RobTand/prismaquant

Adaptive Precision for EXpert Models: MoE-aware mixed-precision quantization

LLM quantization optimizer

PrismQuant: Optimal Null-Space Rotations for Grouped Quantizers

[ICML 2024 Oral] Any-Precision LLM: Low-Cost Deployment of Multiple, Different-Sized LLMs

Gumbel-Softmax post-training quantization for LLMs (1–3 bit scalar, INT/GGUF-compatible).

RCO: Optimization of Non-Decomposable Objectives under Exact Budget Constraints for LLM Quantization and Pruning

LLM model quantization (compression) toolkit with HW acceleration support for Nvidia, AMD, Intel GPU and Intel/AMD/Apple CPU via HF, vLLM, and SGLang.

An easy-to-use LLMs quantization package with user-friendly apis, based on GPTQ algorithm.

Statistically-Lossless Quantization of Large Language Models