Fix type annotation errors in llm_generator.py (#74)

Added explicit list[DatasetCase] type annotations to the cases variable in both LlamaDatasetGenerator.generate() and OpenAIDatasetGenerator.generate() methods to satisfy Mypy type checking.

Closes #67

Author: hargunsiingh

src/agentunit/generators/llm_generator.py

@@ -141,7 +141,7 @@ async def generate(
 
             cases_data = json.loads(response_text)
 
-            cases = []
+            cases: list[DatasetCase] = []
             for case_data in cases_data:
                 case = DatasetCase(
                     id=case_data.get("id", f"generated_{len(cases)}"),
@@ -288,7 +288,7 @@ async def generate(
 
             cases_data = json.loads(response_text)
 
-            cases = []
+            cases: list[DatasetCase] = []
             for case_data in cases_data:
                 case = DatasetCase(
                     id=case_data.get("id", f"generated_{len(cases)}"),