🛠️ T0: We trained a 256M SOTA open-weights time-series forecasting model
Summary
T0 is a 256M parameter open-weights foundation model for zero-shot time-series forecasting. It utilizes historical data and future-known variables to predict outcomes across various domains and frequencies.
Why it’s interesting
It currently ranks first among other time-series foundation models on public benchmarks like GIFT-Eval, fev-bench, and TIME.
Source metrics: Points 8 · Comments 0
HN discussion · Project
Source: #HackerNews / Show HN
Summary
T0 is a 256M parameter open-weights foundation model for zero-shot time-series forecasting. It utilizes historical data and future-known variables to predict outcomes across various domains and frequencies.
Why it’s interesting
It currently ranks first among other time-series foundation models on public benchmarks like GIFT-Eval, fev-bench, and TIME.
Source metrics: Points 8 · Comments 0
HN discussion · Project
Source: #HackerNews / Show HN