Round score
0
weighted tier speedup after failure cap and deduction
Eligible requests
—
trusted baseline exclusions removed
Failed eligible requests
—
each request counted once
Raw weighted speedup
—
before failure cap and deduction
The round score uses weighted completion times for all eligible work in each tier. Per-request speedups below are diagnostics, not score contributions. Only trusted baseline exclusions are removed; candidate failures remain in the eligible count and incur the configured failure penalty.
Reference engine
The baseline is what candidates are measured against, so it has no speedup of its own. Its SLA replay is below.
31 requests
| Request | TTFT | Total | Tokens | Slowest gap | Stream |
|---|---|---|---|---|---|
| hf-000 | 1.732s | 116.511s | 5120 | 2.538s | 2469 gaps for 5120 tokens |
| hf-001 | 3.774s | 113.785s | 4352 | 152.76ms | 2469 gaps for 4352 tokens |
| hf-002 | 3.928s | 63.227s |
Occupied client slots include protocol completion and final drain. Mean occupancy is weighted by elapsed time; it can be below the requested cap.
| Repetition | Group start | Requested | Effective cap | Peak | Mean |
|---|---|---|---|---|---|
| 1 | 0.000s | C32 | 32 | 32 | 26.17 |
| 2 | 0.000s | C32 | 32 | 32 | 25.42 |
| 3 | 0.000s | C32 | 32 | 32 | 25.08 |
Actual settings for each measured request. Each prompt keeps its temperature across repetitions and engines. Generation uses seed 0 throughout qualification, warmups and measured repetitions; fixed settings do not guarantee identical output text.
| Prompt | Repetition | Temperature |
|---|
| 2239 |
| 61.55ms |
| 1211 gaps for 2239 tokens |
| hf-003 | 3.933s | 105.774s | 4162 | 60.46ms | 2146 gaps for 4162 tokens |
| hf-004 | 3.925s | 87.963s | 4387 | 61.15ms | 1723 gaps for 4387 tokens |
| hf-005 | 4.031s | 116.384s | 5120 | 208.07ms | 2466 gaps for 5120 tokens |
| hf-006 | 3.758s | 100.214s | 5120 | 151.84ms | 2037 gaps for 5120 tokens |
| hf-007 | 3.905s | 114.990s | 4245 | 61.11ms | 2397 gaps for 4245 tokens |
| hf-008 | 3.857s | 97.823s | 4562 | 60.82ms | 1975 gaps for 4562 tokens |
| hf-009 | 3.850s | 75.885s | 3592 | 58.80ms | 1477 gaps for 3592 tokens |
| hf-010 | 3.897s | 86.058s | 4559 | 63.51ms | 1682 gaps for 4559 tokens |
| hf-011 | 3.757s | 63.737s | 3000 | 158.42ms | 1227 gaps for 3000 tokens |
| hf-012 | 3.848s | 107.064s | 4714 | 63.68ms | 2248 gaps for 4714 tokens |
