Throughput Measurement on GPQA Diamond
1,440.8Throughput (tokens/s)FLARE-2B
Evaluation Results
| Method | Links | |
|---|---|---|
| FLARE-2BConcurrency level (C)=82026.06 | 1,440.8 | |
| FLARE-4BConcurrency level (C)=82026.06 | 990.4 | |
| FLARE-2BConcurrency level (C)=42026.06 | 949 | |
| FLARE-9BConcurrency level (C)=82026.06 | 786.5 | |
| FLARE-4BConcurrency level (C)=42026.06 | 649.3 | |
| FLARE-9BConcurrency level (C)=42026.06 | 499 | |
| SDAR-1.7BConcurrency level (C)=82026.06 | 461.5 | |
| SDAR-4BConcurrency level (C)=82026.06 | 441.7 | |
| LLaDA-2.1-miniConcurrency level (C)=82026.06 | 401.2 | |
| SDAR-1.7BConcurrency level (C)=42026.06 | 383.1 | |
| SDAR-8BConcurrency level (C)=82026.06 | 379.3 | |
| LLaDA-2.0-miniConcurrency level (C)=82026.06 | 373.2 | |
| FLARE-2BConcurrency level (C)=12026.06 | 330.7 | |
| SDAR-4BConcurrency level (C)=42026.06 | 313.6 | |
| LLaDA-2.1-miniConcurrency level (C)=42026.06 | 283.9 | |
| LLaDA-2.0-miniConcurrency level (C)=42026.06 | 264.2 | |
| FLARE-4BConcurrency level (C)=12026.06 | 237.4 | |
| SDAR-8BConcurrency level (C)=42026.06 | 229.6 | |
| LLaDA-2.1-miniConcurrency level (C)=12026.06 | 226.9 | |
| SDAR-30B-A3BConcurrency level (C)=82026.06 | 210.2 | |
| LLaDA-2.0-miniConcurrency level (C)=12026.06 | 182.6 | |
| FLARE-9BConcurrency level (C)=12026.06 | 170.8 | |
| SDAR-1.7BConcurrency level (C)=12026.06 | 130.4 | |
| SDAR-30B-A3BConcurrency level (C)=42026.06 | 120.6 | |
| SDAR-4BConcurrency level (C)=12026.06 | 95.7 | |
| SDAR-8BConcurrency level (C)=12026.06 | 72.2 | |
| SDAR-30B-A3BConcurrency level (C)=12026.06 | 43.2 |