DyStack is enabled (dynamic_stacking=True). AutoGluon will try to determine whether the input data is affected by stacked overfitting and enable or disable stacking as a consequence.
	This is used to identify the optimal `num_stack_levels` value. Copies of AutoGluon will be fit on subsets of the data. Then holdout validation data is used to detect stacked overfitting.
	Running DyStack for up to 15s of the 60s of remaining time (25%).
	Running DyStack sub-fit in a ray process to avoid memory leakage. Logs will not be shown until this process is complete, due to a limitation in ray. You can force logging by specifying `ds_args={'force_ray_logging': True}`.
		Context path: "AutogluonModels/ag-20240518_004300/ds_sub_fit/sub_fit_ho"
Leaderboard on holdout data from dynamic stacking:
                 model  score_holdout  score_val eval_metric  pred_time_test  pred_time_val  fit_time  pred_time_test_marginal  pred_time_val_marginal  fit_time_marginal  stack_level  can_infer  fit_order
0       XGBoost_BAG_L1      -0.337663  -0.333345    log_loss        0.202160       0.017601  0.657651                 0.202160                0.017601           0.657651            1       True          2
1  WeightedEnsemble_L3      -0.339735  -0.331134    log_loss        0.360134       0.033437  2.938443                 0.001720                0.000586           0.010818            3       True          6
2  WeightedEnsemble_L2      -0.339999  -0.332223    log_loss        0.351907       0.026324  1.826918                 0.001708                0.000594           0.007622            2       True          3
3       XGBoost_BAG_L2      -0.345056  -0.349410    log_loss        0.372647       0.043121  2.426904                 0.022447                0.017391           0.607608            2       True          5
4      LightGBM_BAG_L2      -0.349683  -0.351669    log_loss        0.358413       0.032851  2.927625                 0.008214                0.007121           1.108329            2       True          4
5      LightGBM_BAG_L1      -0.357196  -0.346560    log_loss        0.148039       0.008130  1.161645                 0.148039                0.008130           1.161645            1       True          1
	1	 = Optimal   num_stack_levels (Stacked Overfitting Occurred: False)
	17s	 = DyStack   runtime |	43s	 = Remaining runtime
Starting main fit with num_stack_levels=1.
	For future fit calls on this dataset, you can skip dynamic stacking to save time: `predictor.fit(..., dynamic_stacking=False, num_stack_levels=1)`
Beginning AutoGluon training ... Time limit = 43s

Dynamic stacking is enabled (dynamic_stacking=True). AutoGluon will try to determine whether the input data is affected by stacked overfitting and enable or disable stacking as a consequence.
Detecting stacked overfitting by sub-fitting AutoGluon on the input data. That is, copies of AutoGluon will be sub-fit on subset(s) of the data. Then, the holdout validation data is used to detect stacked overfitting.
Sub-fit(s) time limit is: 60 seconds.
Starting holdout-based sub-fit for dynamic stacking. Context path is: AutogluonModels/ag-20240518_004822/ds_sub_fit/sub_fit_ho.
Running the sub-fit in a ray process to avoid memory leakage.
Spend 17 seconds for the sub-fit(s) during dynamic stacking.
Time left for full fit of AutoGluon: 43 seconds.
Starting full fit now with num_stack_levels 1.
Beginning AutoGluon training ... Time limit = 43s