| hf-013 | 3.868s | 104.146s | 3390 | 58.42ms | 2106 gaps for 3390 tokens |
| hf-014 | 4.100s | 83.027s | 4391 | 74.32ms | 1623 gaps for 4391 tokens |
| hf-015 | 3.909s | 94.279s | 4464 | 64.30ms | 1865 gaps for 4464 tokens |
| hf-016 | 3.863s | 86.452s | 4076 | 60.49ms | 1691 gaps for 4076 tokens |
| hf-017 | 4.082s | 118.755s | 5120 | 72.66ms | 2559 gaps for 5120 tokens |
| hf-018 | 4.065s | 77.891s | 3443 | 73.08ms | 1515 gaps for 3443 tokens |
| hf-019 | 4.016s | 96.735s | 3833 | 209.53ms | 1941 gaps for 3833 tokens |
| hf-020 | 4.109s | 120.818s | 5120 | 75.45ms | 2653 gaps for 5120 tokens |
| hf-021 | 4.069s | 110.866s | 5120 | 77.68ms | 2304 gaps for 5120 tokens |
| hf-022 | 3.907s | 94.833s | 5120 | 60.53ms | 1878 gaps for 5120 tokens |
| hf-023 | 3.892s | 102.874s | 4219 | 64.22ms | 2072 gaps for 4219 tokens |
| hf-024 | 4.106s | 74.184s | 3449 | 75.66ms | 1436 gaps for 3449 tokens |
| hf-025 | 4.158s | 112.538s | 4633 | 77.59ms | 2346 gaps for 4633 tokens |
| hf-027 | 4.074s | 114.778s | 4687 | 73.95ms | 2417 gaps for 4687 tokens |
| hf-028 | 4.092s | 122.847s | 5120 | 78.07ms | 2779 gaps for 5120 tokens |
| hf-029 | 3.870s | 116.723s | 5120 | 62.02ms | 2450 gaps for 5120 tokens |
| hf-030 | 4.081s | 120.991s | 5120 | 75.80ms | 2664 gaps for 5120 tokens |
| hf-031 | 4.064s | 119.242s | 5120 | 73.42ms | 2580 gaps for 5120 tokens |
| Top p |
|---|
| Seed |
|---|
| hf-011 | 1 | 0.76435 | 1 | 0 |
| hf-002 | 1 | 0.440527 | 1 | 0 |
| hf-014 | 1 | 0.91642 | 1 | 0 |
| hf-024 | 1 | 0.897191 | 1 | 0 |
| hf-008 | 1 | 0.612374 | 1 | 0 |
| hf-010 | 1 | 0.782959 | 1 | 0 |
| hf-016 | 1 | 0.724321 | 1 | 0 |
| hf-012 | 1 | 0.511421 | 1 | 0 |
| hf-004 | 1 | 0.816071 | 1 | 0 |
| hf-026 | 1 | 0.864741 | 1 | 0 |
| hf-015 | 1 | 0.992364 | 1 | 0 |
| hf-018 | 1 | 0.72616 | 1 | 0 |
| hf-022 | 1 | 0.90875 | 1 | 0 |
| hf-006 | 1 | 0.832691 | 1 | 0 |
| hf-009 | 1 | 1.009705 | 1 | 0 |
| hf-023 | 1 | 0.446593 | 1 | 0 |
| hf-013 | 1 | 0.321389 | 1 | 0 |
| hf-003 | 1 | 0.313504 | 1 | 0 |
| hf-001 | 1 | 0.143471 | 1 | 0 |
| hf-021 | 1 | 0.56603 | 1 | 0 |
| hf-007 | 1 | 0.251762 | 1 | 0 |
| hf-000 | 1 | 0.567334 | 1 | 0 |
| hf-029 | 1 | 0.498641 | 1 | 0 |
| hf-005 | 1 | 0.528863 | 1 | 0 |
| hf-019 | 1 | 0.460991 | 1 | 0 |
| hf-017 | 1 | 0.397818 | 1 | 0 |
| hf-025 | 1 | 0.397412 | 1 | 0 |
| hf-027 | 1 | 0.44245 | 1 | 0 |
| hf-031 | 1 | 0.506728 | 1 | 0 |
| hf-020 | 1 | 0.245154 | 1 | 0 |
| hf-030 | 1 | 0.321942 | 1 | 0 |
