fix: Improve type conversion for LLM parameters and handle None values properly

Co-Authored-By: Joe Moura <joao@crewai.com>

src/crewai/agent.py

@@ -190,9 +190,71 @@ class Agent(BaseAgent):
                                 if key not in ["prompt", "key_name", "default"]:
                                     # Only add default if the key is already set in os.environ
                                     if key in os.environ:
-                                        llm_params[key] = value
+                                        try:
+                                            # Create a new dictionary for properly typed parameters
+                                            typed_params = {}
+                                            
+                                            # Convert and validate values based on parameter type
+                                            if key in ['temperature', 'top_p', 'presence_penalty', 'frequency_penalty']:
+                                                if value is not None:
+                                                    try:
+                                                        typed_params[key] = float(value)
+                                                    except (ValueError, TypeError):
+                                                        pass
+                                            elif key in ['n', 'max_tokens', 'max_completion_tokens', 'seed']:
+                                                if value is not None:
+                                                    try:
+                                                        typed_params[key] = int(value)
+                                                    except (ValueError, TypeError):
+                                                        pass
+                                            elif key == 'logit_bias' and isinstance(value, str):
+                                                try:
+                                                    bias_dict = {}
+                                                    for pair in value.split(','):
+                                                        token_id, bias = pair.split(':')
+                                                        bias_dict[int(token_id.strip())] = float(bias.strip())
+                                                    typed_params[key] = bias_dict
+                                                except (ValueError, AttributeError):
+                                                    pass
+                                            elif key == 'response_format' and isinstance(value, str):
+                                                try:
+                                                    import json
+                                                    typed_params[key] = json.loads(value)
+                                                except json.JSONDecodeError:
+                                                    pass
+                                            elif key == 'logprobs':
+                                                if value is not None:
+                                                    typed_params[key] = bool(value.lower() == 'true') if isinstance(value, str) else bool(value)
+                                            elif key == 'callbacks':
+                                                typed_params[key] = [] if value is None else [value] if isinstance(value, str) else value
+                                            elif key == 'stop':
+                                                typed_params[key] = [value] if isinstance(value, str) else value
+                                            elif key in ['model', 'base_url', 'api_version', 'api_key']:
+                                                typed_params[key] = value
+                                            
+                                            # Update llm_params with properly typed values
+                                            if typed_params:
+                                                llm_params.update(typed_params)
+                                        except (ValueError, AttributeError, json.JSONDecodeError):
+                                            continue
 
-            self.llm = LLM(**llm_params)
+            # Create LLM instance with properly typed parameters
+            valid_params = {
+                'model', 'timeout', 'temperature', 'top_p', 'n', 'stop',
+                'max_completion_tokens', 'max_tokens', 'presence_penalty',
+                'frequency_penalty', 'logit_bias', 'response_format',
+                'seed', 'logprobs', 'top_logprobs', 'base_url',
+                'api_version', 'api_key', 'callbacks'
+            }
+            
+            # Filter out None values and invalid parameters
+            filtered_params = {}
+            for k, v in llm_params.items():
+                if k in valid_params and v is not None:
+                    filtered_params[k] = v
+            
+            # Create LLM instance with properly typed parameters
+            self.llm = LLM(**filtered_params)
         else:
             # For any other type, attempt to extract relevant attributes
             llm_params = {

tests/crew_test.py

@@ -642,9 +642,10 @@ def test_task_tools_override_agent_tools():
     crew.kickoff()
 
     # Verify task tools override agent tools
-    assert len(task.tools) == 1  # AnotherTestTool
-    assert any(isinstance(tool, AnotherTestTool) for tool in task.tools)
-    assert not any(isinstance(tool, TestTool) for tool in task.tools)
+    tools = task.tools or []
+    assert len(tools) == 1  # AnotherTestTool
+    assert any(isinstance(tool, AnotherTestTool) for tool in tools)
+    assert not any(isinstance(tool, TestTool) for tool in tools)
 
     # Verify agent tools remain unchanged
     assert len(new_researcher.tools) == 1