Refactor One-shot configuration #156
Labels
improvement
Improvement that don't change the behaviour for a user
performance
Temporal/memory performance of the algorithm
Currently, the one-shot option is implemented by running the HyperOpt stage only once, so the setup values are automatically put in the
best_bps_model
variable.To improve performance, might be interesting to avoid running HyperOpt stages and directly take the setup values as the final ones. This is, directly jump to the discovery of the
final_bps_model
.The text was updated successfully, but these errors were encountered: