simplified COT and rebased main

Author: rmenziejr

sdks/python/src/opik/evaluation/metrics/llm_judges/g_eval/metric.py

@@ -1,6 +1,5 @@
 from typing import Any, Optional, Union
 import pydantic
-import asyncstdlib as a
 
 from opik.evaluation.metrics import base_metric, score_result
 from opik.evaluation.models import base_model, models_factory
@@ -52,11 +51,14 @@ def __init__(
         self._chain_of_thought_response = ""
 
     def llm_chain_of_thought(self) -> str:
-        prompt = template.G_EVAL_COT_TEMPLATE.format(
-            task_introduction=self.task_introduction,
-            evaluation_criteria=self.evaluation_criteria,
-        )
-        return self._model.generate_string(input=prompt)
+        if not self._chain_of_thought_response:
+            prompt = template.G_EVAL_COT_TEMPLATE.format(
+                task_introduction=self.task_introduction,
+                evaluation_criteria=self.evaluation_criteria,
+            )
+            self._chain_of_thought_response=self._model.generate_string(input=prompt)
+
+        return self._chain_of_thought_response
 
 
     async def allm_chain_of_thought(self) -> str:
@@ -144,10 +146,6 @@ async def ascore(
         llm_query = template.G_EVAL_QUERY_TEMPLATE.format(
             task_introduction=self.task_introduction,
             evaluation_criteria=self.evaluation_criteria,
-<<<<<<< HEAD
-
-=======
->>>>>>> 857e26be (COT simplified)
             chain_of_thought=await self.allm_chain_of_thought(),
             input=output,
         )
@@ -170,4 +168,4 @@ async def ascore(
             content=model_output,
             name=self.name,
             log_probs_supported=self._log_probs_supported,
-        )
+        )