Actions: Azure/azure-sdk-for-python
Actions
351 workflow run results
351 workflow run results
outputs = await _SafetyEvaluation_instance( evaluators=[evaluator], //"content_safety,protected_material,indirect_attack,direct_attack,eci" evaluation_name="safety evaluation", target=conversation_generator.get_llm_reply, // call back num_turns=1, num_rows=200, concurrent_async_tasks=30, output_path="evaluation_outputs_safety.jsonl", randomization_seed=1 ) We are using the same randomization_seed to get the sin...
Copilot coding agent
#149:
by
Copilot
AI