lmdeploy对DeepSeek-R1-Distill-Llama-70B进行api推理时，使用H100的两张显卡，共160G，会出现out of memory

[winni0]

trafficstars

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

lmdeploy对DeepSeek-R1-Distill-Llama-70B进行api推理时，使用H100的两张显卡，共160G，会出现out of memory；当指定4张卡，共320G，且在命令中加入--tp 4时，70B能够正常运行。我测试了deepseek蒸馏版的70B、32B、14B、8B，api推理时所需显存分别为292G、280G、74G、73G，为什么会出现这种情况，显存为什么用了这么多。一般情况下70B不是140G显存就够了吗？

Reproduction

CUDA_VISIBLE_DEVICES=3,4 lmdeploy serve api_server /app123/model/DeepSeek-R1-Distill-Llama-70B --backend turbomind --server-port 8000 --device cuda --chat-template deepseek

Environment

sys.platform: linux
Python: 3.10.12 (main, Jan 17 2025, 14:35:34) [GCC 11.4.0]
CUDA available: True
MUSA available: False
numpy_random_seed: 2147483648
GPU 0,1,2,3,4,5,6,7: NVIDIA H100 80GB HBM3
CUDA_HOME: /usr/local/cuda
NVCC: Cuda compilation tools, release 11.8, V11.8.89
GCC: x86_64-linux-gnu-gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0
PyTorch: 2.4.1+cu118
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 v3.4.2 (Git Hash 1137e04ec0b5251ca2b4400a4fd3c667ce843d67)
  - OpenMP 201511 (a.k.a. OpenMP 4.5)
  - LAPACK is enabled (usually provided by MKL)
  - NNPACK is enabled
  - CPU capability usage: AVX512
  - CUDA Runtime 11.8
  - 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_37,code=sm_37;-gencode;arch=compute_90,code=sm_90
  - CuDNN 90.1
  - Magma 2.6.1
  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.8, CUDNN_VERSION=9.1.0, 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_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=2.4.1, 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: 0.19.1+cu118
LMDeploy: 0.6.5+
transformers: 4.48.1
gradio: 5.13.1
fastapi: 0.115.7
pydantic: 2.10.6
triton: 3.0.0
NVIDIA Topology: 
        GPU0    GPU1    GPU2    GPU3    GPU4    GPU5    GPU6    GPU7    NIC0    NIC1    NIC2    NIC3    NIC4    NIC5    NIC6    NIC7    CPU Affinity    NUMA Affinity   GPU NUMA ID
GPU0     X      NV18    NV18    NV18    NV18    NV18    NV18    NV18    PXB     PXB     SYS     SYS     SYS     SYS     SYS     SYS     0-47,96-143     0               N/A
GPU1    NV18     X      NV18    NV18    NV18    NV18    NV18    NV18    PXB     PXB     SYS     SYS     SYS     SYS     SYS     SYS     0-47,96-143     0               N/A
GPU2    NV18    NV18     X      NV18    NV18    NV18    NV18    NV18    SYS     SYS     PXB     PXB     SYS     SYS     SYS     SYS     0-47,96-143     0               N/A
GPU3    NV18    NV18    NV18     X      NV18    NV18    NV18    NV18    SYS     SYS     PXB     PXB     SYS     SYS     SYS     SYS     0-47,96-143     0               N/A
GPU4    NV18    NV18    NV18    NV18     X      NV18    NV18    NV18    SYS     SYS     SYS     SYS     PXB     PXB     SYS     SYS     48-95,144-191   1               N/A
GPU5    NV18    NV18    NV18    NV18    NV18     X      NV18    NV18    SYS     SYS     SYS     SYS     PXB     PXB     SYS     SYS     48-95,144-191   1               N/A
GPU6    NV18    NV18    NV18    NV18    NV18    NV18     X      NV18    SYS     SYS     SYS     SYS     SYS     SYS     PXB     PXB     48-95,144-191   1               N/A
GPU7    NV18    NV18    NV18    NV18    NV18    NV18    NV18     X      SYS     SYS     SYS     SYS     SYS     SYS     PXB     PXB     48-95,144-191   1               N/A
NIC0    PXB     PXB     SYS     SYS     SYS     SYS     SYS     SYS      X      PXB     SYS     SYS     SYS     SYS     SYS     SYS
NIC1    PXB     PXB     SYS     SYS     SYS     SYS     SYS     SYS     PXB      X      SYS     SYS     SYS     SYS     SYS     SYS
NIC2    SYS     SYS     PXB     PXB     SYS     SYS     SYS     SYS     SYS     SYS      X      PXB     SYS     SYS     SYS     SYS
NIC3    SYS     SYS     PXB     PXB     SYS     SYS     SYS     SYS     SYS     SYS     PXB      X      SYS     SYS     SYS     SYS
NIC4    SYS     SYS     SYS     SYS     PXB     PXB     SYS     SYS     SYS     SYS     SYS     SYS      X      PXB     SYS     SYS
NIC5    SYS     SYS     SYS     SYS     PXB     PXB     SYS     SYS     SYS     SYS     SYS     SYS     PXB      X      SYS     SYS
NIC6    SYS     SYS     SYS     SYS     SYS     SYS     PXB     PXB     SYS     SYS     SYS     SYS     SYS     SYS      X      PXB
NIC7    SYS     SYS     SYS     SYS     SYS     SYS     PXB     PXB     SYS     SYS     SYS     SYS     SYS     SYS     PXB      X 