| hf-028 | 1 | 0.243526 | 1 | 0 |
| hf-025 | 2 | 0.397412 | 1 | 0 |
| hf-011 | 2 | 0.76435 | 1 | 0 |
| hf-018 | 2 | 0.72616 | 1 | 0 |
| hf-031 | 2 | 0.506728 | 1 | 0 |
| hf-009 | 2 | 1.009705 | 1 | 0 |
| hf-027 | 2 | 0.44245 | 1 | 0 |
| hf-015 | 2 | 0.992364 | 1 | 0 |
| hf-010 | 2 | 0.782959 | 1 | 0 |
| hf-002 | 2 | 0.440527 | 1 | 0 |
| hf-023 | 2 | 0.446593 | 1 | 0 |
| hf-014 | 2 | 0.91642 | 1 | 0 |
| hf-024 | 2 | 0.897191 | 1 | 0 |
| hf-000 | 2 | 0.567334 | 1 | 0 |
| hf-019 | 2 | 0.460991 | 1 | 0 |
| hf-022 | 2 | 0.90875 | 1 | 0 |
| hf-008 | 2 | 0.612374 | 1 | 0 |
| hf-026 | 2 | 0.864741 | 1 | 0 |
| hf-007 | 2 | 0.251762 | 1 | 0 |
| hf-016 | 2 | 0.724321 | 1 | 0 |
| hf-006 | 2 | 0.832691 | 1 | 0 |
| hf-012 | 2 | 0.511421 | 1 | 0 |
| hf-004 | 2 | 0.816071 | 1 | 0 |
| hf-003 | 2 | 0.313504 | 1 | 0 |
| hf-021 | 2 | 0.56603 | 1 | 0 |
| hf-001 | 2 | 0.143471 | 1 | 0 |
| hf-005 | 2 | 0.528863 | 1 | 0 |
| hf-029 | 2 | 0.498641 | 1 | 0 |
| hf-030 | 2 | 0.321942 | 1 | 0 |
| hf-017 | 2 | 0.397818 | 1 | 0 |
| hf-020 | 2 | 0.245154 | 1 | 0 |
| hf-028 | 2 | 0.243526 | 1 | 0 |
| hf-013 | 2 | 0.321389 | 1 | 0 |
| hf-026 | 3 | 0.864741 | 1 | 0 |
| hf-009 | 3 | 1.009705 | 1 | 0 |
| hf-002 | 3 | 0.440527 | 1 | 0 |
| hf-024 | 3 | 0.897191 | 1 | 0 |
| hf-011 | 3 | 0.76435 | 1 | 0 |
| hf-013 | 3 | 0.321389 | 1 | 0 |
| hf-018 | 3 | 0.72616 | 1 | 0 |
| hf-014 | 3 | 0.91642 | 1 | 0 |
| hf-004 | 3 | 0.816071 | 1 | 0 |
| hf-016 | 3 | 0.724321 | 1 | 0 |
| hf-022 | 3 | 0.90875 | 1 | 0 |
| hf-010 | 3 | 0.782959 | 1 | 0 |
| hf-015 | 3 | 0.992364 | 1 | 0 |
| hf-019 | 3 | 0.460991 | 1 | 0 |
| hf-003 | 3 | 0.313504 | 1 | 0 |
| hf-006 | 3 | 0.832691 | 1 | 0 |
| hf-008 | 3 | 0.612374 | 1 | 0 |
| hf-021 | 3 | 0.56603 | 1 | 0 |
| hf-023 | 3 | 0.446593 | 1 | 0 |
| hf-025 | 3 | 0.397412 | 1 | 0 |
| hf-012 | 3 | 0.511421 | 1 | 0 |
| hf-027 | 3 | 0.44245 | 1 | 0 |
| hf-005 | 3 | 0.528863 | 1 | 0 |
| hf-029 | 3 | 0.498641 | 1 | 0 |
| hf-000 | 3 | 0.567334 | 1 | 0 |
| hf-007 | 3 | 0.251762 | 1 | 0 |
| hf-017 | 3 | 0.397818 | 1 | 0 |
| hf-031 | 3 | 0.506728 | 1 | 0 |
| hf-020 | 3 | 0.245154 | 1 | 0 |
| hf-030 | 3 | 0.321942 | 1 | 0 |
| hf-028 | 3 | 0.243526 | 1 | 0 |
| hf-001 | 3 | 0.143471 | 1 | 0 |
role
evidence
request concurrency
requests per s
sla goodput ratio
output tokens per s
p99 e2e ms rel range
p99 itl ms rel range
p99 ttft ms rel range
tier completion max rel range
Image digest