[Bug] chat with gemma-2-27b-it, response is empty.

[zhulinJulia24]

Checklist

• 1. I have searched related issues but cannot get the expected help.
• 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

[Bug] chat with gemma-2-27b-it, response is empty.

Reproduction

lmdeploy chat /nvme/qa_test_models/google/gemma-2-27b-it --backend pytorch --session-len 4096 --tp 2

Environment

sys.platform: linux
Python: 3.10.12 (main, Nov  6 2024, 20:22:13) [GCC 11.4.0]
CUDA available: True
MUSA available: False
numpy_random_seed: 2147483648
GPU 0,1,2,3,4,5,6,7: NVIDIA A100-SXM4-80GB
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: 2.4.0+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=2.4.0, 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: 0.19.0+cu118
LMDeploy: 0.6.4+
transformers: 4.47.1
gradio: 5.9.1
fastapi: 0.115.6
pydantic: 2.10.3
triton: 3.0.0
NVIDIA Topology: 
        GPU0    GPU1    GPU2    GPU3    GPU4    GPU5    GPU6    GPU7    CPU Affinity    NUMA Affinity
GPU0     X      NV12    NV12    NV12    NV12    NV12    NV12    NV12    0-27,56-83      0
GPU1    NV12     X      NV12    NV12    NV12    NV12    NV12    NV12    0-27,56-83      0
GPU2    NV12    NV12     X      NV12    NV12    NV12    NV12    NV12    0-27,56-83      0
GPU3    NV12    NV12    NV12     X      NV12    NV12    NV12    NV12    0-27,56-83      0
GPU4    NV12    NV12    NV12    NV12     X      NV12    NV12    NV12    28-55,84-111    1
GPU5    NV12    NV12    NV12    NV12    NV12     X      NV12    NV12    28-55,84-111    1
GPU6    NV12    NV12    NV12    NV12    NV12    NV12     X      NV12    28-55,84-111    1
GPU7    NV12    NV12    NV12    NV12    NV12    NV12    NV12     X      28-55,84-111    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

