91 lines
3 KiB
Python
91 lines
3 KiB
Python
import importlib.util
|
|
from pathlib import Path
|
|
|
|
import pytest
|
|
from datasets import Dataset
|
|
from evaluations.evaluators import LLMJudge
|
|
|
|
from haiku.rag.agents.qa.agent import QuestionAnswerAgent
|
|
from haiku.rag.client import HaikuRAG
|
|
from haiku.rag.config.models import ModelConfig
|
|
|
|
HAS_ANTHROPIC = importlib.util.find_spec("anthropic") is not None
|
|
|
|
|
|
@pytest.fixture(scope="module")
|
|
def vcr_cassette_dir():
|
|
return str(Path(__file__).parent.parent.parent / "cassettes" / "test_qa")
|
|
|
|
|
|
@pytest.mark.vcr()
|
|
async def test_qa_ollama(allow_model_requests, qa_corpus: Dataset, temp_db_path):
|
|
"""Test Ollama QA with LLM judge (VCR recorded)."""
|
|
client = HaikuRAG(temp_db_path, create=True)
|
|
qa = QuestionAnswerAgent(
|
|
client, ModelConfig(provider="ollama", name="gpt-oss", enable_thinking=True)
|
|
)
|
|
llm_judge = LLMJudge()
|
|
|
|
doc = qa_corpus[1]
|
|
await client.create_document(
|
|
content=doc["document_extracted"], uri=doc["document_id"]
|
|
)
|
|
|
|
question = doc["question"]
|
|
expected_answer = doc["answer"]
|
|
|
|
answer, _ = await qa.answer(question)
|
|
is_equivalent = await llm_judge.judge_answers(question, answer, expected_answer)
|
|
|
|
assert is_equivalent, (
|
|
f"Generated answer not equivalent to expected answer.\nQuestion: {question}\nGenerated: {answer}\nExpected: {expected_answer}"
|
|
)
|
|
|
|
|
|
@pytest.mark.vcr()
|
|
async def test_qa_openai(allow_model_requests, qa_corpus: Dataset, temp_db_path):
|
|
"""Test OpenAI QA with LLM judge (VCR recorded)."""
|
|
client = HaikuRAG(temp_db_path, create=True)
|
|
qa = QuestionAnswerAgent(client, ModelConfig(provider="openai", name="gpt-4o-mini"))
|
|
llm_judge = LLMJudge()
|
|
|
|
doc = qa_corpus[1]
|
|
await client.create_document(
|
|
content=doc["document_extracted"], uri=doc["document_id"]
|
|
)
|
|
|
|
question = doc["question"]
|
|
expected_answer = doc["answer"]
|
|
|
|
answer, _ = await qa.answer(question)
|
|
is_equivalent = await llm_judge.judge_answers(question, answer, expected_answer)
|
|
|
|
assert is_equivalent, (
|
|
f"Generated answer not equivalent to expected answer.\nQuestion: {question}\nGenerated: {answer}\nExpected: {expected_answer}"
|
|
)
|
|
|
|
|
|
@pytest.mark.vcr()
|
|
@pytest.mark.skipif(not HAS_ANTHROPIC, reason="Anthropic not installed")
|
|
async def test_qa_anthropic(allow_model_requests, qa_corpus: Dataset, temp_db_path):
|
|
"""Test Anthropic QA with LLM judge (VCR recorded)."""
|
|
client = HaikuRAG(temp_db_path, create=True)
|
|
qa = QuestionAnswerAgent(
|
|
client, ModelConfig(provider="anthropic", name="claude-3-5-haiku-20241022")
|
|
)
|
|
llm_judge = LLMJudge()
|
|
|
|
doc = qa_corpus[1]
|
|
await client.create_document(
|
|
content=doc["document_extracted"], uri=doc["document_id"]
|
|
)
|
|
|
|
question = doc["question"]
|
|
expected_answer = doc["answer"]
|
|
|
|
answer, _ = await qa.answer(question)
|
|
is_equivalent = await llm_judge.judge_answers(question, answer, expected_answer)
|
|
|
|
assert is_equivalent, (
|
|
f"Generated answer not equivalent to expected answer.\nQuestion: {question}\nGenerated: {answer}\nExpected: {expected_answer}"
|
|
)
|