[Bug] internlm2_5-7b-chat多卡部署报错 aborted

### 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.
- [ ] 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

4卡T4的服务器，使用lmdeploy部署internlm2_5-7b-chat，张量并行tp=2
模型成功加载到显存，api接口服务正常。
调用推理接口，模型报错aborted，进程结束

使用internlm/internlm2_5-7b-chat-4bit，可以在单卡正常部署

### Reproduction

使用modelscope
`export LMDEPLOY_USE_MODELSCOPE=True`
用cli工具部署服务
`lmdeploy serve api_server Shanghai_AI_Laboratory/internlm2_5-7b-chat --backend turbomind --chat-template internlm2 --tp 2`
用其他机器post请求推理接口
`ip:23333/v1/chat/completions`
`{
"model": "/root/.cache/modelscope/hub/Shanghai_AI_Laboratory/internlm2_5-7b-chat",
"messages": [
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": "讲一个三国故事"}
],
"temperature": 0.7,
"top_p": 0.8
}`
进程报错跳出
`(lmdeploy) [root@local-gpu models]# lmdeploy serve api_server Shanghai_AI_Laboratory/internlm2_5-7b-chat --backend turbomind --chat-template internlm2 --tp 2
[WARNING] gemm_config.in is not found; using default GEMM algo
[WARNING] gemm_config.in is not found; using default GEMM algo
HINT: Please open http://0.0.0.0:23333 in a browser for detailed api usage!!!
HINT: Please open http://0.0.0.0:23333 in a browser for detailed api usage!!!
HINT: Please open http://0.0.0.0:23333 in a browser for detailed api usage!!!
INFO: Started server process [32752]
INFO: Waiting for application startup.
INFO: Application startup complete.
INFO: Uvicorn running on http://0.0.0.0:23333 (Press CTRL+C to quit)
已放弃`

### Environment

```Shell
'(lmdeploy) [root@local-gpu models]# lmdeploy check_env
sys.platform: linux
Python: 3.8.19 (default, Mar 20 2024, 19:58:24) [GCC 11.2.0]
CUDA available: True
MUSA available: False
numpy_random_seed: 2147483648
GPU 0,1,2,3: Tesla T4
CUDA_HOME: /usr/local/cuda
NVCC: Cuda compilation tools, release 12.0, V12.0.140
GCC: gcc (GCC) 4.8.5 20150623 (Red Hat 4.8.5-28)
PyTorch: 2.3.1+cu121
PyTorch compiling details: PyTorch built with:
- GCC 9.3
- C++ Version: 201703
- Intel(R) oneAPI Math Kernel Library Version 2022.2-Product Build 20220804 for Intel(R) 64 architecture applications
- Intel(R) MKL-DNN v3.3.6 (Git Hash 86e6af5974177e513fd3fee58425e1063e7f1361)
- OpenMP 201511 (a.k.a. OpenMP 4.5)
- LAPACK is enabled (usually provided by MKL)
- NNPACK is enabled
- CPU capability usage: AVX512
- CUDA Runtime 12.1
- NVCC architecture flags: -gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-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
- CuDNN 8.9.2
- Magma 2.6.1
- Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=12.1, CUDNN_VERSION=8.9.2, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -D_GLIBCXX_USE_CXX11_ABI=0 -fabi-version=11 -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Werror=bool-operation -Wnarrowing -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-stringop-overflow -Wsuggest-override -Wno-psabi -Wno-error=pedantic -Wno-error=old-style-cast -Wno-missing-braces -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=2.3.1, USE_CUDA=ON, USE_CUDNN=ON, USE_CUSPARSELT=1, USE_EXCEPTION_PTR=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,

TorchVision: 0.18.1+cu121
LMDeploy: 0.6.0+
transformers: 4.44.2
gradio: Not Found
fastapi: 0.115.0
pydantic: 2.9.2
triton: 2.3.1
NVIDIA Topology:
GPU0 GPU1 GPU2 GPU3 CPU Affinity NUMA Affinity
GPU0 X NODE NODE SYS 0-15,32-47 0
GPU1 NODE X PHB SYS 0-15,32-47 0
GPU2 NODE PHB X SYS 0-15,32-47 0
GPU3 SYS SYS SYS X 16-31,48-63 1

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
'
模型是
```

### Error traceback

```Shell
(lmdeploy) [root@local-gpu models]# lmdeploy serve api_server Shanghai_AI_Laboratory/internlm2_5-7b-chat --backend turbomind --chat-template internlm2 --tp 2
[WARNING] gemm_config.in is not found; using default GEMM algo
[WARNING] gemm_config.in is not found; using default GEMM algo
HINT: Please open http://0.0.0.0:23333 in a browser for detailed api usage!!!
HINT: Please open http://0.0.0.0:23333 in a browser for detailed api usage!!!
HINT: Please open http://0.0.0.0:23333 in a browser for detailed api usage!!!
INFO: Started server process [32752]
INFO: Waiting for application startup.
INFO: Application startup complete.
INFO: Uvicorn running on http://0.0.0.0:23333 (Press CTRL+C to quit)
已放弃

因为internlm/internlm2_5-7b-chat-4bit可以在单卡正常部署，我认为可能和显卡驱动和nccl有关
```

Start by reproducing the `lmdeploy serve api_server` command with the InternLM model, `--backend turbomind`, and `--tp 2`, then inspect the reported CUDA/NCCL and multi-GPU behavior. Compare it with the working 4-bit single-GPU deployment; done means the unquantized model can serve the provided chat completion request without the process aborting.

Written by the indexing model from the issue text.