[Bug]: Medusa speculation hangs when tp > 1 #16477
Description
Your current environment
The output of `python collect_env.py`
INFO 04-11 09:55:02 [init.py:239] Automatically detected platform cuda.
Collecting environment information...
/opt/conda/lib/python3.11/site-packages/_distutils_hack/init.py:32: UserWarning: Setuptools is replacing distutils. Support for replacing an already imported distutils is deprecated. In the future, this condition will fail. Register concerns at https://github.com/pypa/setuptools/issues/new?template=distutils-deprecation.yml
warnings.warn(
PyTorch version: 2.6.0+cu124
Is debug build: False
CUDA used to build PyTorch: 12.4
ROCM used to build PyTorch: N/A
OS: Ubuntu 22.04.4 LTS (x86_64)
GCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0
Clang version: Could not collect
CMake version: version 3.30.5
Libc version: glibc-2.35
Python version: 3.11.10 | packaged by conda-forge | (main, Oct 16 2024, 01:27:36) [GCC 13.3.0] (64-bit runtime)
Python platform: Linux-5.15.0-131-generic-x86_64-with-glibc2.35
Is CUDA available: True
CUDA runtime version: 12.4.131
CUDA_MODULE_LOADING set to: LAZY
GPU models and configuration:
GPU 0: NVIDIA H100 80GB HBM3
GPU 1: NVIDIA H100 80GB HBM3
Nvidia driver version: 550.54.15
cuDNN version: Could not collect
HIP runtime version: N/A
MIOpen runtime version: N/A
Is XNNPACK available: True
CPU:
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Address sizes: 43 bits physical, 57 bits virtual
Byte Order: Little Endian
CPU(s): 128
On-line CPU(s) list: 0-127
Vendor ID: GenuineIntel
Model name: Intel(R) Xeon(R) Platinum 8468
CPU family: 6
Model: 143
Thread(s) per core: 2
Core(s) per socket: 32
Socket(s): 2
Stepping: 8
BogoMIPS: 4200.00
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts mmx fxsr sse sse2 ss ht syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology cpuid tsc_known_freq pni pclmulqdq dtes64 ssse3 fma cx16 pdcm pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand hypervisor lahf_lm abm 3dnowprefetch cpuid_fault invpcid_single ssbd ibrs ibpb stibp ibrs_enhanced fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves avx_vnni avx512_bf16 wbnoinvd arat avx512vbmi umip pku ospke waitpkg avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg avx512_vpopcntdq la57 rdpid bus_lock_detect cldemote movdiri movdir64b fsrm md_clear serialize tsxldtrk amx_bf16 avx512_fp16 amx_tile amx_int8 arch_capabilities
Hypervisor vendor: KVM
Virtualization type: full
L1d cache: 4 MiB (128 instances)
L1i cache: 4 MiB (128 instances)
L2 cache: 256 MiB (64 instances)
L3 cache: 32 MiB (2 instances)
NUMA node(s): 2
NUMA node0 CPU(s): 0-63
NUMA node1 CPU(s): 64-127
Vulnerability Gather data sampling: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Unknown: No mitigations
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl and seccomp
Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2: Mitigation; Enhanced / Automatic IBRS; IBPB conditional; RSB filling; PBRSB-eIBRS SW sequence; BHI SW loop, KVM SW loop
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Mitigation; TSX disabled
