[Test] Added call of hls library in .cpp file and created a shell script with the compile command (should only stay during dev->delete later)

[Test] Changed execute_custom_node.py and test cpp file to check if passing arrays via .npy files works

[Test] Omit removing of the temporary file which contains the input values and created a .cpp for testing the single node execution

[Execution] Added if tree to read out the attribute backend to enable the path for executing of hls lib custom nodes

since w1a2 is coming

[Code generation] Changed code_gen so that the input and output names for the compute command for all layers, except the last one, are correct

[Code generation] Added inputs and output of node as layer attribute in layer.py, changed code_gen accordingly

[Code generation] Added new (empty) python script to create an interface between onnx representation and corresponding fpga layers

[Code generation] Added some of the missing layer parameters (WMEM, TMEM and API) to get_layer_parameters()

[Code generation] Added function to determine the number of FIFOs and intergrated it into the generation of the stream declarations to help writing the parameters as variables into the commands