List your company

LLM Inference Acceleration on MBPP

6.26SpeedupFLy
0.78962.20983.635.0502Nov 28, 2025

Evaluation results

MethodLinks
FLyModel=Llama-3.1-Instruct-405B, Training-based=false, Temperature=02025.116.2618.56
FLyModel=Llama-3.1-Instruct-405B, Training-based=false, Temperature=12025.115.9717.8
EAGLE-3Model=Meta-Llama-3.3-Instruct-70B, Training-based=true, Temperature=02025.113.85.47
EAGLE-3Model=Meta-Llama-3.3-Instruct-70B, Training-based=true, Temperature=12025.113.585.15
SpSModel=Llama-3.1-Instruct-405B, Training-based=false, Temperature=02025.113.0312.67
FLyModel=Llama-3.1-Instruct-70B, Training-based=false, Temperature=02025.112.7912.84
FLyModel=Meta-Llama-3.3-Instruct-70B, Training-based=false, Temperature=12025.112.6912.33
EAGLE-2Model=Meta-Llama-3-Instruct-70B, Training-based=true, Temperature=02025.112.613.9
FLyModel=Meta-Llama-3-Instruct-70B, Training-based=false, Temperature=12025.112.5911.37
FLyModel=Llama-3.1-Instruct-70B, Training-based=false, Temperature=12025.112.5912.57
SpSModel=Llama-3.1-Instruct-405B, Training-based=false, Temperature=12025.112.5510.67
FLyModel=Meta-Llama-3-Instruct-70B, Training-based=false, Temperature=02025.112.5311.21
FLyModel=Meta-Llama-3.3-Instruct-70B, Training-based=false, Temperature=02025.112.4711.76
EAGLE-2Model=Meta-Llama-3-Instruct-70B, Training-based=true, Temperature=12025.112.433.63
RESTModel=Llama-3.1-Instruct-405B, Training-based=false, Temperature=02025.112.362.46
TokenRecyclingModel=Llama-3.1-Instruct-70B, Training-based=false, Temperature=12025.112.153.01
TokenRecyclingModel=Llama-3.1-Instruct-70B, Training-based=false, Temperature=02025.112.133.01
RESTModel=Llama-3.1-Instruct-70B, Training-based=false, Temperature=02025.112.032.34
RESTModel=Llama-3.1-Instruct-405B, Training-based=false, Temperature=12025.112.022.11
RESTModel=Llama-3.1-Instruct-70B, Training-based=false, Temperature=12025.111.942.24
SpSModel=Llama-3.1-Instruct-70B, Training-based=false, Temperature=02025.111.7210.47
TokenRecyclingModel=Llama-3.1-Instruct-405B, Training-based=false, Temperature=12025.111.672.92
TokenRecyclingModel=Llama-3.1-Instruct-405B, Training-based=false, Temperature=02025.111.662.92
SpSModel=Llama-3.1-Instruct-70B, Training-based=false, Temperature=12025.1116.08