arm
se_gemma_l_pubtextcw_n100_v1 · pair
One config on summeval_100, scored one way; its repeats averaged per item.
- dataset
- summeval_100
- scored against
- gold/summeval_100 gold
- goal
- A scorer that agrees with experts on summaries (SummEval)
- runs
- 1
- headline
- tau 0.540
- last run
- 2026-10-19 07:22
Inspect itAgainst the baselineWith the 35 arms scored the same way
01metrics
| metric | mean | items | better |
|---|---|---|---|
| tau | 0.540 | 100 | higher |
| cost_usd | $0.000276 | 100 | lower |
| latency_ms | 3268.1 | 100 | lower |
| parsed | 1.000 | 100 | neutral |
| reasoning_tokens | 0.000 | 100 | neutral |
| rho | 0.598 | 100 | higher |
| rho_coherence | 0.654 | 100 | higher |
| rho_consistency | 0.636 | 100 | higher |
| rho_fluency | 0.539 | 100 | higher |
| rho_relevance | 0.564 | 100 | higher |
| tau_coherence | 0.562 | 100 | higher |
| tau_consistency | 0.610 | 100 | higher |
| tau_fluency | 0.499 | 100 | higher |
| tau_relevance | 0.487 | 100 | higher |
| tie_rate | 0.415 | 100 | neutral |
| tokens_in | 2023.1 | 100 | lower |
| tokens_out | 217.9 | 100 | lower |
| truncated | 0.000 | 100 | neutral |
02cost and speed
- cost per item
- $0.000276 · 2,241 tok
- speed per item
- p50 2,639 · p95 6,340 ms
Per item, over its items: cost as dollars and tokens (a local model is free but still uses tokens), speed as the median and the slow tail (p95). A goal that names this arm weighs them in its decision.
03runs
| run | made | items | cost | how |
|---|---|---|---|---|
se_gemma_l_pubtextcw_n100_v1_20261019T071647Z | 2026-10-19 07:22 | 100 | $0.027640 | measured |
04config
config se_gemma_l_pubtextcw_n100_v1 · adapter examples/summeval-judging/adapter.py · credentials are never shown
- family
gemma- inputs
published- max_tokens
1500- model
eval-gemma-31b- price_tier
large- prompt_file
prompt.txt- provider
litellm- provider_routing
{"allow_fallbacks": false, "order": ["coreweave/fp4"]}- reasoning
{"enabled": false}- temperature
0.0- usd_per_mtok_in
0.1- usd_per_mtok_out
0.2
05conditions
- data.items_sha256
b59999b281b2cbe87c57e4af34cc475c01ba1d531c7b77cb93a136665119c613- deployment.model
eval-gemma-31b- deployment.source
adapter- deployment.upstream
openrouter/google/gemma-4-31b-it- fingerprint.version
5- library.numpy
1.26.4- library.openai
3.31.0- library.rouge-score
0.1.2- library.torch
2.14.1- library.transformers
4.55.4- machine.cpu_count
32- machine.label
omarchy- machine.platform
Linux/x86_64- machine.python
3.11.16- machine.release
7.2.5-3-omarchy- reference.sha256
7b5e618e23b7f2b222665d83e99006eb0e9a51066e63e2e5ff8c51ae3c696c21- scorer.id
examples/summeval-judging/adapter.py- scorer.sha256
52000c5a06f5324e6803f9db340fab51cc132c7c671706f0df5f5fa06e513785
06rounds
- Round r2 of A scorer that agrees with experts on summaries (SummEval), closed: retired in it.