Versions of relevant libraries:
[pip3] numpy==2.1.2
[pip3] nvidia-cublas-cu12==12.4.5.8
[pip3] nvidia-cuda-cupti-cu12==12.4.127
[pip3] nvidia-cuda-nvrtc-cu12==12.4.127
[pip3] nvidia-cuda-runtime-cu12==12.4.127
[pip3] nvidia-cudnn-cu12==9.1.0.70
[pip3] nvidia-cufft-cu12==11.2.1.3
[pip3] nvidia-curand-cu12==10.3.5.147
[pip3] nvidia-cusolver-cu12==11.6.1.9
[pip3] nvidia-cusparse-cu12==12.3.1.170
[pip3] nvidia-cusparselt-cu12==0.6.2
[pip3] nvidia-nccl-cu12==2.21.5
[pip3] nvidia-nvjitlink-cu12==12.4.127
[pip3] nvidia-nvtx-cu12==12.4.127
[pip3] optree==0.13.0
[pip3] pyzmq==26.4.0
[pip3] torch==2.6.0
[pip3] torchaudio==2.6.0
[pip3] torchelastic==0.2.2
[pip3] torchvision==0.21.0
[pip3] transformers==4.51.1
[pip3] triton==3.2.0
[conda] numpy 2.1.2 py311h71ddf71_0 conda-forge
[conda] nvidia-cublas-cu12 12.4.5.8 pypi_0 pypi
[conda] nvidia-cuda-cupti-cu12 12.4.127 pypi_0 pypi
[conda] nvidia-cuda-nvrtc-cu12 12.4.127 pypi_0 pypi
[conda] nvidia-cuda-runtime-cu12 12.4.127 pypi_0 pypi
[conda] nvidia-cudnn-cu12 9.1.0.70 pypi_0 pypi
[conda] nvidia-cufft-cu12 11.2.1.3 pypi_0 pypi
[conda] nvidia-curand-cu12 10.3.5.147 pypi_0 pypi
[conda] nvidia-cusolver-cu12 11.6.1.9 pypi_0 pypi
[conda] nvidia-cusparse-cu12 12.3.1.170 pypi_0 pypi
[conda] nvidia-cusparselt-cu12 0.6.2 pypi_0 pypi
[conda] nvidia-nccl-cu12 2.21.5 pypi_0 pypi
[conda] nvidia-nvjitlink-cu12 12.4.127 pypi_0 pypi
[conda] nvidia-nvtx-cu12 12.4.127 pypi_0 pypi
[conda] optree 0.13.0 pypi_0 pypi
[conda] pyzmq 26.4.0 pypi_0 pypi
[conda] torch 2.6.0 pypi_0 pypi
[conda] torchaudio 2.6.0 pypi_0 pypi
[conda] torchelastic 0.2.2 pypi_0 pypi
[conda] torchvision 0.21.0 pypi_0 pypi
[conda] transformers 4.51.1 pypi_0 pypi
[conda] triton 3.2.0 pypi_0 pypi
ROCM Version: Could not collect
Neuron SDK Version: N/A
vLLM Version: 0.1.dev5707+g40b4284 (git sha: 40b4284
vLLM Build Flags:
CUDA Archs: Not Set; ROCm: Disabled; Neuron: Disabled
GPU Topology:
GPU0 GPU1 NIC0 NIC1 NIC2 NIC3 NIC4 NIC5 NIC6 NIC7 CPU Affinity NUMA Affinity GPU NUMA ID
GPU0 X NV18 PHB PHB PIX PHB SYS SYS SYS SYS 64-127 1 N/A
GPU1 NV18 X PHB PHB PHB PIX SYS SYS SYS SYS 64-127 1 N/A
NIC0 PHB PHB X PHB PHB PHB SYS SYS SYS SYS
NIC1 PHB PHB PHB X PHB PHB SYS SYS SYS SYS
NIC2 PIX PHB PHB PHB X PHB SYS SYS SYS SYS
NIC3 PHB PIX PHB PHB PHB X SYS SYS SYS SYS
NIC4 SYS SYS SYS SYS SYS SYS X PHB PHB PHB
NIC5 SYS SYS SYS SYS SYS SYS PHB X PHB PHB
NIC6 SYS SYS SYS SYS SYS SYS PHB PHB X PHB
NIC7 SYS SYS SYS SYS SYS SYS PHB PHB PHB 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
NVIDIA_VISIBLE_DEVICES=GPU-3ad756e5-3f5c-c2bf-e0db-73166e8a21e9,GPU-901f596e-e924-303e-897b-9153339612bc
