A collection of systems programming, embedded, and low-level projects. Lorem ipsum dolor sit amet, consectetur adipiscing elit.
A miniature PyTorch built end-to-end: reverse-mode autograd engine, tensor library with broadcasting, neural network modules (Linear, Conv2d, LayerNorm, attention), Adam/SGD optimizers, and a CUDA backend with hand-written kernels for matmul, softmax, and im2col convolution. Trained a character-level transformer on TinyShakespeare and a small ResNet on CIFAR-10 using only the framework. Optimized matmul kernel reaches a meaningful fraction of cuBLAS performance on 1024×1024 matrices.
Vestibulum ante ipsum primis in faucibus orci luctus et ultrices posuere cubilia curae. Donec velit neque, auctor sit amet aliquam vel, ullamcorper sit amet ligula.
Curabitur aliquet quam id dui posuere blandit. Nulla quis lorem ut libero malesuada feugiat. Proin eget tortor risus. Nulla porttitor accumsan tincidunt.
Pellentesque in ipsum id orci porta dapibus. Curabitur non nulla sit amet nisl tempus convallis quis ac lectus. Sed porttitor lectus nibh.
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam.
Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident.