TimeGPT-1: production ready pre-trained Time Series Foundation Model for forecasting and anomaly detection. Generative pretrained transformer for time series trained on over 100B data points. It's capable of accurately predicting various domains such as retail, electricity, finance, and IoT with just a few lines of code 🚀.
Hi, I am checking to see if objective function can enter as parameter in any of the models (in neuralforecast I could see, but not in other ones. )
In this example, there is objective parameter (objective = 'tweedie') for the tsforecast()
source
However couldn't find it in the actual implementation.
link
Just wondering ways to scale different models and weather one can specify different loss functions at this level ..