Mmap | martinuke0's Blog

From‑Scratch Safetensors & GGUF Weight Loader with mmap‑Backed Zero‑Copy Views

A practical, runnable guide to constructing a from‑scratch safetensors/GGUF weight loader with zero‑copy mmap views, dtype/quantization decoding, and shape validation, perfect for a CV‑boosting side project.

Hands‑On: Building a Zero‑Dependency Pure‑Python GGUF Model Loader with Memory‑Mapped Tensor Access

A lightweight, dependency‑free Python loader for GGUF models that uses memory‑mapped files and on‑the‑fly dtype conversion, perfect for a standout CV project.

Lazy GGUF/Safetensors Checkpoint Loader: A Hands-On Build Guide

A practical guide to building a lazy checkpoint loader that memory-maps GGUF and safetensors files, parses tensor metadata, and dequantizes on demand.

Hands‑On Build Guide: Pure‑Python Safetensors Parser & Weight Shard Assembler with mmap Lazy Loading

A step‑by‑step guide to creating a lightweight, mmap‑based safetensors parser you can ship, test, and extend for real‑world model serving.

From-Scratch GGUF Weight Loader with Blockwise Dequantization, mmap Streaming, and Safetensors Export

A practical guide to implementing a GGUF loader with blockwise dequantization and safetensors export, demonstrating low-level systems expertise.