{% extends "test/base.html" %} {% set active_test_page = "run" %} {% block title %}Test — Run | SPIKEE{% endblock %} {% block content %}
Test Dataset
{# ================================================================ TARGET + DATASET ================================================================ #}
Target & Dataset
{# Target #}
Built-in or workspace target module.
{# Target options #}
e.g. model=openai/gpt-4o, temperature=0
{# Datasets (multi-select) #}
{% if datasets %} {% for ds in datasets %}
{% endfor %} {% else %} No datasets found in datasets/. {% endif %}
{# Judge options — own row #}
e.g. offline, model=bedrock/claude45-haiku
{# ================================================================ OPTIONS ================================================================ #}
Options
0–1
Fraction of dataset to test.
Reproducibility seed.
How to handle existing results files.
Appended to results filename.
{# ================================================================ EXECUTION ================================================================ #}
Execution
Parallel workers.
Per-entry attempts.
On API errors.
Delay between requests.
Attack (optional — runs if standard attempts fail)
{# Left col: module picker #}
Loading attacks…
Runs if standard attempts fail.
{# Right col: iterations + options + attack only, stacked #}
Number of attack iterations.
Key=value options passed to the attack module.
{# ================================================================ SUBMIT ================================================================ #}
{% endblock %}