Curated 9-paper reading list for LLM quantization beginners, from QLoRA to BitNet
East-Muffin-6472 · reddit · 2026-11-12
A Reddit user curated a beginner-friendly reading list on LLM quantization, covering post-training quantization, QAT, and 1-bit weights across 9 papers: BitNet (1-bit Transformers), AWQ, The Era of 1-bit LLMs (1.58 bits), EfficientQAT, LLM-QAT, LLM.int8(), QLoRA, SmoothQuant, and ZeroQuant.
The companion site papercase.dev presents them in community reading order and clearly states the audience (engineers who know transformers), coverage, and notable omissions like GPTQ and newer formats (NVFP4, MXFP4).
More from Infra
- Local LLM on 3090 + 64GB RAM hits 70-80 t/s with QFN quants; eyeing a $3-3.5k Strix Halo box — makkalot · 2026-11-12
- Running ComfyUI on NVIDIA DGX Spark with Docker Compose — setup shared — Ok_Joke5408 · 2026-11-12
- Beff Jezos: Even the toilets in Japan are running 3B-parameter models — beffjezos · 2026-11-12
- llama.cpp PR #31915 fixes Gemma4 tool calling, user reports it "completely fixed" agentic coding — jacek2023 · 2026-11-12
- Nvidia board reportedly working with outside firm to find Jensen Huang successor — kimmonismus · 2026-11-12
- RDNA3-tuned llama.cpp fork doubles Qwen3.8-27B speed on RX 7900 XTX to ~100 tok/s — Jebbyk1 · 2026-11-12