Large Language Model Evaluation on LLaMA 3B 3.2 benchmark leaderboard · SOTA2 Research

Large Language Model Evaluation on LLaMA 3B 3.2

7.81PPLbaseline

Evaluation Results

Method	Links
baseline2026.03		7.81	62.66	54.06
SERQ2026.03		9.15	58.5	45.7
QuaRot2026.03		9.73	55.76	44.75
SpinQuant2026.03		10.15	56.88	42.42
SmoothQ(g128)2026.03		53.33	44.17	27.31
SmoothQ2026.03		37,300	35.72	23.41