[EHN] FBLightConvNet

[bruAristimunha]

No description provided.

[Copilot]

Copilot reviewed 5 out of 6 changed files in this pull request and generated no comments.

Files not reviewed (1)

• braindecode/models/summary.csv: Language not supported

Comments suppressed due to low confidence (1)

braindecode/models/fblightconvnet.py:50

• The docstring contains a grammatical error in the 'Final Layer' description ('to with kernel sizes'); consider rephrasing it for clarity.

through a linear layer to with kernel sizes matching the input dimensions to integrate features across different channels generate the final output predictions.

[codecov]

Codecov Report

Attention: Patch coverage is 92.94118% with 6 lines in your changes missing coverage. Please review.

Project coverage is 87.46%. Comparing base (8b443e5) to head (ab7abfc).

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #723      +/-   ##
==========================================
+ Coverage   87.41%   87.46%   +0.04%     
==========================================
  Files          79       80       +1     
  Lines        7313     7388      +75     
==========================================
+ Hits         6393     6462      +69     
- Misses        920      926       +6     

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[Copilot]

Copilot reviewed 5 out of 7 changed files in this pull request and generated no comments.

Files not reviewed (2)

• braindecode/models/summary.csv: Language not supported
• docs/whats_new.rst: Language not supported

Comments suppressed due to low confidence (1)

braindecode/models/fblightconvnet.py:180

• The input_size passed to _LightweightConv1d is self.n_filters_spat, but the forward method of _LightweightConv1d reshapes the input to use the number of heads (self.heads) as the channel dimension. Ensure that self.n_filters_spat equals self.heads or adjust the reshaping logic in _LightweightConv1d to prevent a potential dimension mismatch.

self.attn_conv = _LightweightConv1d(self.n_filters_spat, (self.n_times // self.win_len), heads=self.heads, weight_softmax=weight_softmax, bias=bias)

[bruAristimunha]

thank you @martinwimpff