I tested to run EQTransformer (conservative model) on one-day data multiple times.
I would get different results when the input and parameters were not changed. By changing the threshold values, the event and phase probabilities surprisingly differ to some extent. For example, if I lowered the threshold, the output probabilities decrease (sometimes notably).
The package was last accessed and installed in Nov 2022.
Hi Mostafa,
I tested to run EQTransformer (conservative model) on one-day data multiple times.
I would get different results when the input and parameters were not changed. By changing the threshold values, the event and phase probabilities surprisingly differ to some extent. For example, if I lowered the threshold, the output probabilities decrease (sometimes notably).
The package was last accessed and installed in Nov 2022.
Best, Junhao