SimpleConvPCLayer.h File Reference

ConvPCLayer with precision removed, AdamW/Adam support, now routed through IComputeBackend for GPU portability (mirrors SimplePCLayer's own IComputeBackend port). More...

#include <vector>
#include <stdexcept>
#include <random>
#include <memory>
#include <cstdlib>
#include <deepity/utils/Activations.h>
#include <deepity/utils/AdamOptimizer.h>
#include <deepity/layers/Layer.h>
#include <deepity/utils/MemoryArena.h>
#include <deepity/utils/DeviceMemoryArena.h>
#include <deepity/utils/Im2Col.h>
#include <deepity/backend/IComputeBackend.h>

Go to the source code of this file.

Classes
class	Deep::SimpleConvPCLayer
	Convolutional PC layer, precision-free, AdamW-capable, routed through IComputeBackend for GPU portability – see the file-level warning above for the GPU path's verification status. More...

Namespaces
namespace	Deep
	This file exists to provide both Tensor.h and IComputeBackend.h with device types without introducing circular dependencies.

Detailed Description

ConvPCLayer with precision removed, AdamW/Adam support, now routed through IComputeBackend for GPU portability (mirrors SimplePCLayer's own IComputeBackend port).

Warning

CPU correctness re-verified tonight via five independent, hand-computable tests (single-channel, multi-channel, real spatial kernel, padding, and multi-channel+real-kernel combined) plus a from-scratch Python im2col+GEMM reference comparison – all matched to float32 precision. The GPU path (once CUDABackend implements Im2Col/Col2Im/RepackForBatchedGemm/MultiplyInto/Fill/ AddBiasPerChannel) has NOT yet been tested at all.