We propose adding the Amazon Chronos-2 pipeline into Orion as a pre-trained model. This would introduce a new primitive wrapping Chronos2Pipeline. The primitive will integrate into Orion’s collection of forecasting-based anomaly detection methods. The primitive will support multivariate inputs and univariate outputs for now, with a later PR allowing for multivariate outputs as well.
We would also introduce a new pipeline which uses the primitive as its prediction step. We can use this pipeline to perform anomaly detection and include it in our benchmark suite.
We propose adding the Amazon Chronos-2 pipeline into Orion as a pre-trained model. This would introduce a new primitive wrapping Chronos2Pipeline. The primitive will integrate into Orion’s collection of forecasting-based anomaly detection methods. The primitive will support multivariate inputs and univariate outputs for now, with a later PR allowing for multivariate outputs as well.
We would also introduce a new pipeline which uses the primitive as its prediction step. We can use this pipeline to perform anomaly detection and include it in our benchmark suite.