This weights and architecture of this NN will be used in my Pipeline project as an example of a "general neural network".
Python 3.9+ (with PyTorch + pandas)
Generate the dataset file (totals.csv) with:
$ python ./gen_numbers.py
Run the trainer:
$ python ./trainer.py
After training finishes the weights will be saved to output file. The file name format is:
{number of features}ftrs_{number of epochs}epochs_{batch size}batch.pth
Number of input features is two in our case (we compute total of two numbers).
Number of epochs and batch size can be changed in the Trainer's
constructor.