[mintlify]

Summary

Documents the StatsForecast constructor defaults and explains how NeuralForecast max_steps relates to epochs shown in training logs.

Changes

• statsforecast/src/core/core.html.mdx: add constructor signature and parameter table (n_jobs defaults to 1), verified against statsforecast/core.py.
• StatsForecast end-to-end guides (pandas and polars): note that n_jobs defaults to 1.
• NeuralForecast cross-validation page: explain that max_steps counts gradient steps, how steps map to epochs, and why the progress bar ends at epoch max_steps - 1 for small datasets.

Context

Recurring unanswered assistant questions asked for the default n_jobs value and why training showed 999 epochs with max_steps=1000 during cross-validation.

[mintlify]

Preview deployment for your docs. Learn more about Mintlify Previews.

Project	Status	Preview	Updated
nixtlaverse_nixtla_io	🟢 Ready	View Preview	Oct 19, 2026, 4:11 PM

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}