After the training, how can I run a single prediction for anomaly detection, without any evaluation? I don't quite understand what type of post-processing I need to do for the log prob output to just achieve AD predictions when I don't have GT data
Would you mind posting the command to replicate the results in the table for ContextFlow? The command:
python model.py --gpu 0 --dataset smap --coupling trans --action-type train-generalist
Appears to be the only one in the command list with dataset as smap. But that seems to be the generalist model both from the "train-generalist" and from the results I get when I run it compared to the results from the table.