Round score
0.014
weighted tier speedup after failure cap and deduction
Eligible requests
28
trusted baseline exclusions removed
Failed eligible requests
10
each request counted once
Raw weighted speedup
+4.97%
before failure cap and deduction
Temperature range
Request concurrency
Tier scheduling
Output policy
Thinking
Context limit
The same eligible inputs and output ceilings apply to every engine. Candidates must emit at least 90% of baseline tokens per request; natural output lengths can affect tier completion time. Slots refill until the group drains and stay occupied through protocol completion. Baseline exclusions and final drain can reduce actual concurrency below the configured cap.
Raw weighted speedup
Eligible speedup
Request failures cap speed credit at zero: min(0.049709, 0) = 0.049709. Negative speedups remain negative.
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.
28 prompts
| Prompt | Diagnostic speedup | Baseline | Candidate | Aligned output | Diagnostic / failure |
|---|---|---|---|---|---|
| hf-026 | — | — | — | 2340 | Per-token timing unavailable (not a request failure) |
| hf-001 | — | — | — | 2820 | Per-token timing unavailable (not a request failure) |
| hf-022 | — | — | 70.044s |
28 requests
| Request | TTFT | Total | Tokens | Slowest gap | Stream |
|---|---|---|---|---|---|
| hf-001 | 664.51ms | 78.428s | 5120 | 38.87ms | 2701 gaps for 5120 tokens |
| hf-002 | 582.28ms | 80.022s | 5120 | 100.21ms | 2510 gaps for 5120 tokens |
| hf-003 | 667.62ms |
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 | 28 | 28 | 20.97 |
| 2 | 0.000s | C32 | 28 | 28 | 20.57 |
| 3 | 0.000s | C32 | 28 | 28 | 21.93 |
Each latency-median response is compared with the lowest valid opening-baseline ratio for the same prompt. Drops above the recorded limit fail; absolute checks also apply. These checks detect repetition, not overall writing quality.
| Prompt | Candidate char-16 |
|---|
Failed requests
Deduction
Final score
Each duration is the median across repetitions, from the group start until the tier's last request finishes. This includes client queueing. Weighted speedup = sum of weight × (1 − candidate / baseline).
| Input tier | Eligible requests | Weight | Baseline | Candidate | Speedup | Contribution |
|---|---|---|---|---|---|---|
| 2k | 6 | 25% | 103.068s | 92.708s | +10.05% | +2.51% |
| 4k | 7 | 25% | 104.854s | 88.874s | +15.24% | +3.81% |
| 8k | 7 | 25% | 101.621s | 99.202s | +2.38% | +0.59% |
| 16k | 8 | 25% | 99.499s | 107.247s | -7.79% | -1.95% |
| 2132 |
| Per-token timing unavailable (not a request failure) |
| hf-002 | — | — | — | 3187 | Per-token timing unavailable (not a request failure) |
| hf-008 | — | — | — | 5120 | Per-token timing unavailable (not a request failure) |
| hf-031 | — | — | — | 1998 | candidate output below tolerance |
