1.8 KiB
1.8 KiB
| 1 | ctx_tokens | prefill_tokens | prefill_tps | gen_tokens | gen_tps | gen_first_ms | gen_steady_tokens | gen_steady_tps | kvcache_bytes |
|---|---|---|---|---|---|---|---|---|---|
| 2 | 2048 | 2048 | 825.76 | 128 | 18.05 | 71.721 | 127 | 18.20 | 52184460 |
| 3 | 4096 | 2048 | 899.52 | 128 | 15.47 | 57.061 | 127 | 15.54 | 80373132 |
| 4 | 6144 | 2048 | 888.62 | 128 | 15.37 | 65.899 | 127 | 15.45 | 108561804 |
| 5 | 8192 | 2048 | 882.80 | 128 | 15.21 | 65.667 | 127 | 15.29 | 136750476 |
| 6 | 10240 | 2048 | 874.95 | 128 | 15.13 | 67.430 | 127 | 15.21 | 164939148 |
| 7 | 12288 | 2048 | 872.74 | 128 | 15.08 | 67.877 | 127 | 15.16 | 193127820 |
| 8 | 14336 | 2048 | 867.76 | 128 | 15.02 | 67.140 | 127 | 15.10 | 221316492 |
| 9 | 16384 | 2048 | 872.44 | 128 | 15.10 | 67.128 | 127 | 15.18 | 249505164 |
| 10 | 18432 | 2048 | 871.75 | 128 | 15.04 | 67.562 | 127 | 15.12 | 277693836 |
| 11 | 20480 | 2048 | 869.55 | 128 | 15.00 | 67.678 | 127 | 15.08 | 305882508 |
| 12 | 22528 | 2048 | 866.37 | 128 | 14.96 | 68.248 | 127 | 15.04 | 334071180 |
| 13 | 24576 | 2048 | 864.65 | 128 | 14.91 | 68.667 | 127 | 14.99 | 362259852 |
| 14 | 26624 | 2048 | 861.34 | 128 | 14.86 | 68.919 | 127 | 14.94 | 390448524 |
| 15 | 28672 | 2048 | 859.07 | 128 | 14.82 | 69.044 | 127 | 14.90 | 418637196 |
| 16 | 30720 | 2048 | 856.88 | 128 | 14.80 | 68.233 | 127 | 14.87 | 446825868 |
| 17 | 32768 | 2048 | 855.94 | 128 | 14.43 | 68.704 | 127 | 14.51 | 475014540 |
| 18 | 34816 | 2048 | 850.96 | 128 | 14.36 | 70.661 | 127 | 14.43 | 503203212 |
| 19 | 36864 | 2048 | 849.83 | 128 | 14.34 | 70.827 | 127 | 14.41 | 531391884 |
| 20 | 38912 | 2048 | 849.59 | 128 | 14.30 | 70.971 | 127 | 14.38 | 559580556 |
| 21 | 40960 | 2048 | 847.40 | 128 | 14.28 | 70.881 | 127 | 14.36 | 587769228 |
| 22 | 43008 | 2048 | 843.24 | 128 | 14.24 | 71.968 | 127 | 14.32 | 615957900 |
| 23 | 45056 | 2048 | 840.41 | 128 | 14.19 | 71.406 | 127 | 14.26 | 644146572 |
| 24 | 47104 | 2048 | 839.95 | 128 | 14.16 | 71.523 | 127 | 14.23 | 672335244 |
| 25 | 49152 | 2048 | 836.64 | 128 | 14.12 | 72.153 | 127 | 14.19 | 700523916 |
| 26 | 51200 | 2048 | 837.72 | 128 | 14.08 | 71.579 | 127 | 14.15 | 728712588 |
| 27 | 53248 | 2048 | 837.01 | 128 | 14.04 | 72.138 | 127 | 14.11 | 756901260 |
| 28 | 55296 | 2048 | 829.92 | 128 | 14.00 | 73.733 | 127 | 14.08 | 785089932 |
| 29 | 57344 | 2048 | 830.16 | 128 | 13.98 | 72.862 | 127 | 14.05 | 813278604 |
| 30 | 59392 | 2048 | 829.11 | 128 | 13.94 | 73.157 | 127 | 14.01 | 841467276 |
| 31 | 61440 | 2048 | 827.26 | 128 | 13.90 | 73.924 | 127 | 13.97 | 869655948 |
| 32 | 63488 | 2048 | 825.04 | 128 | 13.86 | 73.787 | 127 | 13.93 | 897844620 |
| 33 | 65536 | 2048 | 822.98 | 128 | 13.84 | 73.966 | 127 | 13.91 | 0 |