refresh card (as_of=2026-07-29)

README.md

@@ -33,11 +33,11 @@ values + a model's quantiles, and the KOTH cluster-bootstrap groups windows by
 `source`. Re-cut windows of `context_length + horizon`, forecast the horizon,
 and the metrics reproduce byte-for-byte.
 
-## Latest snapshot — `2026-07-28`
+## Latest snapshot — `2026-07-29`
 
-- **series:** 3285
-- **domains:** transport=2365, energy=348, healthcare=223, nature=188, sales=66, econ_fin=65, web_cloudops=30
-- **cadences:** min=2381, D=337, 30min=235, 5min=124, 15min=123, H=50, 10min=15, 6min=11, 240S=6, 15S=1, 10S=1, 1S=1
+- **series:** 3318
+- **domains:** transport=2385, energy=353, healthcare=224, nature=192, econ_fin=68, sales=66, web_cloudops=30
+- **cadences:** min=2401, D=338, 30min=236, 15min=130, 5min=128, H=51, 10min=15, 6min=10, 240S=6, 15S=1, 10S=1, 1S=1
 
 > This is an **evaluation** set, not training data. Publishing it to train on
 > would contaminate the benchmark it exists to measure.