TY  - RPRT
TI  - Low-Bit Quantization Favors Undertrained LLMs: Scaling Laws for Quantized LLMs with 100T Training Tokens
AU  - Xu Ouyang
AU  - Tao Ge
AU  - Thomas Hartvigsen
AU  - Zhisong Zhang
AU  - Haitao Mi
AU  - Dong Yu
PY  - 2024
UR  - https://arxiv.org/abs/2411.17691
ID  - 2411.17691
ER  -