feat: add AveragePooling2D layer support to backend

[paramananda-15]

Description

Adds support for the AveragePooling2D layer in the TensorMap backend model generation service.

This change allows ReactFlow graphs to include AveragePooling nodes that are translated into
tf.keras.layers.AveragePooling2D when constructing the Keras functional model.

The implementation adds handling for a new node type (customavgpool) in
app/services/model_generation.py, enabling TensorMap to support average pooling operations
commonly used in convolutional neural networks.

This contribution aligns with the TensorMap GSoC project goal of expanding the
supported layer types in the backend.

Fixes #N/A

Type of Change

• Bug fix
• New feature
• Breaking change
• Documentation update

How Has This Been Tested?

All existing backend tests were executed locally after implementing the change
to ensure no regressions were introduced.

Command used:
python -m pytest

Results:
130 passed in ~3 seconds

• Existing tests pass
• New tests added
• Manual testing

Screenshots (if applicable)

N/A

Checklist

• My code follows the project's style guidelines
• I have performed a self-review
• I have added/updated documentation as needed
• My changes generate no new warnings
• Tests pass locally

[ivantha]

The backend model generation code for AveragePooling2D looks correct, but this PR only implements the backend half of the feature. There is no matching frontend implementation (no customavgpool node component, no sidebar button, and no properties panel entry). Please add the corresponding frontend pieces so the layer is actually usable, then this can be merged.

[paramananda-15]

The backend model generation code for AveragePooling2D looks correct, but this PR only implements the backend half of the feature. There is no matching frontend implementation (no customavgpool node component, no sidebar button, and no properties panel entry). Please add the corresponding frontend pieces so the layer is actually usable, then this can be merged.

Added full stack support for AveragePooling2D:

• Backend: model generation support
• Frontend:
  • AvgPoolNode component
  • Sidebar entry (customavgpool)
  • Canvas integration
  • NodePropertiesPanel support

Also fixed CI issues (lock file + formatting) and squashed commits.

Ready for review.