deepseek Swarm parallel tool calls

[LoveIsLive]

My Deepseek big model returns two tool_calls, and parallel_tool_calls=False has no effect

Code from https://langchain-ai.github.io/langgraph/agents/multi-agent/#handoffs

from langchain_core.tools import tool, InjectedToolCallId
from langgraph.prebuilt import create_react_agent, InjectedState
from langgraph.graph import StateGraph, START, MessagesState
from langgraph.types import Command

def create_handoff_tool(*, agent_name: str, description: str | None = None):
    name = f"transfer_to_{agent_name}"
    description = description or f"Transfer to {agent_name}"

    @tool(name, description=description)
    def handoff_tool(
        state: Annotated[MessagesState, InjectedState], 
        tool_call_id: Annotated[str, InjectedToolCallId],
    ) -> Command:
        tool_message = {
            "role": "tool",
            "content": f"Successfully transferred to {agent_name}",
            "name": name,
            "tool_call_id": tool_call_id,
        }
        return Command(  
            goto=agent_name,  
            update={"messages": state["messages"] + [tool_message]},  
            graph=Command.PARENT,  
        )
    return handoff_tool

# Handoffs
transfer_to_hotel_assistant = create_handoff_tool(
    agent_name="hotel_assistant",
    description="Transfer user to the hotel-booking assistant.",
)
transfer_to_flight_assistant = create_handoff_tool(
    agent_name="flight_assistant",
    description="Transfer user to the flight-booking assistant.",
)

# Simple agent tools
def book_hotel(hotel_name: str):
    """Book a hotel"""
    return f"Successfully booked a stay at {hotel_name}."

def book_flight(from_airport: str, to_airport: str):
    """Book a flight"""
    return f"Successfully booked a flight from {from_airport} to {to_airport}."

# Define agents
flight_assistant = create_react_agent(
    model="anthropic:claude-3-5-sonnet-latest",
    tools=[book_flight, transfer_to_hotel_assistant],
    prompt="You are a flight booking assistant",
    name="flight_assistant"
)
hotel_assistant = create_react_agent(
    model="anthropic:claude-3-5-sonnet-latest",
    tools=[book_hotel, transfer_to_flight_assistant],
    prompt="You are a hotel booking assistant",
    name="hotel_assistant"
)

# Define multi-agent graph
multi_agent_graph = (
    StateGraph(MessagesState)
    .add_node(flight_assistant)
    .add_node(hotel_assistant)
    .add_edge(START, "flight_assistant")
    .compile()
)

# Run the multi-agent graph
for chunk in multi_agent_graph.stream(
    {
        "messages": [
            {
                "role": "user",
                "content": "book a flight from BOS to JFK and a stay at McKittrick Hotel"
            }
        ]
    }
):
    print(chunk)
    print("\n")

Error:
ValueError: Found AIMessages with tool_calls that do not have a corresponding ToolMessage. Here are the first few of those tool calls: [{'name': 'book_flight', 'args': {'from_airport': 'BOS', 'to_airport': 'JFK'}, 'id': 'call_0_826ae813-27ca-4943-b634-de9cd9a1c69f', 'type': 'tool_call'}].

Every tool call (LLM requesting to call a tool) in the message history MUST have a corresponding ToolMessage (result of a tool invocation to return to the LLM) - this is required by most LLM providers.
For troubleshooting, visit: https://python.langchain.com/docs/troubleshooting/errors/INVALID_CHAT_HISTORY
During task with name 'agent' and id '4fd28bf8-33c9-c779-ab2e-05c0fb8a13e4'
During task with name 'hotel_assistant' and id 'd20ed6c3-6e4a-b728-14f2-5ccd050edb7d'

smith:
flight assistant deepseek output:

