Adaptive

Adaptive Models

This module contains the small neural network I built to inspect how forward passes, gradients, and weight updates compose on top of SicLib tensors. It is a learning and verification tool, not a benchmark against modern machine learning libraries.

Implementation Notes

Tensor operations are not optimized. Use an established library such as PyTorch for production machine learning.

---

Documentation

class ProtoNet:
    def __init__(self, arg0: int, arg1: int, arg2: int, arg3: int, arg4: float) -> None: ...
    def query_net(self, arg0: _pysiclib.linalg.Tensor) -> _pysiclib.linalg.Tensor: ...
    def run_epoch(self, arg0: _pysiclib.linalg.Tensor, arg1: _pysiclib.linalg.Tensor) -> None: ...
    @property
    def m_bias(self) -> List[_pysiclib.linalg.Tensor]: ...
    @property
    def m_transform(self) -> Callable[[float],float]: ...
    @property
    def m_transform_deriv(self) -> Callable[[float],float]: ...
    @property
    def m_weights(self) -> List[_pysiclib.linalg.Tensor]: ...