BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Palaner//palaner Single Event//EN
METHOD:PUBLISH
CALSCALE:GREGORIAN
BEGIN:VEVENT
UID:luma-1f6cc23402@palaner.app
DTSTAMP:20260802T140653Z
DTSTART:20260829T100000Z
DTEND:20260829T113000Z
SUMMARY:Daytona & Give(a)Go HackSprint - Dublin\, August 2026
LOCATION:Baseline
DESCRIPTION:Saved in Palaner as: saved\n\nPalaner match: 25% · why: matches
  your usual pattern\n\nA one-day HackSprint for fine-tuning\, reinforcemen
 t learning and AI sandboxes\nOn 29 August\, Daytona\, Give(a)Go and Baseli
 ne are bringing Dublin builders together for a full day of experiments at 
 the edge of model training and agent infrastructure.\nThe challenge is sim
 ple: take a model or agent\, give it a measurable task\, and make it impro
 ve.\nYou might fine-tune a small model on a focused dataset\, build an eva
 luation loop\, train an agent with reinforcement learning\, or use isolate
 d sandboxes to run many experiments safely in parallel. The day is about w
 orking systems\, visible results and lessons earned by building.\nWhat you
  will explore\nFine-tuning: prepare datasets\, run training jobs\, compare
  checkpoints and understand where specialised models beat prompting alone.
 \nReinforcement learning: design rewards\, run rollouts\, evaluate behavio
 ur and improve an agent through feedback.\nSandboxes: give every experimen
 t an isolated\, reproducible computer with the dependencies\, files and to
 ols it needs.\nParallel experimentation: fork environments\, test multiple
  approaches at once and keep the strongest results.\nSandboxes make this p
 ractical. Each training job\, rollout or agent variant can run inside its 
 own disposable environment\, without contaminating the others. Teams can r
 eproduce failures\, compare results and scale experiments without spending
  the day rebuilding infrastructure.\nEvery builder will receive Daytona cr
 edits to build with during the HackSprint.\nProject directions\nEvolutiona
 ry fine-tuning lab: run fine-tuning jobs across datasets and hyperparamete
 rs\, score them automatically\, then promote the strongest checkpoints.\nR
 L agent arena: create a task environment where agents attempt the same goa
 l\, receive rewards and improve over repeated rollouts.\nSelf-improving co
 ding agent: benchmark agent variants against real repository issues and te
 sts\, then use the results to refine prompts\, tools\, memory or models.\n
 Sandboxed evaluation harness: spin up a clean environment for every model 
 or agent evaluation so each run is reproducible and safely isolated.\nSynt
 hetic data factory: use agents to generate\, critique and filter training 
 examples\, then measure whether the resulting dataset improves a small mod
 el.\nYou can arrive with a team or join one on the day. Bring a laptop\, a
  model or agent idea\, and a problem you can measure.\nWhat the day looks 
 like\n11:00 AM: Doors open\n11:30 AM: Introduction to Daytona\, sandboxes 
 and the challenge\n12:00 PM: Hacking begins\n1:00 PM: Lunch\n1:30 PM: Back
  to building\n4:45 PM: Demos and judging\n5:30 PM: Awards and close\nAbout
  Daytona\nDaytona builds secure\, elastic infrastructure for running AI-ge
 nerated code. Its sandboxes launch quickly\, fork into parallel branches a
 nd snapshot mid-execution\, giving agents and training workflows isolated 
 computers they can create and discard on demand.\nVenue partner\nBaseline 
 is an independent community and fund for early-stage founders in Ireland. 
 They back technical founders with €100K first checks and create a home for
  ambitious builders in Dublin.\nJoin the community\nGive(a)Go is Ireland’s
  largest community of builders and founders\, with 2\,500+ attendees acros
 s 45+ hands-on events. From AI filmmaking and agentic hackathons to builde
 r days\, we bring ambitious people together to turn new technology into wo
 rking projects.\nFollow us on LinkedIn\, Instagram\, our website and GitHu
 b.\nWe will take photos and short video clips during the day for recap pos
 ts and marketing. By registering\, you consent to being included. If you p
 refer not to appear\, tell one of the hosts when you arrive.
URL:https://luma.com/daytona-dublin
CATEGORIES:palaner,saved
STATUS:TENTATIVE
END:VEVENT
END:VCALENDAR