  "generations": [
    [
      {
        "text": "",
        "generation_info": {
          "finish_reason": "tool_calls",
          "logprobs": null
        },
        "type": "ChatGeneration",
        "message": {
          "lc": 1,
          "type": "constructor",
          "id": [
            "langchain",
            "schema",
            "messages",
            "AIMessage"
          ],
          "kwargs": {
            "content": "",
            "additional_kwargs": {
              "tool_calls": [
                {
                  "id": "call_0_826ae813-27ca-4943-b634-de9cd9a1c69f",
                  "function": {
                    "arguments": "{\"from_airport\": \"BOS\", \"to_airport\": \"JFK\"}",
                    "name": "book_flight"
                  },
                  "type": "function",
                  "index": 0
                },
                {
                  "id": "call_1_7ca668e5-fd1e-4277-98d4-073e3ae23ec5",
                  "function": {
                    "arguments": "{}",
                    "name": "transfer_to_hotel_assistant"
                  },
                  "type": "function",
                  "index": 1
                }
              ],
              "refusal": null
            },
            "response_metadata": {
              "token_usage": {
                "completion_tokens": 55,
                "prompt_tokens": 231,
                "total_tokens": 286,
                "completion_tokens_details": null,
                "prompt_tokens_details": {
                  "audio_tokens": null,
                  "cached_tokens": 192
                },
                "prompt_cache_hit_tokens": 192,
                "prompt_cache_miss_tokens": 39
              },
              "model_name": "deepseek-chat",
              "system_fingerprint": "fp_8802369eaa_prod0225",
              "id": "87f575c1-0d19-425d-bc22-86251fb4f633",
              "finish_reason": "tool_calls",
              "logprobs": null
            },
            "type": "ai",
            "id": "run-40057ec9-4c41-42be-954f-235a10076565-0",
            "tool_calls": [
              {
                "name": "book_flight",
                "args": {
                  "from_airport": "BOS",
                  "to_airport": "JFK"
                },
                "id": "call_0_826ae813-27ca-4943-b634-de9cd9a1c69f",
                "type": "tool_call"
              },
              {
                "name": "transfer_to_hotel_assistant",
                "args": {},
                "id": "call_1_7ca668e5-fd1e-4277-98d4-073e3ae23ec5",
                "type": "tool_call"
              }
            ],
            "usage_metadata": {
              "input_tokens": 231,
              "output_tokens": 55,
              "total_tokens": 286,
              "input_token_details": {
                "cache_read": 192
              },
              "output_token_details": {}
            },
            "invalid_tool_calls": []
          }
        }
      }
    ]
  ],
  "llm_output": {
    "token_usage": {
      "completion_tokens": 55,
      "prompt_tokens": 231,
      "total_tokens": 286,
      "completion_tokens_details": null,
      "prompt_tokens_details": {
        "audio_tokens": null,
        "cached_tokens": 192
      },
      "prompt_cache_hit_tokens": 192,
      "prompt_cache_miss_tokens": 39
    },
    "model_name": "deepseek-chat",
    "system_fingerprint": "fp_8802369eaa_prod0225",
    "id": "87f575c1-0d19-425d-bc22-86251fb4f633"
  },
  "run": null,
  "type": "LLMResult"
}

hotel assistant agent input:

 "messages": [
   {
     "content": "book a flight from BOS to JFK and a stay at McKittrick Hotel",
     "additional_kwargs": {},
     "response_metadata": {},
     "type": "human",
     "id": "e1d8e20e-075d-492f-8413-50eb256aef4b",
     "example": false
   },
   {
     "content": "",
     "additional_kwargs": {
       "tool_calls": [
         {
           "id": "call_0_826ae813-27ca-4943-b634-de9cd9a1c69f",
           "function": {
             "arguments": "{\"from_airport\": \"BOS\", \"to_airport\": \"JFK\"}",
             "name": "book_flight"
           },
           "type": "function",
           "index": 0
         },
         {
           "id": "call_1_7ca668e5-fd1e-4277-98d4-073e3ae23ec5",
           "function": {
             "arguments": "{}",
             "name": "transfer_to_hotel_assistant"
           },
           "type": "function",
           "index": 1
         }
       ],
       "refusal": null
     },
     "response_metadata": {
       "token_usage": {
         "completion_tokens": 55,
         "prompt_tokens": 231,
         "total_tokens": 286,
         "completion_tokens_details": null,
         "prompt_tokens_details": {
           "audio_tokens": null,
           "cached_tokens": 192
         },
         "prompt_cache_hit_tokens": 192,
         "prompt_cache_miss_tokens": 39
       },
       "model_name": "deepseek-chat",
       "system_fingerprint": "fp_8802369eaa_prod0225",
       "id": "87f575c1-0d19-425d-bc22-86251fb4f633",
       "finish_reason": "tool_calls",
       "logprobs": null
     },
     "type": "ai",
     "id": "run-40057ec9-4c41-42be-954f-235a10076565-0",
     "example": false,
     "tool_calls": [
       {
         "name": "book_flight",
         "args": {
           "from_airport": "BOS",
           "to_airport": "JFK"
         },
         "id": "call_0_826ae813-27ca-4943-b634-de9cd9a1c69f",
         "type": "tool_call"
       },
       {
         "name": "transfer_to_hotel_assistant",
         "args": {},
         "id": "call_1_7ca668e5-fd1e-4277-98d4-073e3ae23ec5",
         "type": "tool_call"
       }
     ],
     "invalid_tool_calls": [],
     "usage_metadata": {
       "input_tokens": 231,
       "output_tokens": 55,
       "total_tokens": 286,
       "input_token_details": {
         "cache_read": 192
       },
       "output_token_details": {}
     }
   },
   {
     "content": "Successfully transferred to hotel_assistant",
     "additional_kwargs": {},
     "response_metadata": {},
     "type": "tool",
     "name": "transfer_to_hotel_assistant",
     "id": "f7a66f18-506a-4cb3-b250-21770af9c4d8",
     "tool_call_id": "call_1_7ca668e5-fd1e-4277-98d4-073e3ae23ec5",
     "status": "success"
   }
 ],
 "is_last_step": false,
 "remaining_steps": 24
}

[NithinXD]

parallel tool calls only works for openai models
handled at model provider level
https://www.reddit.com/r/LangChain/comments/1dmtcyn/looking_for_ideas_how_to_handle_parallel_tool/