The race to construct the Top foundation forecasting model is on!
Salesforce’s MOIRAI, one in all the early foundation models, achieved high benchmark results and was open-sourced together with its pretraining dataset, LOTSA.
We extensively analyzed how MOIRAI works here — and built an end-to-end project comparing MOIRAI with popular statistical models.
Salesforce has now released an upgraded version — MOIRAI-MOE — with significant improvements, particularly the addition of Mixture-of-Experts (MOE). We briefly discussed MOE when one other model, Time-MOE, also used multiple experts.
In this text, we’ll cover:
- How MOIRAI-MOE works and why it’s a strong model.
- Key differences between MOIRAI and MOIRAI-MOE.
- How MOIRAI-MOE’s use of Mixture-of-Experts enhances accuracy.
- How Mixture-of-Experts generally solves frequency variation issues in foundation time-series models.
Let’s start.
✅ I’ve launched AI Horizon Forecast, a newsletter specializing in time-series and revolutionary…
