Re:TimeGPT-1
by dist-epoch on Friday October 13, 2023 @05:22PM
(#37875782)
ARIMA is not very good for anything but highly predictable time series (of the summer-hot winter-cold kind).
[ Parent
| on HN ]
Ok but the claim is about high training times… Wtf?
High training times could be cost prohibitive. Currently, its over $100mil to train GPT4 from scratch (which possibly includes other costs related to RLHF and data acquisition). Not sure how this model compares, but its likely not cheap.
Re:TimeGPT-1
by warkdarrior on Friday October 13, 2023 @05:59PM
(#37876082)
The claim in the paper was the ARIMA has high training times.
Maybe they're using a pure Python implementation or something...
Re:TimeGPT-1
by carbocation on Friday October 13, 2023 @05:55PM
(#37876059)
If true, then beating it and looking good will be easy.
Having trained ARIMA models in my day, I will say that long training times and training cost -- compared to any deep learning model -- is not something that ever crossed my mind.