[fix] fix bug #1580

[MaiBe-ctrl]

🔬 Background

• closes RuntimeError: Subtraction, the - operator, with a bool tensor is not supported. If you are trying to invert a mask, use the ~ or logical_not() operator instead. #1580
• closes Torch error under some optimization parameters during attempted Optuna outer-loop optimization #1575

🔮 Key changes

• Change the order of init of trainer and trend

📋 Review Checklist

• I have performed a self-review of my own code.
• I have commented my code, added docstrings and data types to function definitions.
• I have added pytests to check whether my feature / fix works.

Please make sure to follow our best practices in the Contributing guidelines.

[github-actions]

Model Benchmark

Benchmark	Metric	main	current	diff
PeytonManning	MAE_val	-	0.3496	-
PeytonManning	RMSE_val	-	0.5005	-
PeytonManning	Loss_val	-	0.0177	-
PeytonManning	MAE	-	0.3465	-
PeytonManning	RMSE	-	0.4931	-
PeytonManning	Loss	-	0.0146	-
PeytonManning	time	-	14.01	-
AirPassengers	MAE_val	-	30.6285	-
AirPassengers	RMSE_val	-	31.5255	-
AirPassengers	Loss_val	-	0.0128	-
AirPassengers	MAE	-	6.1686	-
AirPassengers	RMSE	-	7.8523	-
AirPassengers	Loss	-	0.0006	-
AirPassengers	time	-	8.95	-
EnergyPriceDaily	MAE_val	-	5.4116	-
EnergyPriceDaily	RMSE_val	-	6.7154	-
EnergyPriceDaily	Loss_val	-	0.0253	-
EnergyPriceDaily	MAE	-	5.9494	-
EnergyPriceDaily	RMSE	-	7.9833	-
EnergyPriceDaily	Loss	-	0.0258	-
EnergyPriceDaily	time	-	17.39	-
YosemiteTemps	MAE_val	-	0.5641	-
YosemiteTemps	RMSE_val	-	0.8316	-
YosemiteTemps	Loss_val	-	0.0004	-
YosemiteTemps	MAE	-	0.9845	-
YosemiteTemps	RMSE	-	1.7539	-
YosemiteTemps	Loss	-	0.0013	-
YosemiteTemps	time	-	37.94	-

\n

Model training plots

\n ## Model Training ### PeytonManning  ### YosemiteTemps  ### AirPassengers  ### EnergyPriceDaily  \n