From a0d5102bd8a704b5075c2e51fa27ddcdbb59e97e Mon Sep 17 00:00:00 2001
From: Rita Aleksziev <alekszievr@gmail..com>
Date: Tue, 3 Dec 2024 17:22:23 +0100
Subject: [PATCH] add some spaces for readability

---
 evals/llm_as_a_judge.py | 15 +++++++++++++++
 1 file changed, 15 insertions(+)

diff --git a/evals/llm_as_a_judge.py b/evals/llm_as_a_judge.py
index c84795642..4deee7d7d 100644
--- a/evals/llm_as_a_judge.py
+++ b/evals/llm_as_a_judge.py
@@ -37,8 +37,10 @@ async def answer_without_cognee(instance):
 async def answer_with_cognee(instance):
     await cognee.prune.prune_data()
     await cognee.prune.prune_system(metadata=True)
+    
     for (title, sentences) in instance["context"]:
         await cognee.add("\n".join(sentences), dataset_name = "HotPotQA")
+    
     await cognee.cognify("HotPotQA")
 
     search_results = await cognee.search(
@@ -58,11 +60,13 @@ async def answer_with_cognee(instance):
         system_prompt=system_prompt,
         response_model=str,
     )
+    
     return answer_prediction
 
 
 async def eval_answers(instances, answers, eval_metric):
     test_cases = []
+    
     for instance, answer in zip(instances, answers):
         test_case = LLMTestCase(
             input=instance["question"],
@@ -70,19 +74,24 @@ async def eval_answers(instances, answers, eval_metric):
             expected_output=instance["answer"]
         )
         test_cases.append(test_case)
+    
     eval_set = EvaluationDataset(test_cases)
     eval_results = eval_set.evaluate([eval_metric])
+    
     return eval_results
 
 async def eval_on_hotpotQA(answer_provider, num_samples, eval_metric):
     base_config = get_base_config()
     data_root_dir = base_config.data_root_directory
+    
     if not Path(data_root_dir).exists():
         Path(data_root_dir).mkdir()
+    
     filepath = data_root_dir / Path("hotpot_dev_fullwiki_v1.json")
     if not filepath.exists():
         url = 'http://curtis.ml.cmu.edu/datasets/hotpot/hotpot_dev_fullwiki_v1.json'
         wget.download(url, out=data_root_dir)
+    
     with open(filepath, "r") as file:
         dataset = json.load(file)
     
@@ -91,15 +100,19 @@ async def eval_on_hotpotQA(answer_provider, num_samples, eval_metric):
     for instance in tqdm(instances, desc="Getting answers"):
         answer = await answer_provider(instance)
         answers.append(answer)
+    
     eval_results = await eval_answers(instances, answers, eval_metric)
     avg_score = statistics.mean([result.metrics_data[0].score for result in eval_results.test_results])
+    
     return avg_score
 
 if __name__ == "__main__":
     parser = argparse.ArgumentParser()
+    
     parser.add_argument("--with_cognee", action="store_true")
     parser.add_argument("--num_samples", type=int, default=500)
     parser.add_argument("--metric", type=str, default="correctness_metric")
+    
     args = parser.parse_args()
 
     try:
@@ -107,9 +120,11 @@ if __name__ == "__main__":
         metric = metric_cls()
     except AttributeError:
         metric = getattr(evals.deepeval_metrics, args.metric)
+    
     if args.with_cognee:
         answer_provider = answer_with_cognee
     else:
         answer_provider = answer_without_cognee
+    
     avg_score = asyncio.run(eval_on_hotpotQA(answer_provider, args.num_samples, metric))
     print(f"Average {args.metric}: {avg_score}")
\ No newline at end of file