Fix/unittests m1

[dennisbader]

Summary

Fixes unit tests to run locally on M1 devices

• run all torch models on CPU
  • by default most time series are float64 which is not supported by M1 GPU
  • not all torch functionalities that we use in our TorchForecastingModel are supported on M1 GPU
• fixes a bug in TiDE model where it was not possible to predict with n < output_chunk_length

[codecov-commenter]

Codecov Report

Patch coverage: 90.00% and project coverage change: -0.02% ⚠️

Comparison is base (2f40d97) 93.88% compared to head (42e3f82) 93.87%.

❗ Your organization is not using the GitHub App Integration. As a result you may experience degraded service beginning May 15th. Please install the Github App Integration for your organization. Read more.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1933      +/-   ##
==========================================
- Coverage   93.88%   93.87%   -0.02%     
==========================================
  Files         132      132              
  Lines       12667    12655      -12     
==========================================
- Hits        11893    11880      -13     
- Misses        774      775       +1     

Files Changed	Coverage Δ
darts/models/forecasting/tft_model.py	97.25% <ø> (-0.03%)	⬇️
darts/models/forecasting/tide_model.py	97.32% <90.00%> (+0.09%)	⬆️

... and 7 files with indirect coverage changes

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[madtoinou]

LGTM, some minor comments about typo or coherence