[Bug] KeyError when processing assistant messages with tool_calls but missing content field

### Checklist

- [x] 1. I have searched related issues but cannot get the expected help.
- [x] 2. The bug has not been fixed in the latest version.
- [x] 3. Please note that if the bug-related issue you submitted lacks corresponding environment info and a minimal reproducible demo, it will be challenging for us to reproduce and resolve the issue, reducing the likelihood of receiving feedback.

### Describe the bug

When processing chat messages from models like Qwen or GPT-OSS that return tool calls, the assistant message may have a `tool_calls` field but no `content` field (or `content: null`). This causes a `KeyError` when the code tries to check `isinstance(msg['content'], list)`.

### Reproduction

1. Start an LMDeploy server with Qwen3-8B:
```bash
lmdeploy serve api_server Qwen/Qwen3-8B --tool-call-parser qwen3
```

2. Send a chat completion request with tool calling:
```python
from openai import OpenAI

client = OpenAI(
    api_key='EMPTY',
    base_url='http://localhost:23333/v1'
)

model = 'Qwen/Qwen3-8B'

# Define a simple weather tool
tools = [{
    'type': 'function',
    'function': {
        'name': 'get_weather',
        'description': 'Get the current weather for a location',
        'parameters': {
            'type': 'object',
            'properties': {
                'location': {
                    'type': 'string',
                    'description': 'The city name, e.g. San Francisco'
                }
            },
            'required': ['location']
        }
    }
}]

# Initial user request
messages = [
    {'role': 'user', 'content': 'What is the weather like in Paris?'}
]

response = client.chat.completions.create(
    model=model,
    messages=messages,
    tools=tools,
)

# The assistant responds with a tool call
assistant_message = response.choices[0].message
print(assistant_message)
# Output shows:
# {
#     'role': 'assistant',
#     'content': None,  # or missing entirely
#     'tool_calls': [{
#         'id': 'call_abc123',
#         'type': 'function',
#         'function': {
#             'name': 'get_weather',
#             'arguments': '{"location": "Paris"}'
#         }
#     }]
# }

# Now append the assistant message and tool result to continue the conversation
messages.append({
    'role': 'assistant',
    'tool_calls': assistant_message.tool_calls
    # Note: 'content' field is missing or None
})

messages.append({
    'role': 'tool',
    'tool_call_id': assistant_message.tool_calls[0].id,
    'content': '{"temperature": 18, "condition": "sunny"}'
})

# Send the follow-up request - this triggers the bug!
response = client.chat.completions.create(
    model=model,
    messages=messages
)
```

BUG ANALYSIS: In `lmdeploy/serve/async_engine.py`, the code directly accesses `msg['content']` without checking if the field exists:
```python
if isinstance(prompt, list) and any(isinstance(msg['content'], list) for msg in prompt):
```

### Environment

```Shell
/bin/sh: 1: /usr/local/cuda/bin/nvcc: not found
/bin/sh: 1: gcc: not found
sys.platform: linux
Python: 3.11.13 (main, Jun  5 2025, 13:12:00) [GCC 11.2.0]
CUDA available: True
MUSA available: False
numpy_random_seed: 2147483648
GPU 0,1,2,3: Tesla V100-SXM2-16GB
CUDA_HOME: /usr/local/cuda
NVCC: Not Available
GCC: n/a
PyTorch: 2.8.0+cu128
PyTorch compiling details: PyTorch built with:
  - GCC 13.3
  - C++ Version: 201703
  - Intel(R) oneAPI Math Kernel Library Version 2024.2-Product Build 20240605 for Intel(R) 64 architecture applications
  - Intel(R) MKL-DNN v3.7.1 (Git Hash 8d263e693366ef8db40acc569cc7d8edf644556d)
  - OpenMP 201511 (a.k.a. OpenMP 4.5)
  - LAPACK is enabled (usually provided by MKL)
  - NNPACK is enabled
  - CPU capability usage: AVX2
  - CUDA Runtime 12.8
  - NVCC architecture flags: -gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86;-gencode;arch=compute_90,code=sm_90;-gencode;arch=compute_100,code=sm_100;-gencode;arch=compute_120,code=sm_120
  - CuDNN 91.0.2  (built against CUDA 12.9)
    - Built with CuDNN 90.8
  - Magma 2.6.1
  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, COMMIT_SHA=a1cb3cc05d46d198467bebbb6e8fba50a325d4e7, CUDA_VERSION=12.8, CUDNN_VERSION=9.8.0, CXX_COMPILER=/opt/rh/gcc-toolset-13/root/usr/bin/c++, CXX_FLAGS= -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DLIBKINETO_NOXPUPTI=ON -DUSE_FBGEMM -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -DC10_NODEPRECATED -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Werror=range-loop-construct -Werror=bool-operation -Wnarrowing -Wno-missing-field-initializers -Wno-unknown-pragmas -Wno-unused-parameter -Wno-strict-overflow -Wno-strict-aliasing -Wno-stringop-overflow -Wsuggest-override -Wno-psabi -Wno-error=old-style-cast -faligned-new -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-dangling-reference -Wno-error=dangling-reference -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, TORCH_VERSION=2.8.0, USE_CUDA=ON, USE_CUDNN=ON, USE_CUSPARSELT=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_GLOO=ON, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=1, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, USE_ROCM_KERNEL_ASSERT=OFF, USE_XCCL=OFF, USE_XPU=OFF,

TorchVision: 0.23.0+cu128
LMDeploy: 0.10.1+unknown
transformers: 4.56.2
fastapi: 0.117.1
pydantic: 2.11.9
triton: 3.4.0
NVIDIA Topology:
        GPU0    GPU1    GPU2    GPU3    CPU Affinity    NUMA Affinity   GPU NUMA ID
GPU0     X      NV2     NV2     NV2     0,2,4,6,8,10    0               N/A
GPU1    NV2      X      NV2     NV2     0,2,4,6,8,10    0               N/A
GPU2    NV2     NV2      X      NV2     0,2,4,6,8,10    0               N/A
GPU3    NV2     NV2     NV2      X      0,2,4,6,8,10    0               N/A

Legend:

  X    = Self
  SYS  = Connection traversing PCIe as well as the SMP interconnect between NUMA nodes (e.g., QPI/UPI)
  NODE = Connection traversing PCIe as well as the interconnect between PCIe Host Bridges within a NUMA node
  PHB  = Connection traversing PCIe as well as a PCIe Host Bridge (typically the CPU)
  PXB  = Connection traversing multiple PCIe bridges (without traversing the PCIe Host Bridge)
  PIX  = Connection traversing at most a single PCIe bridge
  NV#  = Connection traversing a bonded set of # NVLinks
```

