Benchmarks: Kumo, Nori, Chronos-2 vs tuned LightGBM for power prices
nkafr · reddit · 2026-11-07
The author benchmarked tabular foundation models (Kumo, Nori), time-series foundation models (Chronos-2, TimesFM-3), and tuned LightGBM/CatBoost on German day-ahead electricity price forecasting, using the foundation models zero-shot on electricity market data with comparable information across models. The writeup covers experimental setup, using tabular foundation models for forecasting without data leakage, comparison against tuned tree-based baselines, and how extra training data shifts the traditional ML results. Full tutorial on Substack.
More from Models
- $50K live trading battles: AI bots currently edge out Claude and GPT — Mr_Hodlerr · 2026-11-07
- Brazil's Maritaca Releases S5-nano 4B, Claiming Best Open Weights in Brazil — ApprehensiveAd3629 · 2026-11-07
- Grok Bot's computer internet access issue is fixed — nima_owji · 2026-11-07
- Consistency test: Kimi K3.1 Pro passes 45% of VIALS tasks on all 3 tries vs 63% for GPT-6 Astra — ArtificialAnlys · 2026-11-07
- Consistency separates frontier models: Kimi K3.1 Pro passes 45% of VIALS tasks 3/3 vs Astra's 63% — ArtificialAnlys · 2026-11-07
- New VIALS-AA leaderboard: Gemini 4 leads at 64%, Claude costs 29x more per task — ArtificialAnlys · 2026-11-07