NVIDIA_REQUIRE_CUDA=cuda>=12.4 brand=tesla,driver>=470,driver<471 brand=unknown,driver>=470,driver<471 brand=nvidia,driver>=470,driver<471 brand=nvidiartx,driver>=470,driver<471 brand=geforce,driver>=470,driver<471 brand=geforcertx,driver>=470,driver<471 brand=quadro,driver>=470,driver<471 brand=quadrortx,driver>=470,driver<471 brand=titan,driver>=470,driver<471 brand=titanrtx,driver>=470,driver<471 brand=tesla,driver>=525,driver<526 brand=unknown,driver>=525,driver<526 brand=nvidia,driver>=525,driver<526 brand=nvidiartx,driver>=525,driver<526 brand=geforce,driver>=525,driver<526 brand=geforcertx,driver>=525,driver<526 brand=quadro,driver>=525,driver<526 brand=quadrortx,driver>=525,driver<526 brand=titan,driver>=525,driver<526 brand=titanrtx,driver>=525,driver<526 brand=tesla,driver>=535,driver<536 brand=unknown,driver>=535,driver<536 brand=nvidia,driver>=535,driver<536 brand=nvidiartx,driver>=535,driver<536 brand=geforce,driver>=535,driver<536 brand=geforcertx,driver>=535,driver<536 brand=quadro,driver>=535,driver<536 brand=quadrortx,driver>=535,driver<536 brand=titan,driver>=535,driver<536 brand=titanrtx,driver>=535,driver<536
NCCL_VERSION=2.21.5-1
NVIDIA_DRIVER_CAPABILITIES=compute,utility
NVIDIA_PRODUCT_NAME=CUDA
CUDA_VERSION=12.4.1
PYTORCH_VERSION=2.5.1
LD_LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64
NCCL_CUMEM_ENABLE=0
PYTORCH_NVML_BASED_CUDA_CHECK=1
TORCHINDUCTOR_COMPILE_THREADS=1
CUDA_MODULE_LOADING=LAZY
🐛 Describe the bug
Hi team!
Launching a model with Medusa heads as a drafter with TP > 1 causes hangs when collecting the unembedding layer. Code to reproduce:
from vllm.entrypoints.llm import LLM
if __name__ == "__main__":
MODEL_NAME = "JackFram/llama-68m"
SPEC_MODEL = "abhigoyal/vllm-medusa-llama-68m-random"
llm = LLM(
model=MODEL_NAME,
max_model_len=1024,
speculative_config={
"model": SPEC_MODEL,
"num_speculative_tokens": 5,
},
tensor_parallel_size=2,
seed=0,
)
outputs = llm.generate(prompts=["Hi! How are you doing?"], use_tqdm=True)Output with `VLLM_TRACE_FUNCTION=1`
[rank0]:[E411 10:18:06.598287174 ProcessGroupNCCL.cpp:629] [Rank 0] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=11, OpType=GATHER, NumelIn=16000, NumelOut=32000, Timeout(ms)=600000) ran for 600012 milliseconds before timing out. [rank0]:[E411 10:18:06.598683243 ProcessGroupNCCL.cpp:2168] [PG ID 2 PG GUID 3 Rank 0] failure detected by watchdog at work sequence id: 11 PG status: last enqueued work: 11, last completed work: 10 [rank0]:[E411 10:18:06.598698816 ProcessGroupNCCL.cpp:667] Stack trace of the failed collective not found, potentially because FlightRecorder is disabled. You can enable it by setting TORCH_NCCL_TRACE_BUFFER_SIZE to a non-zero value. [rank0]:[E411 10:18:06.598702813 ProcessGroupNCCL.cpp:681] [Rank 0] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data. [rank0]:[E411 10:18:06.598707200 ProcessGroupNCCL.cpp:695] [Rank 0] To avoid data inconsistency, we are taking the entire process down. [rank0]:[E411 10:18:06.599898745 ProcessGroupNCCL.cpp:1895] [PG ID 2 PG GUID 3 Rank 0] Process group watchdog thread terminated with exception: [Rank 0] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=11, OpType=GATHER, NumelIn=16000, NumelOut=32000, Timeout(ms)=600000) ran for 600012 milliseconds before timing out. Exception raised from checkTimeout at /pytorch/torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:632 (most recent call first): frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x96 (0x7f802576c1b6 in /opt/conda/lib/python3.11/site-packages/torch/lib/libc10.so) frame #1: c10d::ProcessGroupNCCL::WorkNCCL::checkTimeout(std::optional > >) + 0x2b4 (0x7f7fd3a1bc74 in /opt/conda/lib/python3.11/site-packages/torch/lib/libtorch_cuda.so) frame #2: c10d::ProcessGroupNCCL::watchdogHandler() + 0x890 (0x7f7fd3a1d7d0 in /opt/conda/lib/python3.11/site-packages/torch/lib/libtorch_cuda.so) frame #3: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x14d (0x7f7fd3a1e6ed in /opt/conda/lib/python3.11/site-packages/torch/lib/libtorch_cuda.so) frame #4: + 0x145c0 (0x7f80260615c0 in /opt/conda/lib/python3.11/site-packages/torch/lib/libtorch.so) frame #5: + 0x94ac3 (0x7f80264d7ac3 in /lib/x86_64-linux-gnu/libc.so.6) frame #6: + 0x126850 (0x7f8026569850 in /lib/x86_64-linux-gnu/libc.so.6)terminate called after throwing an instance of 'c10::DistBackendError'
what(): [PG ID 2 PG GUID 3 Rank 0] Process group watchdog thread terminated with exception: [Rank 0] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=11, OpType=GATHER, NumelIn=16000, NumelOut=32000, Timeout(ms)=600000) ran for 600012 milliseconds before timing out.
Exception raised from checkTimeout at /pytorch/torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:632 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x96 (0x7f802576c1b6 in /opt/conda/lib/python3.11/site-packages/torch/lib/libc10.so)
frame #1: c10d::ProcessGroupNCCL::WorkNCCL::checkTimeout(std::optional<std::chrono::duration<long, std::ratio<1l, 1000l> > >) + 0x2b4 (0x7f7fd3a1bc74 in /opt/conda/lib/python3.11/site-packages/torch/lib/libtorch_cuda.so)
frame #2: c10d::ProcessGroupNCCL::watchdogHandler() + 0x890 (0x7f7fd3a1d7d0 in /opt/conda/lib/python3.11/site-packages/torch/lib/libtorch_cuda.so)
frame #3: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x14d (0x7f7fd3a1e6ed in /opt/conda/lib/python3.11/site-packages/torch/lib/libtorch_cuda.so)
frame #4: + 0x145c0 (0x7f80260615c0 in /opt/conda/lib/python3.11/site-packages/torch/lib/libtorch.so)
frame #5: + 0x94ac3 (0x7f80264d7ac3 in /lib/x86_64-linux-gnu/libc.so.6)
frame #6: + 0x126850 (0x7f8026569850 in /lib/x86_64-linux-gnu/libc.so.6)
Exception raised from ncclCommWatchdog at /pytorch/torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1901 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x96 (0x7f802576c1b6 in /opt/conda/lib/python3.11/site-packages/torch/lib/libc10.so)
frame #1: + 0xe5c6fc (0x7f7fd36796fc in /opt/conda/lib/python3.11/site-packages/torch/lib/libtorch_cuda.so)
frame #2: + 0x145c0 (0x7f80260615c0 in /opt/conda/lib/python3.11/site-packages/torch/lib/libtorch.so)
frame #3: + 0x94ac3 (0x7f80264d7ac3 in /lib/x86_64-linux-gnu/libc.so.6)
frame #4: + 0x126850 (0x7f8026569850 in /lib/x86_64-linux-gnu/libc.so.6)
/opt/conda/lib/python3.11/multiprocessing/resource_tracker.py:254: UserWarning: resource_tracker: There appear to be 1 leaked shared_memory objects to clean up at shutdown
warnings.warn('resource_tracker: There appear to be %d '
Aborted (core dumped)
Before submitting a new issue...
- Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.