Yhn ScienceBiology first seen 14 h ago, last 49 min ago, peak #8

T0: compact open-weights time-series forecasting model released

Original: Show HN: T0: We trained a 256M SOTA open-weights time-series forecasting model

A small team has released T0, a 256-million-parameter time-series forecasting model it says achieves state-of-the-art results while being small enough to run cheaply, with weights published openly on GitHub. Developer communities are discussing the claim, weighing the unusually compact size against the state-of-the-art benchmark performance and what an open release means for forecasting work.

Why now: A claimed state-of-the-art open-weights release challenges the assumption that better forecasting requires much larger models, prompting technical debate.

T0The Forecasting CompanyGitHub

Open on hn →

Rank over time, top of the chart is #1. 13 snapshots from 11 h ago to 49 min ago.

Evidence

• Show HN: T0: We trained a 256M SOTA open-weights time-series forecasting model · LTMeyer · 12

API: https://socialmediatrends-api.osmike.com/v1/trends/2248762