tiramisu-ml 0.1.1

ML framework with tensors, autograd, NN modules, optimizers, transformers.

tiramisu

A from-scratch ML framework in C++20 — about 5,000 lines of readable, PyTorch-familiar code with a Python binding.

• Stdlib-only compute (no Eigen, no BLAS, no PyTorch at link time)
• Tensor, requires_grad, backward(), Module, Linear, Adam — Python and C++
• AVX2/FMA matmul on x86, scalar fallback elsewhere
• Optional CUDA backend for GPU training
• End-to-end MNIST and char-level GPT examples

Install

pip install tiramisu-ml

Requires Python 3.10+. Wheels ship for Linux (x86_64, aarch64) and macOS (x86_64, arm64); other platforms build from sdist and need CMake + a C++20 compiler.

Quick start

import numpy as np
import tiramisu as tr

x = tr.from_numpy(np.random.randn(2, 784).astype(np.float32))
layer = tr.nn.Linear(784, 10)
out = layer.forward(x)
print(out.shape())  # [2, 10]

Training step:

opt.zero_grad()
h = tr.relu(layer1.forward(batch_x))
logits = layer2.forward(h)
loss = tr.nn.cross_entropy_loss(logits, batch_y)
loss.backward()
opt.step()

C++ mirrors the Python API — see examples/mnist.cpp.

API

Tensor ops — add, sub, mul, div, neg, matmul, sum, mean, reshape, transpose, contiguous, relu, gelu, softmax, from_numpy, backward

Modules — nn.Linear, nn.LayerNorm, nn.GPT, nn.cross_entropy_loss

Optimizers — optim.Adam

Full binding reference in python/README.md.

Examples

MNIST (CPU, ~10 epochs to 95%+):

cmake -S . -B build -DCMAKE_BUILD_TYPE=Release
cmake --build build --target mnist
./build/examples/mnist

Data: MNIST IDX files in data/.

Char-level GPT on Tiny Shakespeare:

cmake --build build --target train_shakespeare
./build/examples/train_shakespeare --preset tiny --epochs 3

Presets: tiny, 2m, 10m. Add --cuda for GPU.

Build from source

cmake -S . -B build -G Ninja -DCMAKE_BUILD_TYPE=Release
cmake --build build --parallel
ctest --test-dir build --output-on-failure

CUDA: -DTIRAMISU_ENABLE_CUDA=ON. Debug builds enable ASan+UBSan by default.

Layout

core/       Storage, Tensor, dtype, device
ops/cpu/    Forward kernels (elementwise, reduce, matmul, normalization)
ops/cuda/   Optional CUDA kernels
autograd/   Differentiable wrappers, backward(), gradcheck
nn/         Module, Linear, GPT, LayerNorm, loss
optim/      SGD, Adam, AdamW, grad clipping, cosine LR
python/     pybind11 bindings
serialize/  GPT checkpoint save/load
examples/   hello_tiramisu, mnist, train_shakespeare

GoogleTest is the only non-stdlib fetch at configure time. License: MIT.