| hf-024 | — | — | — | 3845 | Per-token timing unavailable (not a request failure) |
| hf-009 | — | — | — | 2081 | candidate output below tolerance |
| hf-011 | — | — | — | 3482 | Per-token timing unavailable (not a request failure) |
| hf-010 | — | — | — | 2745 | Per-token timing unavailable (not a request failure) |
| hf-005 | — | — | — | 4448 | Per-token timing unavailable (not a request failure) |
| hf-018 | — | — | — | 4258 | Per-token timing unavailable (not a request failure) |
| hf-006 | — | — | — | 3462 | candidate output below tolerance |
| hf-016 | — | — | — | 2624 | candidate output below tolerance |
| hf-029 | — | — | — | 2867 | candidate output below tolerance |
| hf-014 | — | — | — | 5120 | Per-token timing unavailable (not a request failure) |
| hf-020 | — | — | — | 2104 | candidate output below tolerance |
| hf-027 | — | — | — | 2880 | candidate output below tolerance |
| hf-023 | — | — | — | 4400 | Per-token timing unavailable (not a request failure) |
| hf-003 | — | — | — | 5120 | Per-token timing unavailable (not a request failure) |
| hf-019 | — | — | — | 5120 | Per-token timing unavailable (not a request failure) |
| hf-015 | — | — | — | 4170 | candidate output below tolerance |
| hf-012 | — | — | — | 5120 | Per-token timing unavailable (not a request failure) |
| hf-025 | — | — | — | 4162 | Per-token timing unavailable (not a request failure) |
| hf-021 | — | — | — | 4170 | candidate output below tolerance |
| hf-030 | — | — | — | 5120 | Per-token timing unavailable (not a request failure) |
| hf-013 | — | — | — | 4491 | Per-token timing unavailable (not a request failure) |
| hf-007 | — | — | — | 4248 | candidate output below tolerance |
| 81.585s |
| 5120 |
| 37.43ms |
| 2678 gaps for 5120 tokens |
| hf-005 | 661.56ms | 75.696s | 4727 | 39.00ms | 2534 gaps for 4727 tokens |
| hf-006 | 662.33ms | 63.988s | 3462 | 38.95ms | 1919 gaps for 3462 tokens |
| hf-007 | 662.84ms | 70.361s | 4248 | 39.22ms | 2240 gaps for 4248 tokens |
| hf-008 | 665.29ms | 82.504s | 5120 | 39.16ms | 2996 gaps for 5120 tokens |
| hf-009 | 660.17ms | 36.799s | 2081 | 37.48ms | 1067 gaps for 2081 tokens |
| hf-010 | 657.75ms | 66.852s | 4695 | 37.46ms | 2054 gaps for 4695 tokens |
| hf-011 | 665.37ms | 73.172s | 4596 | 37.42ms | 2238 gaps for 4596 tokens |
| hf-012 | 659.75ms | 82.997s | 5120 | 37.53ms | 2671 gaps for 5120 tokens |
| hf-013 | 654.78ms | 87.458s | 4777 | 37.54ms | 2983 gaps for 4777 tokens |
| hf-014 | 659.24ms | 80.985s | 5120 | 39.29ms | 2872 gaps for 5120 tokens |
| hf-015 | 656.42ms | 80.893s | 4170 | 37.42ms | 2554 gaps for 4170 tokens |
| hf-016 | 660.21ms | 54.959s | 2624 | 39.63ms | 1630 gaps for 2624 tokens |
| hf-018 | 662.10ms | 76.048s | 5120 | 39.43ms | 2555 gaps for 5120 tokens |
| hf-019 | 661.50ms | 84.342s | 5120 | 37.41ms | 2849 gaps for 5120 tokens |