### Error traceback

```Shell
2025-10-11 14:08:43,577 - lmdeploy - INFO - turbomind.py:773 - [async_stream_infer] session 5 done
2025-10-11 14:08:43,577 - lmdeploy - INFO - async_engine.py:879 - session 5 finished, reason "stop", input_tokens 162, output_tokens 21
INFO:     127.0.0.1:41294 - "POST /v1/chat/completions HTTP/1.1" 200 OK
INFO:     127.0.0.1:41294 - "POST /v1/chat/completions HTTP/1.1" 500 Internal Server Error
ERROR:    Exception in ASGI application
Traceback (most recent call last):
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/uvicorn/protocols/http/h11_impl.py", line 403, in run_asgi
    result = await app(  # type: ignore[func-returns-value]
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/uvicorn/middleware/proxy_headers.py", line 60, in __call__
    return await self.app(scope, receive, send)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/fastapi/applications.py", line 1082, in __call__
    await super().__call__(scope, receive, send)
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/starlette/applications.py", line 113, in __call__
    await self.middleware_stack(scope, receive, send)
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/starlette/middleware/errors.py", line 186, in __call__
    raise exc
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/starlette/middleware/errors.py", line 164, in __call__
    await self.app(scope, receive, _send)
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/starlette/middleware/cors.py", line 85, in __call__
    await self.app(scope, receive, send)
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/starlette/middleware/exceptions.py", line 63, in __call__
    await wrap_app_handling_exceptions(self.app, conn)(scope, receive, send)
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/starlette/_exception_handler.py", line 53, in wrapped_app
    raise exc
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/starlette/_exception_handler.py", line 42, in wrapped_app
    await app(scope, receive, sender)
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/starlette/routing.py", line 716, in __call__
    await self.middleware_stack(scope, receive, send)
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/starlette/routing.py", line 736, in app
    await route.handle(scope, receive, send)
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/starlette/routing.py", line 290, in handle
    await self.app(scope, receive, send)
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/starlette/routing.py", line 78, in app
    await wrap_app_handling_exceptions(app, request)(scope, receive, send)
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/starlette/_exception_handler.py", line 53, in wrapped_app
    raise exc
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/starlette/_exception_handler.py", line 42, in wrapped_app
    await app(scope, receive, sender)
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/starlette/routing.py", line 75, in app
    response = await f(request)
               ^^^^^^^^^^^^^^^^
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/fastapi/routing.py", line 308, in app
    raw_response = await run_endpoint_function(
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/fastapi/routing.py", line 219, in run_endpoint_function
    return await dependant.call(**values)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/lmdeploy/serve/openai/api_server.py", line 572, in chat_completions_v1
    async for res in result_generator:
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/lmdeploy/serve/async_engine.py", line 730, in generate
    prompt_input = await self._get_prompt_input(prompt,
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/lmdeploy/serve/async_engine.py", line 612, in _get_prompt_input
    if isinstance(prompt, list) and any(isinstance(msg['content'], list) for msg in prompt):
                                    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/ma/lmdeploy-0.10.1/lib/python3.11/site-packages/lmdeploy/serve/async_engine.py", line 612, in <genexpr>
    if isinstance(prompt, list) and any(isinstance(msg['content'], list) for msg in prompt):
                                                   ~~~^^^^^^^^^^^
KeyError: 'content'
```

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[Bug] KeyError when processing assistant messages with tool_calls but missing content field #4035

Checklist

Describe the bug

Reproduction

Environment

Error traceback

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

[Bug] KeyError when processing assistant messages with tool_calls but missing content field #4035

Description

Checklist

Describe the bug

Reproduction

Environment

Error traceback

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions