nexus-agents - v2.80.0
GitHub
npm
Preparing search index...
runBenchmark
Function runBenchmark
runBenchmark
<
TInstance
,
TPrediction
,
TEvalResult
>
(
adapter
:
BenchmarkAdapter
<
TInstance
,
TPrediction
,
TEvalResult
>
,
config
:
Record
<
string
,
unknown
>
,
options
?:
BenchmarkOrchestratorOptions
,
)
:
Promise
<
BenchmarkRunSummary
>
Execute one adapter end-to-end. Returns the adapter-produced summary.
Behavioral notes:
An instance failure (either runInstance or evaluate throws) is captured as a failure count in summary metadata; the run continues.
Timeouts cancel via AbortController; adapters should honor
ctx.signal
.
Type Parameters
TInstance
TPrediction
TEvalResult
Parameters
adapter
:
BenchmarkAdapter
<
TInstance
,
TPrediction
,
TEvalResult
>
config
:
Record
<
string
,
unknown
>
options
:
BenchmarkOrchestratorOptions
= {}
Returns
Promise
<
BenchmarkRunSummary
>
Settings
Member Visibility
Protected
Internal
Theme
OS
Light
Dark
GitHub
npm
nexus-agents - v2.80.0
Loading...
Execute one adapter end-to-end. Returns the adapter-produced summary.
Behavioral notes:
ctx.signal.