refactor(async): remove unused global evaluation task cache

We now track and gather tasks per loop. The global cache is no longer
read and is not needed. The cache posed cross loop task management
risks, which is why we have moved away from it.

- Delete `global_evaluation_tasks` singleton and related imports.
- Stop appending to the global cache in `EvaluationDataset.evaluate()`;
  still coerce the task to ensure validation/side effects run.
- Remove finalizers that cleared the global cache in
  `execute.py` looped paths.

Author: BloggerBust

deepeval/dataset/dataset.py

@@ -49,7 +49,6 @@
 from deepeval.test_run import (
     global_test_run_manager,
 )
-from deepeval.dataset.types import global_evaluation_tasks
 from deepeval.openai.utils import openai_test_case_pairs
 from deepeval.tracing import trace_manager
 from deepeval.tracing.tracing import EVAL_DUMMY_SPAN_NAME
@@ -1238,7 +1237,7 @@ def evals_iterator(
                 )
 
     def evaluate(self, task: Task):
-        global_evaluation_tasks.append(coerce_to_task(task))
+        coerce_to_task(task)
 
     def _start_otel_test_run(self, tracer: Optional[Tracer] = None) -> Context:
         _tracer = check_tracer(tracer)

deepeval/dataset/types.py

@@ -23,6 +23,3 @@ def clear_tasks(self) -> None:
             if not t.done():
                 t.cancel()
         self._tasks.clear()
-
-
-global_evaluation_tasks = EvaluationTasks()

deepeval/evaluate/execute.py

@@ -42,7 +42,6 @@
     BaseApiSpan,
 )
 from deepeval.dataset import Golden
-from deepeval.dataset.types import global_evaluation_tasks
 from deepeval.errors import MissingTestCaseParamsError
 from deepeval.metrics.utils import copy_metrics
 from deepeval.utils import (
@@ -1719,10 +1718,6 @@ def dfs(
         local_trace_manager.evaluating = False
         local_trace_manager.traces_to_evaluate_order.clear()
         local_trace_manager.traces_to_evaluate.clear()
-        try:
-            global_evaluation_tasks.clear_tasks()
-        except Exception:
-            pass
 
 
 def a_execute_agentic_test_cases_from_loop(
@@ -2069,10 +2064,6 @@ def on_task_done(t: asyncio.Task):
         local_trace_manager.evaluating = False
         local_trace_manager.traces_to_evaluate_order.clear()
         local_trace_manager.traces_to_evaluate.clear()
-        try:
-            global_evaluation_tasks.clear_tasks()
-        except Exception:
-            pass
 
 
 async def _a_evaluate_traces(