Open rlnsanz opened 7 years ago
We have a way to do fan-out of multi-trial experiments, but no way to do fan-in (such as picking the best model from many trials, and then using it later in the same pipeline). Aggregation must be expressed in Jarvis to be more expressive.
We have a way to do fan-out of multi-trial experiments, but no way to do fan-in (such as picking the best model from many trials, and then using it later in the same pipeline). Aggregation must be expressed in Jarvis to be more expressive.