fix deep copying of TFM trainer parameters

[dennisbader]

Fixes #1455.

Summary

• fixes mutable copying of trainer parameters in TorchForecastingModels. This resulted in PyTorch Lightning appending their callbacks to the original trainer_params.

[codecov-commenter]

Codecov Report

Base: 93.93% // Head: 93.92% // Decreases project coverage by -0.00% ⚠️

Coverage data is based on head (a57d39c) compared to base (fab7ddf).
Patch coverage: 100.00% of modified lines in pull request are covered.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1459      +/-   ##
==========================================
- Coverage   93.93%   93.92%   -0.01%     
==========================================
  Files         122      122              
  Lines       10727    10714      -13     
==========================================
- Hits        10076    10063      -13     
  Misses        651      651              

Impacted Files	Coverage Δ
...arts/models/forecasting/torch_forecasting_model.py	89.52% <100.00%> (-0.03%)	⬇️
darts/ad/anomaly_model/filtering_am.py	91.93% <0.00%> (-0.13%)	⬇️
darts/timeseries.py	91.78% <0.00%> (-0.06%)	⬇️
darts/models/forecasting/block_rnn_model.py	98.24% <0.00%> (-0.04%)	⬇️
darts/models/forecasting/nhits.py	99.27% <0.00%> (-0.01%)	⬇️
darts/datasets/__init__.py	100.00% <0.00%> (ø)

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.