search

zhihanyue / ts2vec

A universal time series representation learning framework
MIT License
593 stars 143 forks source link

Unexpected evaluation results with zeroed Loss (anomaly detection) #40

Open Alegelx24 opened 10 months ago

Alegelx24 commented 10 months ago

Hello, I have been studying the ts2vec model recently. When I use the yahoo dataset for training, I encountered an inconsistency about the contribution of the loss. Even if i set the loss manually to zero during the training phase, the evaluation results are almost the same, sometimes even better. I was wondering if it is something that I am missing or if it is a possible issue encountered, perhaps related to the evaluation phase itself. I appreciate any guidance or assistance you can provide regarding this matter and I hope to receive your answer. Thank you for your time and attention to this issue!

HuntZhaozq commented 10 months ago

@Alegelx24 Hello, I wonder how to download the Yahoo dataset without register in the website.

Alegelx24 commented 10 months ago

@HuntZhaozq I think it is unavailable elsewhere