Because A2 creates 2 models and combines their ESR together (Lite and Full), the comment about the ESR value at the end of training is no longer relevant.
For example, see below. This model turned out with a Full model ESR of 0.006 which is well within the acceptable range.
I would say that it probably shouldn't say that it might sound okay unless the combined ESR is over 0.08
Error-signal ratio (channels_3) = 0.02888
Error-signal ratio (channels_8) = 0.007281
Aggregate error-signal ratio = 0.03616
...This *might* sound ok!
Model training complete!
Because A2 creates 2 models and combines their ESR together (Lite and Full), the comment about the ESR value at the end of training is no longer relevant.
For example, see below. This model turned out with a Full model ESR of 0.006 which is well within the acceptable range.
I would say that it probably shouldn't say that it might sound okay unless the combined ESR is over 0.08