that can run on local hardware\, ensuring higher privacy.\n\nUnfortunat
 ely\, SLMs struggle with this task - they pass wrong arguments when callin
 g functions with many parameters\, and make mistakes when the conversation
  spans multiple turns. On the other hand\, for production applications wit
 h specific API sets\, we often don't need general-purpose LLMs - we need r
 eliable\, specialized models.\n\nThis talk demonstrates how to increase th
 e accuracy of SLMs (under 8B parameters) for custom tool calling tasks. We
  will share how leveraging knowledge distillation helps to get the most ou
 t of SLMs in low-data settings - they can even outperform LLMs! We will pr
 esent the whole pipeline from data generation\, fine-tuning\, and local de
 ployment.
DTSTAMP:20260523T180009Z
LOCATION:Helium [3rd Floor]
SUMMARY:Small Language Models for Tool Calling Are Better Than You Think - 
 Gabi Kadlecova
URL:https://pretalx.com/pyconde-pydata-2026/talk/AZ7GD3/
END:VEVENT
END:VCALENDAR