List your company

Long-Context Generation on PG-19 60K context length

6.29Throughput Speedup (micro)SpecPV-2K
1.94283.07144.25.3286Dec 2, 2025

Evaluation results

MethodLinks
SpecPV-2KBackbone=LLaMA3.1-8B-Instruct, Partial KV Budget=2K2025.126.293.57
SpecPV-4KBackbone=LLaMA3.1-8B-Instruct, Partial KV Budget=4K2025.126.153.56
SpecPV-8KBackbone=LLaMA3.1-8B-Instruct, Partial KV Budget=8K2025.125.83.57
TriForceBackbone=LLaMA3.1-8B-Instruct, Partial KV Budget=4K2025.123.172.61
EAGLE3-YARNBackbone=LLaMA3.1-8B-Instruct2025.123.113.48
TokenSwiftBackbone=LLaMA3.1-8B-Instruct, Partial KV Budget=4K2025.122.111.75