Error while verifying Backpropagation_Single_Xor.ino with Arduino IDE on an Arduino DUE board

[simi3103]

Hi,

First, thank you for your work and I'm please to use your library for my studies.

Second, I had encountered an error while verifying the Backpropagation_Single_Xor.ino file on the IDE.

The output was:
`𝖀𝖓𝖈𝖔𝖓𝖉𝖎𝖙𝖎𝖔𝖓𝖆𝖑 𝕷𝖔𝖛𝖊 "VERSION: 4.1.8-Alpha"

• " [1] 0B00010000 [Δ] [𝗥𝗲𝗺𝗶𝗻𝗱𝗲𝗿] Using (REDUCE_RAM_WEIGHTS_LVL2)."
• " [1] 0B00001000 [Ι] [𝗥𝗲𝗺𝗶𝗻𝗱𝗲𝗿] Using (REDUCED_SKETCH_SIZE_DOT_PROD)."

𝗨𝗦𝗜𝗡𝗚 [𝗗𝗘𝗡𝗦𝗘] [ƒx] |> ELU |~|

#pragma message( STR(INFORMATION) )`

Could you help me fix it please?

Regards

[GiorgosXou]

This is not an error, this is just a #pragma message informing\reminding you, that you are using those 2 optimizations: REDUCE_RAM_WEIGHTS_LVL2 and REDUCED_SKETCH_SIZE_DOT_PROD. See also: https://github.com/GiorgosXou/NeuralNetworks?tab=readme-ov-file#define-macro-properties

[simi3103]

Thanks for your reply.

What those two optimizations mean for the XOR example?
Is it possible to modify them?

Regards

[GiorgosXou]

What those two optimizations mean for the XOR example?

For the XOR-example specifically, nothing. But generally speaking, for the Neural-Network itself, means:

REDUCE_RAM_WEIGHTS_LVL2: changes the way weights are kept\represented inside the NN. Practically, it recudes the overall logic from a 2D Array to a pointer to a single dimentional array. Simply said, it uses the weights directly as passed. (as far as my knowledge and testing goes) doesn't effect the overall performance negatively at all.

REDUCED_SKETCH_SIZE_DOT_PROD: Optimizes the (internal) accumulatedDotProduct function, minimizes sketch consumption. But may slightly reduce the performance in some case scenarios (haven't done many performance tests so I can't say for sure).

Is it possible to modify them?

Of course you can, simply change the 1s to 0s . but first see https://github.com/GiorgosXou/NeuralNetworks?tab=readme-ov-file#define-macro-properties

[GiorgosXou]

And btw, if you need any other help feel free to join https://discord.gg/R2RVeD3gY8 I'd be more than happy to help you.

[simi3103]

Thanks again for your help.

I will probably join it :)