Introduction to artificial neural networks, with the implementation of a multilayer perceptron.
uv venv --python 3.12 .venv && uv pip install -e .uv run python examples/train_mlp.py
uv run python examples/inference_mlp.py- split train and prediction program
- refacto loss function
- implement SGD
my intuition is that giving 80% of the dataset in one pass of our model will make it learn so fast. it does not need to reajust much its weight.