The V Tensor Library

VTL is a pure-V tensor library for numerical computing and machine learning. It provides n-dimensional arrays with automatic differentiation, linear algebra via VSL, and a full neural network module.

Documentation · Tutorials · VSL Backend

Features

• Tensor operations — create, slice, reshape, transpose, broadcast, map/reduce
• Autograd — reverse-mode automatic differentiation; build arbitrary computational graphs
• Neural networks — Sequential model API; layers (Linear, Conv2D, LSTM, Attention, …); losses (MSE, BCE, CrossEntropy, Huber, …); optimizers (Adam, AdamW, RMSProp, AdaGrad, SGD)
• Linear algebra — VSL-backed: matmul, solve, lstsq, qr, lu, cholesky, pinv, trace, norm, SVD
• Hardware acceleration — zero-copy sharing with C libraries via vtl.Tensor.data

Quick start

import vtl
import vtl.autograd
import vtl.nn.layers
import vtl.nn.models
import vtl.nn.optimizers

// 1. Build a two-layer network using Sequential API
mut ctx := autograd.ctx[f64]()
mut model := models.sequential_from_ctx[f64](ctx)
model.input([784])
model.linear(256) // 784 -> 256
model.linear(10) // 256 -> 10

// 2. Forward pass
input_tensor := vtl.zeros[f64]([64, 784])
mut x := ctx.variable(input_tensor)
y_pred := model.forward(x)!

// 3. Loss
target := vtl.zeros[f64]([64, 10])
mut loss_val := model.loss(y_pred, target)!

// 4. Backward + update
loss_val.backprop()!
mut opt := optimizers.adam_optimizer[f64](optimizers.AdamOptimizerConfig{
	learning_rate: 0.001
})
opt.build_params(model.info.layers)
opt.update()!

Module overview

Module	Purpose
vtl	Core Tensor[T] type; creation, slicing, broadcasting
vtl.autograd	Context, Variable, gates, backprop()
vtl.la	Linear algebra (wraps VSL)
vtl.nn	Neural network layers, losses, optimizers
vtl.nn.models	Sequential model API
vtl.nn.internal	Weight initialisation (Kaiming, Xavier)
vtl.nn.gates	Autograd gate implementations

Installation

VTL depends on VSL. Follow the VSL install instructions before installing VTL.

v install vtl

Testing

v test ~/.vmodules/vtl

Documentation

All tutorials live in docs/:

Tutorial	Topic
TUTORIAL_FIRST_STEPS.md	Tensor creation, indexing, slicing
TUTORIAL_MAP_REDUCE.md	map / nmap and reductions
TUTORIAL_AUTOGRAD.md	Autograd: Variable, gates, backprop
TUTORIAL_REDUCTIONS.md	argmax / argmin / cumsum / cumprod
TUTORIAL_NEURAL_NETWORKS.md	Layers, losses, optimizers, Sequential
TUTORIAL_OPTIMIZERS.md	Adam / AdamW / RMSProp / AdaGrad / SGD + schedulers
TUTORIAL_LINEAR_ALGEBRA.md	LA basics via VSL
TUTORIAL_ADVANCED_LA.md	trace / norm / qr / lu / cholesky / pinv
TUTORIAL_BROADCASTING.md	Broadcasting rules
TUTORIAL_SLICING.md	Slicing and views

License

MIT