Purpose
A learning lab for foundations of machine learning and deep learning with Python and PyTorch, including experiments with local inference.
Proposed method
Frame a bounded task, prepare a suitable dataset, and separate training, validation, and test partitions. Use the validation set for model choices and reserve the test set for a final evaluation.
Evaluation
Choose a baseline and a metric appropriate to the task before training; inspect errors and document data limitations alongside results. Prevent leakage between partitions and record the environment and random seeds where practical.
Current status
The lab is in progress. The next milestone is a reproducible notebook with baseline results.