Rocky a9609d6475 fix Dropout to use the module-level seeded rng instead of global np.random (#1869) ...

The module declares rng = np.random.default_rng(7) at the top so that all
stochastic operations are reproducible for a given seed. Dropout ignored this
and called np.random.random(shape), which draws from the global unseeded state.
Masks were different on every run even when the module seed was set, making
Dropout unit tests non-reproducible and causing hard-to-debug numerical
differences between training runs.

Changed to rng.random(shape) so Dropout participates in the same reproducible
stream as every other random operation in the module.

2026-06-16 18:58:38 -04:00

..

03_layers.py

fix Dropout to use the module-level seeded rng instead of global np.random (#1869)

2026-06-16 18:58:38 -04:00

module.yaml

feat(tinytorch): add module.yaml metadata files for CLI module descriptions

2025-12-15 20:30:32 -05:00