| hf-020 | 661.16ms | 45.897s | 2104 | 37.43ms | 1337 gaps for 2104 tokens |
| hf-021 | 662.02ms | 80.788s | 4170 | 37.41ms | 2633 gaps for 4170 tokens |
| hf-022 | 658.80ms | 86.835s | 5120 | 37.40ms | 3034 gaps for 5120 tokens |
| hf-023 | 655.98ms | 78.463s | 5120 | 39.96ms | 2704 gaps for 5120 tokens |
| hf-024 | 653.73ms | 68.962s | 4000 | 40.03ms | 2166 gaps for 4000 tokens |
| hf-025 | 651.32ms | 85.534s | 5120 | 40.11ms | 3302 gaps for 5120 tokens |
| hf-026 | 652.18ms | 48.255s | 2388 | 40.05ms | 1420 gaps for 2388 tokens |
| hf-027 | 657.96ms | 47.084s | 2880 | 37.40ms | 1374 gaps for 2880 tokens |
| hf-029 | 656.93ms | 48.440s | 2867 | 37.41ms | 1417 gaps for 2867 tokens |
| hf-030 | 658.31ms | 86.568s | 5120 | 37.40ms | 3012 gaps for 5120 tokens |
| hf-031 | 654.57ms | 35.657s | 1998 | 40.01ms | 1036 gaps for 1998 tokens |
| Baseline char-16 |
|---|
| Drop |
|---|
| Limit |
|---|
| Result |
|---|
| hf-001 | 0.9866 | 0.9203 | -0.0664 | 0.1000 | Passed |
| hf-002 | 0.9455 | 0.9763 | 0.0309 | 0.1000 | Passed |
| hf-003 | 0.9630 | 0.8336 | -0.1294 | 0.1000 | Passed |
| hf-005 | 0.9716 | 0.9704 | -0.0012 | 0.1000 | Passed |
| hf-006 | 0.9876 | 0.9593 | -0.0283 | 0.1000 | Passed |
| hf-007 | 0.9807 | 0.7025 | -0.2782 | 0.1000 | Passed |
| hf-008 | 0.9746 | 0.7725 | -0.2021 | 0.1000 | Passed |
| hf-009 | 0.9678 | 0.9725 | 0.0046 | 0.1000 | Passed |
| hf-010 | 0.9532 | 0.9876 | 0.0345 | 0.1000 | Passed |
| hf-011 | 0.9899 | 0.9432 | -0.0467 | 0.1000 | Passed |
| hf-012 | 0.9688 | 0.9591 | -0.0097 | 0.1000 | Passed |
| hf-013 | 0.9660 | 0.9490 | -0.0170 | 0.1000 | Passed |
| hf-014 | 0.9887 | 0.9603 | -0.0284 | 0.1000 | Passed |
| hf-015 | 0.9929 | 0.9432 | -0.0497 | 0.1000 | Passed |
| hf-016 | 0.9943 | 0.9732 | -0.0211 | 0.1000 | Passed |
| hf-018 | 0.9675 | 0.5820 | -0.3854 | 0.1000 | Passed |
| hf-019 | 0.9702 | 0.9726 | 0.0024 | 0.1000 | Passed |
| hf-020 | 0.9742 | 0.9137 | -0.0605 | 0.1000 | Passed |
| hf-021 | 0.9786 | 0.9389 | -0.0397 | 0.1000 | Passed |
| hf-022 | 0.9598 | 0.9048 | -0.0550 | 0.1000 | Passed |
| hf-023 | 0.8757 | 0.8942 | 0.0184 | 0.1000 | Passed |
| hf-024 | 0.9939 | 0.9312 | -0.0627 | 0.1000 | Passed |
| hf-025 | 0.9526 | 0.9246 | -0.0279 | 0.1000 | Passed |
| hf-026 | 0.9248 | 0.9627 | 0.0379 | 0.1000 | Passed |
| hf-027 | 0.9389 | 0.8809 | -0.0580 | 0.1000 | Passed |
| hf-029 | 0.9554 | 0.9088 | -0.0466 | 0.1000 | Passed |
| hf-030 | 0.9681 | 0.9677 | -0.0004 | 0.1000 | Passed |
| hf-031 | 0.9557 | 0.8850 | -0.0707 | 0.1000 | Passed |
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 | Top p | Seed |
|---|---|---|---|---|
| hf-007 | 1 | 0.714609 | 1 | 0 |
| hf-018 | 1 | 0.737151 | 1 | 0 |
| hf-031 | 1 | 0.776567 | 1 | 0 |
| hf-016 | 1 | 0.996831 | 1 | 0 |
| hf-020 | 1 | 0.854241 | 1 | 0 |
| hf-027 | 1 | 0.911383 | 1 | 0 |
