Hotfix/fix quickstart example for query creation

deepeval/dataset.py

@@ -12,6 +12,7 @@
 from deepeval.run_test import run_test
 from deepeval.metrics.metric import Metric
 from deepeval.test_case import LLMTestCase
+from dataclasses import asdict
 
 
 class EvaluationDataset(UserList):
@@ -230,7 +231,7 @@ def from_dict(
         return cls(test_cases)
 
     def to_dict(self):
-        return [x.dict() for x in self.data]
+        return [asdict(x) for x in self.data]
 
     def to_csv(self, csv_filename: str):
         import pandas as pd
@@ -255,7 +256,7 @@ def sample(self, n: int = 5):
         if len(self.data) <= n:
             n = len(self.data)
         result = random.sample(self.data, n)
-        return [r.dict() for r in result]
+        return [asdict(r) for r in result]
 
     def head(self, n: int = 5):
         return self.data[:n]

docs/docs/quickstart/quickstart.md

@@ -87,11 +87,12 @@ Now we often don't want to write our own tests or at least be given a variety of
 You can automatically create tests in DeepEval in just a few lines of code:
 
 ```python
-from deepeval.dataset import create_evaluation_dataset
+from deepeval.dataset import create_evaluation_query_answer_pairs
+
 dataset = create_evaluation_query_answer_pairs(
     openai_api_key="<YOUR_OPENAI_API_KEY>",
     context="FastAPI is a modern, fast (high-performance), web framework for building APIs with Python 3.7+ based on standard Python type hints.",
-    n=3    
+    n=3,
 )
 
 ```

examples/create_tests.py

@@ -0,0 +1,12 @@
+import os
+from deepeval.dataset import (
+    create_evaluation_query_answer_pairs,
+    EvaluationDataset,
+)
+
+dataset: EvaluationDataset = create_evaluation_query_answer_pairs(
+    openai_api_key=os.environ["OPENAI_API_KEY"],
+    context="FastAPI is a modern, fast (high-performance), web framework for building APIs with Python 3.7+ based on standard Python type hints.",
+    n=3,
+)
+dataset.review()

examples/review_datasets.py

@@ -0,0 +1,34 @@
+# Define your completion protocol
+import openai
+from deepeval.dataset import EvaluationDataset
+from deepeval.metrics.factual_consistency import FactualConsistencyMetric
+
+ds = EvaluationDataset.from_csv(
+    "review-test.csv",
+    query_column="query",
+    expected_output_column="expected_output",
+)
+print(ds.sample())
+
+
+def generate_chatgpt_output(query: str):
+    response = openai.ChatCompletion.create(
+        model="gpt-3.5-turbo",
+        messages=[
+            {"role": "system", "content": "You are a helpful assistant."},
+            {
+                "role": "assistant",
+                "content": "The customer success phone line is 1200-231-231 and the customer success state is in Austin.",
+            },
+            {"role": "user", "content": query},
+        ],
+    )
+    expected_output = response.choices[0].message.content
+    return expected_output
+
+
+factual_consistency_metric = FactualConsistencyMetric()
+
+ds.run_evaluation(
+    completion_fn=generate_chatgpt_output, metrics=[factual_consistency_metric]
+)