Dave LumAI·Apr 24Episode 2: Bits, Bytes, and Data Representation, or How Everything Becomes Zeros, Ones, and…By the time most people start programming, they already know the machine can store numbers, print text, show images, play music, and…
Mohamed Hisham·Feb 22I Found a Bug in Hugging Face’s 4-bit Quantization While Trying to Make It FasterTL;DR: In September 2025, I submitted two PRs to bitsandbytes: one fixing an integer packing bug that was affecting quantization accuracy…
Inbits2botsbyChaitanya V·Oct 21, 2025Logic Gates: How Computers Make DecisionsEvery intelligent machine — from your smartwatch to ChatGPT — relies on one simple trick: turning electricity into logic.
InAI AdvancesbyPhrugsa Limbunlom (Gift)·Aug 12, 2025Choosing the right model size to your memory: Understanding bits & bytes and Model QuantizationA Deep Dive into Quantization Techniques and Bits & Bytes Consideration for Achieving Significant Model Compression.A response icon1A response icon1
Akanksha Sinha·Jun 14, 2025When Fine-Tuning Fails: 3 Mistakes to Avoid + What I’d Do DifferentlyDay 74 of #100DaysOfAI | Practical lessons from debugging a LoRA + QLoRA fine-tune on a local machine — without a GPU
Sophie Kearney·May 7, 2025How I Fixed Bitsandbytes CUDA Errors for QLoRA LLM Fine-Tuning on an HPCA few weeks ago, I found myself stuck in a programmer’s purgatory — endlessly trying to align conflicting package versions just to begin…A response icon1A response icon1
LambdaFlux·Apr 2, 2025Packing and Unpacking Weights in 4 bit QuantizationThis article explains how packing and unpacking of quantized weights works for efficient low-bit quantization and how bitsandbytes library…