| hf-029 | 1 | 0.752413 | 1 | 0 |
| hf-005 | 1 | 0.663059 | 1 | 0 |
| hf-006 | 1 | 0.852864 | 1 | 0 |
| hf-002 | 1 | 0.711992 | 1 | 0 |
| hf-023 | 1 | 0.502734 | 1 | 0 |
| hf-011 | 1 | 0.8957 | 1 | 0 |
| hf-024 | 1 | 0.995315 | 1 | 0 |
| hf-014 | 1 | 0.690995 | 1 | 0 |
| hf-001 | 1 | 0.732727 | 1 | 0 |
| hf-021 | 1 | 0.295024 | 1 | 0 |
| hf-003 | 1 | 0.506793 | 1 | 0 |
| hf-009 | 1 | 0.970461 | 1 | 0 |
| hf-015 | 1 | 0.468658 | 1 | 0 |
| hf-012 | 1 | 0.444494 | 1 | 0 |
| hf-019 | 1 | 0.450884 | 1 | 0 |
| hf-025 | 1 | 0.119657 | 1 | 0 |
| hf-008 | 1 | 0.391391 | 1 | 0 |
| hf-030 | 1 | 0.282037 | 1 | 0 |
| hf-022 | 1 | 0.407771 | 1 | 0 |
| hf-013 | 1 | 0.234207 | 1 | 0 |
| hf-026 | 1 | 0.258588 | 1 | 0 |
| hf-010 | 1 | 0.938026 | 1 | 0 |
| hf-027 | 2 | 0.911383 | 1 | 0 |
| hf-031 | 2 | 0.776567 | 1 | 0 |
| hf-026 | 2 | 0.258588 | 1 | 0 |
| hf-001 | 2 | 0.732727 | 1 | 0 |
| hf-021 | 2 | 0.295024 | 1 | 0 |
| hf-009 | 2 | 0.970461 | 1 | 0 |
| hf-011 | 2 | 0.8957 | 1 | 0 |
| hf-020 | 2 | 0.854241 | 1 | 0 |
| hf-024 | 2 | 0.995315 | 1 | 0 |
| hf-008 | 2 | 0.391391 | 1 | 0 |
| hf-010 | 2 | 0.938026 | 1 | 0 |
| hf-002 | 2 | 0.711992 | 1 | 0 |
| hf-029 | 2 | 0.752413 | 1 | 0 |
| hf-015 | 2 | 0.468658 | 1 | 0 |
| hf-007 | 2 | 0.714609 | 1 | 0 |
| hf-014 | 2 | 0.690995 | 1 | 0 |
| hf-006 | 2 | 0.852864 | 1 | 0 |
| hf-012 | 2 | 0.444494 | 1 | 0 |
| hf-005 | 2 | 0.663059 | 1 | 0 |
| hf-003 | 2 | 0.506793 | 1 | 0 |
| hf-018 | 2 | 0.737151 | 1 | 0 |
| hf-023 | 2 | 0.502734 | 1 | 0 |
| hf-019 | 2 | 0.450884 | 1 | 0 |
| hf-013 | 2 | 0.234207 | 1 | 0 |
| hf-016 | 2 | 0.996831 | 1 | 0 |
| hf-022 | 2 | 0.407771 | 1 | 0 |
| hf-030 | 2 | 0.282037 | 1 | 0 |
| hf-025 | 2 | 0.119657 | 1 | 0 |
| hf-029 | 3 | 0.752413 | 1 | 0 |
| hf-009 | 3 | 0.970461 | 1 | 0 |
| hf-031 | 3 | 0.776567 | 1 | 0 |
| hf-026 | 3 | 0.258588 | 1 | 0 |
| hf-016 | 3 | 0.996831 | 1 | 0 |
| hf-022 | 3 | 0.407771 | 1 | 0 |
| hf-019 | 3 | 0.450884 | 1 | 0 |
| hf-030 | 3 | 0.282037 | 1 | 0 |
| hf-020 | 3 | 0.854241 | 1 | 0 |
| hf-006 | 3 | 0.852864 | 1 | 0 |
| hf-010 | 3 | 0.938026 | 1 | 0 |
| hf-027 | 3 | 0.911383 | 1 | 0 |
| hf-003 | 3 | 0.506793 | 1 | 0 |
| hf-024 | 3 | 0.995315 | 1 | 0 |
| hf-007 | 3 | 0.714609 | 1 | 0 |
| hf-011 | 3 | 0.8957 | 1 | 0 |
| hf-005 | 3 | 0.663059 | 1 | 0 |
| hf-018 | 3 | 0.737151 | 1 | 0 |
| hf-001 | 3 | 0.732727 | 1 | 0 |
| hf-023 | 3 | 0.502734 | 1 | 0 |
| hf-015 | 3 | 0.468658 | 1 | 0 |
| hf-002 | 3 | 0.711992 | 1 | 0 |
| hf-014 | 3 | 0.690995 | 1 | 0 |
| hf-012 | 3 | 0.444494 | 1 | 0 |
| hf-021 | 3 | 0.295024 | 1 | 0 |
| hf-008 | 3 | 0.391391 | 1 | 0 |
| hf-013 | 3 | 0.234207 | 1 | 0 |
| hf-025 | 3 | 0.119657 | 1 | 0 |
verdict
evidence
min logprob
num prompts
mean logprob
coverage ratio
quantile logprob
num positions scored
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
Patch hash