Nixtla's Open Source Time Series Ecosystem.

Nixtla Community

image.png

Hi <@U03619XRWKD>,
happy to hear and thank you for providing this new functionality! You said "currently only available in the main branch of the repo and will be released to the pip version early next year". Can you provide further information on how I can access the function? Is it already implemented in pip? What do you mean by the 'main branch of the repo'?

Resolved it :slightly_smiling_face: I post here the solution to share with everyone:
The following lines of code allow you to install the latest repo as of the latest version on GitHub.

Exemplary for environment in Google Colab:
`## Install directly from githup repo with latest functions etc`
`!pip install git+<https://github.com/Nixtla/neuralforecast.git>`

p.s.: @ anyone: please correct me if there is anything wrong.

Thank you, <@U03619XRWKD>! Is there a way to manually use the warm-start with the pip version in an "online" manner? Whenever I get new data, can I fit starting from the previous state and run much fewer iterations?