Discussions

I would like to know how can we use your services in cases in which data are confidential and are not allowed to leave our servers? Is there a way to run the model locally? How does your Azure implementation work in that respect? Thank you.

Qu'est-ce que je peux attendre du jaf et du je le 01 et 04 juillrt a

Qu'est-ce que je peux attendre du jaf et du je le 01 et 04 juillrt a

What is the maximum number of tokens (context length) TimeGPT can handle for a single inference pass?

What is the future of ai in 5 years?

When will the model be available in Microsoft Azure? Do you have a time frame?

When will the model be available in Microsoft Azure? Do you have a time frame?

Is there a plan to also offer classification?

I was wondering, if you use TimeGPT for forecasting, and you incorporate an exogenous variable "Ex", what does the model use exactly to predict the target time series at a given timestamp t? Does it use the value of "Ex" at the same timestamp t, or does it only use past values (so t-1 and previous values)?

the error:
INFO:nixtla.nixtla_client:Validating inputs...
INFO:nixtla.nixtla_client:Preprocessing dataframes...
error: cannot handle a non-unique multi-index!