AMEVA-Forge Logo

AMEVA-Forge

Release 1.0.0
PyPI (pip) Live WebGPU Studio 💖 Sponsor GitHub

Welcome to AMEVA-Forge (Release 1.0.0)

AMEVA-Forge is a lightweight deep learning framework engineered for educational autograd, small model experimentation, and client-side privacy-preserving computation across both native Python environments and modern web browsers via WebGPU.

PyPI Version PyPI Release Python Version WebGPU Platform Tests
1-Line Quick Installation

Install via pip for native/Pyodide usage, or launch the zero-install WebGPU Live Studio:

🐍 Python Edition (PyPI):

pip install ameva-forge

Release 1.0 Scope: 2-Layer MLP & Causal Attention

Our Release 1.0 focus is laser-targeted on one goal: Enable users to build, train, and understand neural networks (2-Layer MLP, Causal Attention, In-Place SGD) entirely in the browser using Python without server compute costs.

Core Capabilities

Supported Operations & Modules in Release 1.0

Category Supported Operations & Features
Tensor Operations add, sub, mul, div, matmul (2D), batched_matmul (3D BMM), transpose, permute, scatter, gather, 8D broadcasting
Neural Network (nn) nn.Linear, nn.LayerNorm, nn.BatchNorm2d (train/eval), nn.PositionalEncoding (LRU cache), nn.Dropout, nn.Sequential
Functional & Loss F.scaled_dot_product_attention (Causal), F.softmax, F.log_softmax, F.cross_entropy, nn.MSELoss, F.relu
Optimization (optim) optim.SGD (In-place parameter updates with zero allocation overhead)
GPU Shader Kernels elementwise.wgsl, matmul.wgsl (16x16 workgroup tiled), softmax.wgsl, layernorm.wgsl, cross_entropy.wgsl