Question Answering on GPQA Diamond (Accuracy and Performance)
86.3Throughput (Req/s)SQUEEZE EVOLVE (p=0)
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| SQUEEZE EVOLVE (p=0)Model 1=GPT-OSS-20B, Model 2=GPT-OSS-120B, Obs. M1 share=100%, GPU split (G2:G1)=0:162026.04 | 86.3 | 2.84 | 79 | |
| SQUEEZE EVOLVE (p=10)Model 1=GPT-OSS-20B, Model 2=GPT-OSS-120B, Obs. M1 share=87.5%, GPU split (G2:G1)=4:122026.04 | 53.54 | 1.76 | 79.5 | |
| RSAModel 1=GPT-OSS-20B, Model 2=GPT-OSS-120B, Obs. M1 share=0%, GPU split (G2:G1)=16:02026.04 | 30.34 | 1 | 79.6 | |
| SQUEEZE EVOLVE (p=0)Model 1=Qwen3-30B-A3B-I, Model 2=Qwen3-235B-A22B-I, Obs. M1 share=100%, GPU split (G2:G1)=0:162026.04 | 22 | 10.71 | 83.8 | |
| SQUEEZE EVOLVE (p=10)Model 1=Qwen3-30B-A3B-I, Model 2=Qwen3-235B-A22B-I, Obs. M1 share=87.5%, GPU split (G2:G1)=8:82026.04 | 8.17 | 3.98 | 84 | |
| RSAModel 1=Qwen3-30B-A3B-I, Model 2=Qwen3-235B-A22B-I, Obs. M1 share=0%, GPU split (G2:G1)=16:02026.04 | 2.05 | 1 | 84.3 |