Description
The num_experiments field defines the number of experiments AutoML will run to optimize your model. Each experiment involves training and evaluating a model with different configurations.
A higher number of experiments increases the chances of finding the best-performing model but also requires more computational resources and time.
num_experimentsmust be>= 1.
Supported Task Types
- All
Default Values (Classification Tasks)
run_mode | Default Value |
|---|---|
| FAST | 2 |
| NORMAL | 4 |
| BEST | 8 |
Default Values (Other Tasks)
run_mode | Default Value |
|---|---|
| FAST | 4 |
| NORMAL | 8 |
| BEST | 16 |
Updated about 1 month ago
