rowboat/apps/simulation_runner/simulation.py
2025-02-17 23:00:15 +05:30

123 lines
No EOL
4.2 KiB
Python

from rowboat import Client, StatefulChat
from typing import List
import json
import os
from openai import OpenAI
from scenario_types import Scenario, SimulationResult, SimulationAggregateResult
from db import write_simulation_result, set_simulation_run_to_completed
openai_client = OpenAI()
MODEL_NAME = "gpt-4o"
ROWBOAT_API_HOST = os.environ.get("ROWBOAT_API_HOST", "http://127.0.0.1:3000").strip()
def simulate_scenario(scenario: Scenario, rowboat_client: Client, workflow_id: str, max_iterations: int = 5) -> str:
"""
Runs a mock simulation for a given scenario.
After simulating several turns of conversation, it evaluates the conversation.
"""
support_chat = StatefulChat(
rowboat_client,
system_prompt=f"{f'Context: {scenario.context}' if scenario.context else ''}",
workflow_id=workflow_id
)
messages = [
{
"role": "system",
"content": f"Simulate the user based on the scenario: \n {scenario.description}"
}
]
# -------------------------
# 1) MAIN SIMULATION LOOP
# -------------------------
for i in range(max_iterations):
openai_input = messages
simulated_user_response = openai_client.chat.completions.create(
model=MODEL_NAME,
messages=openai_input,
temperature=0.0,
)
simulated_content = simulated_user_response.choices[0].message.content
# Feed the model-generated content back into Rowboat's stateful chat
rowboat_response = support_chat.run(simulated_content)
# Store the user message back into `messages` so the conversation continues
messages.append({"role": "assistant", "content": rowboat_response})
# -------------------------
# 2) EVALUATION STEP
# -------------------------
transcript_str = ""
for m in messages:
role = m.get("role", "unknown")
content = m.get("content", "")
transcript_str += f"{role.upper()}: {content}\n"
evaluation_prompt = [
{
"role": "system",
"content": (
f"You are a neutral evaluator. Evaluate based on these criteria:\n{scenario.criteria}\n\nReturn ONLY a JSON object with format: "
'{"verdict": "pass"} if the support bot answered correctly, or {"verdict": "fail"} if not.'
)
},
{
"role": "user",
"content": (
f"Here is the conversation transcript:\n\n{transcript_str}\n\n"
"Did the support bot answer correctly or not? Return only 'pass' or 'fail'."
)
}
]
eval_response = openai_client.chat.completions.create(
model=MODEL_NAME,
messages=evaluation_prompt,
temperature=0.0,
response_format={"type": "json_object"}
)
if not eval_response.choices:
raise Exception("No evaluation response received from model")
else:
response_json = json.loads(eval_response.choices[0].message.content)
evaluation_result = response_json.get("verdict")
if evaluation_result is None:
raise Exception("No verdict field found in evaluation response")
return(evaluation_result, transcript_str)
async def simulate_scenarios(scenarios: List[Scenario], runId: str, workflow_id: str, api_key: str, max_iterations: int = 5):
project_id = scenarios[0].projectId
client = Client(
host=ROWBOAT_API_HOST,
project_id=project_id,
api_key=api_key
)
results = []
for scenario in scenarios:
result, transcript = simulate_scenario(scenario, client, workflow_id, max_iterations)
simulation_result = SimulationResult(
projectId=project_id,
runId=runId,
scenarioId=scenario.id,
result=result,
details=transcript
)
results.append(simulation_result)
write_simulation_result(simulation_result)
aggregate_result = SimulationAggregateResult(**{
"total": len(scenarios),
"pass": sum(1 for result in results if result.result == "pass"),
"fail": sum(1 for result in results if result.result == "fail")
})
return aggregate_result