List your company

Text Generation Throughput on R1-Llama-8B 32K generation

2.51Memory Footprint (%)ThinKV
-1.389624.932751.25577.5773Oct 1, 2025

Evaluation results

MethodLinks
ThinKVTok. Budget=1024, Evaluation Protocol=Standard2025.102.517118,412.293810,578.5
R-KV (seq)Tok. Budget=1024, Evaluation Protocol=Standard2025.105.482681,450.53502,425.8
R-KV (ovl)Tok. Budget=1024, Evaluation Protocol=Standard2025.105.482682,320.93504,311.3
R-KV (seq)Tok. Budget=1024, Evaluation Protocol=Iso-batch, Iso-compression comparison2025.105.482561,769.32562,489.8
R-KV (ovl)Tok. Budget=1024, Evaluation Protocol=Iso-batch, Iso-compression comparison2025.105.482563,539.32565,318.7
ThinKV w/o TBQTok. Budget=1024, Evaluation Protocol=Iso-batch, Iso-compression comparison2025.105.782565,298.42568,079.9
FullKVEvaluation Protocol=Standard2025.1010013297.519453.9