Issue #1: Nested FlowUnit Wrapping in Neural Network Layers

[Shyanil]

Issue #1 Bug Report: Nested FlowUnit Wrapping in Neural Network Layers

Description

When passing an input through the NeuralNetwork layers, FlowUnit objects are getting nested multiple times instead of maintaining a single wrapper. This results in outputs like FlowUnit(FlowUnit(...)), which is incorrect.

Steps to Reproduce

1. Initialize the neural network:

   X = [2.0, 3.0, -1.0]
   n = NeuralNetwork(3, [4, 4, 1])

2. Forward pass through layers:

   layer_1_output = n.layers[0](X)
   print("First Layer Output:", layer_1_output)
   
   layer_2_output = n.layers[1](layer_1_output)
   print("Second Layer Output:", layer_2_output)
   
   final_output = n.layers[2](layer_2_output)
   print("Final Output:", final_output)
   
   print("The final output")
   n(X)

3. Observed Output:

   First Layer Output: [FlowUnit(-0.7321), FlowUnit(1.5557), FlowUnit(-1.0274), FlowUnit(-1.6452)]
   Second Layer Output: [FlowUnit(FlowUnit(0.9508)), FlowUnit(FlowUnit(-2.0511)), ...]
   Final Output: FlowUnit(FlowUnit(FlowUnit(-1.1022)))
   

Expected Behavior

Each layer should return a list of FlowUnit objects, but they should not be nested inside each other. Instead of FlowUnit(FlowUnit(...)), the output should be:

Final Output: FlowUnit(-1.1022)

Possible Cause

• Each layer might be applying FlowUnit() wrapping without checking if the input is already a FlowUnit object.