Error traceback

root@ac035a23f61d:/__w/lmdeploy/lmdeploy# lmdeploy chat /nvme/qa_test_models/google/gemma-2-27b-it --backend pytorch  --session-len 4096 --tp 2
2024-12-23 10:21:55,417 - lmdeploy - WARNING - transformers.py:22 - LMDeploy requires transformers version: [4.33.0 ~ 4.46.1], but found version: 4.47.1
2024-12-23 10:21:55,424 - lmdeploy - INFO - model_agent.py:613 - MASTER_ADDR=127.0.0.1, MASTER_PORT=29500
2024-12-23 10:21:58,142 - lmdeploy - INFO - model_agent.py:335 - build model.
2024-12-23 10:21:58,432 - lmdeploy - INFO - model_agent.py:338 - loading weights.
2024-12-23 10:21:58,433 - lmdeploy - INFO - model_weight_loader.py:142 - rank[0] loading weights - "model-00003-of-00012.safetensors"
2024-12-23 10:21:58,653 - lmdeploy - INFO - model_weight_loader.py:142 - rank[1] loading weights - "model-00001-of-00012.safetensors"
2024-12-23 10:21:59,167 - lmdeploy - INFO - model_weight_loader.py:142 - rank[0] loading weights - "model-00002-of-00012.safetensors"
2024-12-23 10:21:59,672 - lmdeploy - INFO - model_weight_loader.py:142 - rank[1] loading weights - "model-00009-of-00012.safetensors"
2024-12-23 10:22:00,109 - lmdeploy - INFO - model_weight_loader.py:142 - rank[0] loading weights - "model-00010-of-00012.safetensors"
2024-12-23 10:22:00,591 - lmdeploy - INFO - model_weight_loader.py:142 - rank[1] loading weights - "model-00006-of-00012.safetensors"
2024-12-23 10:22:01,090 - lmdeploy - INFO - model_weight_loader.py:142 - rank[0] loading weights - "model-00011-of-00012.safetensors"
2024-12-23 10:22:01,544 - lmdeploy - INFO - model_weight_loader.py:142 - rank[1] loading weights - "model-00007-of-00012.safetensors"
2024-12-23 10:22:02,133 - lmdeploy - INFO - model_weight_loader.py:142 - rank[0] loading weights - "model-00006-of-00012.safetensors"
2024-12-23 10:22:02,426 - lmdeploy - INFO - model_weight_loader.py:142 - rank[1] loading weights - "model-00012-of-00012.safetensors"
2024-12-23 10:22:02,737 - lmdeploy - INFO - model_weight_loader.py:142 - rank[1] loading weights - "model-00002-of-00012.safetensors"
2024-12-23 10:22:03,005 - lmdeploy - INFO - model_weight_loader.py:142 - rank[0] loading weights - "model-00005-of-00012.safetensors"
2024-12-23 10:22:03,500 - lmdeploy - INFO - model_weight_loader.py:142 - rank[1] loading weights - "model-00005-of-00012.safetensors"
2024-12-23 10:22:03,859 - lmdeploy - INFO - model_weight_loader.py:142 - rank[0] loading weights - "model-00009-of-00012.safetensors"
2024-12-23 10:22:04,417 - lmdeploy - INFO - model_weight_loader.py:142 - rank[1] loading weights - "model-00011-of-00012.safetensors"
2024-12-23 10:22:04,810 - lmdeploy - INFO - model_weight_loader.py:142 - rank[0] loading weights - "model-00008-of-00012.safetensors"
2024-12-23 10:22:05,309 - lmdeploy - INFO - model_weight_loader.py:142 - rank[1] loading weights - "model-00003-of-00012.safetensors"
2024-12-23 10:22:05,697 - lmdeploy - INFO - model_weight_loader.py:142 - rank[0] loading weights - "model-00012-of-00012.safetensors"
2024-12-23 10:22:06,071 - lmdeploy - INFO - model_weight_loader.py:142 - rank[0] loading weights - "model-00004-of-00012.safetensors"
2024-12-23 10:22:06,189 - lmdeploy - INFO - model_weight_loader.py:142 - rank[1] loading weights - "model-00008-of-00012.safetensors"
2024-12-23 10:22:06,875 - lmdeploy - INFO - model_weight_loader.py:142 - rank[0] loading weights - "model-00007-of-00012.safetensors"
2024-12-23 10:22:07,089 - lmdeploy - INFO - model_weight_loader.py:142 - rank[1] loading weights - "model-00004-of-00012.safetensors"
2024-12-23 10:22:07,911 - lmdeploy - INFO - model_weight_loader.py:142 - rank[0] loading weights - "model-00001-of-00012.safetensors"
2024-12-23 10:22:08,046 - lmdeploy - INFO - model_weight_loader.py:142 - rank[1] loading weights - "model-00010-of-00012.safetensors"
2024-12-23 10:22:09,740 - lmdeploy - INFO - cache_engine.py:36 - build CacheEngine with config:CacheConfig(max_batches=256, block_size=64, num_cpu_blocks=89, num_gpu_blocks=3235, window_size=4096, cache_max_entry_count=0.8, max_prefill_token_num=4096, enable_prefix_caching=False, quant_policy=0, device_type='cuda')

double enter to end input >>> 你好

<start_of_turn>user
你好<end_of_turn>
<start_of_turn>model


double enter to end input >>> hi

<start_of_turn>user
hi<end_of_turn>
<start_of_turn>model

[AllentDan]

The side effect of #2872. @RunningLeon May have a look.

[RunningLeon]

@zhulinJulia24 can you retest on #2933 ?

[zhulinJulia24]

@zhulinJulia24 can you retest on #2933 ?

fixed on this pr