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

NIC Legend:

  NIC0: mlx5_0
  NIC1: mlx5_1
  NIC2: mlx5_2
  NIC3: mlx5_3
  NIC4: mlx5_4
  NIC5: mlx5_5
  NIC6: mlx5_6
  NIC7: mlx5_7

Error traceback

Traceback (most recent call last):
  File "/opt/py3/bin/lmdeploy", line 33, in <module>
    sys.exit(load_entry_point('lmdeploy', 'console_scripts', 'lmdeploy')())
  File "/opt/lmdeploy/lmdeploy/cli/entrypoint.py", line 42, in run
    args.run(args)
  File "/opt/lmdeploy/lmdeploy/cli/serve.py", line 339, in api_server
    run_api_server(args.model_path,
  File "/opt/lmdeploy/lmdeploy/serve/openai/api_server.py", line 1088, in serve
    VariableInterface.async_engine = pipeline_class(
  File "/opt/lmdeploy/lmdeploy/serve/async_engine.py", line 159, in __init__
    self._build_turbomind(model_path=model_path,
  File "/opt/lmdeploy/lmdeploy/serve/async_engine.py", line 198, in _build_turbomind
    self.engine = tm.TurboMind.from_pretrained(
  File "/opt/lmdeploy/lmdeploy/turbomind/turbomind.py", line 302, in from_pretrained
    return cls(model_path=pretrained_model_name_or_path,
  File "/opt/lmdeploy/lmdeploy/turbomind/turbomind.py", line 112, in __init__
    self.model_comm = self._from_hf(model_source=model_source,
  File "/opt/lmdeploy/lmdeploy/turbomind/turbomind.py", line 226, in _from_hf
    self._create_weight(model_comm)
  File "/opt/lmdeploy/lmdeploy/turbomind/turbomind.py", line 152, in _create_weight
    future.result()
  File "/usr/lib/python3.10/concurrent/futures/_base.py", line 458, in result
    return self.__get_result()
  File "/usr/lib/python3.10/concurrent/futures/_base.py", line 403, in __get_result
    raise self._exception
  File "/usr/lib/python3.10/concurrent/futures/thread.py", line 58, in run
    result = self.fn(*self.args, **self.kwargs)
  File "/opt/lmdeploy/lmdeploy/turbomind/turbomind.py", line 145, in _create_weight_func
    model_comm.create_shared_weights(device_id, rank)
RuntimeError: [TM][ERROR] CUDA runtime error: out of memory /opt/lmdeploy/src/turbomind/utils/memory_utils.cu:31

[lzhangzz]

CUDA_VISIBLE_DEVICES=3,4 lmdeploy serve api_server /app123/model/DeepSeek-R1-Distill-Llama-70B --backend turbomind --server-port 8000 --device cuda --chat-template deepseek

需要加上 --tp 2

[winni0]

加上以后命令为：CUDA_VISIBLE_DEVICES=3,4 lmdeploy serve api_server /app123/model/DeepSeek-R1-Distill-Llama-70B --backend turbomind --server-port 8000 --device cuda --chat-template deepseek --tp 2 报错：[WARNING] gemm_config.in is not found; using default GEMM algo
[WARNING] gemm_config.in is not found; using default GEMM algo terminate called after throwing an instance of 'std::runtime_error' what(): [TM][ERROR] pointer_mapping_ does not have information of ptr at 0x1647104000. Assertion fail: /opt/lmdeploy/src/turbomind/utils/allocator.h:284

Aborted (core dumped) 怎么解决？

[sunxiaoyu12]

你好，请问你部署deepseek蒸馏模型的时候，它会think吗？