pytorch - 💡(How to fix) Fix DistributedDataParallel + Conv2d = CUDA error: an illegal memory access was encountered [1 comments, 1 participants]

pytorch2026-04-13 21:17:48

ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

GitHub issue URL

Helpful · Quick feedback

GitHub stats

pytorch/pytorch#180280•Fetched 2026-04-15 06:18:52

View on GitHub

Comments

Participants

Timeline

Reactions

Author

Douvahkiin

Participants

Douvahkiin

Timeline (top)

subscribed ×14mentioned ×13labeled ×2closed ×1

Error Message

python Example_edit.py

/root/Documents/ES6D/lib/transformations.py:1912: UserWarning: failed to import module _transformations warnings.warn('failed to import module %s' % name) world size 2 /root/Documents/ES6D/lib/transformations.py:1912: UserWarning: failed to import module _transformations warnings.warn('failed to import module %s' % name) /root/Documents/ES6D/lib/transformations.py:1912: UserWarning: failed to import module _transformations warnings.warn('failed to import module %s' % name) init finished init finished [rank0]:[E414 05:10:32.178007984 ProcessGroupNCCL.cpp:1899] [PG ID 0 PG GUID 0(default_pg) Rank 0] Process group watchdog thread terminated with exception: CUDA error: an illegal memory access was encountered CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. For debugging consider passing CUDA_LAUNCH_BLOCKING=1 Compile with TORCH_USE_CUDA_DSA to enable device-side assertions.

Exception raised from c10_cuda_check_implementation at /pytorch/c10/cuda/CUDAException.cpp:43 (most recent call first): frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0x98 (0x7f9d03d785e8 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so) frame #1: c10::detail::torchCheckFail(char const*, char const*, unsigned int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) + 0xe0 (0x7f9d03d0d4a2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so) frame #2: c10::cuda::c10_cuda_check_implementation(int, char const*, char const*, int, bool) + 0x3c2 (0x7f9d0417a422 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10_cuda.so) frame #3: c10d::ProcessGroupNCCL::WorkNCCL::finishedGPUExecutionInternal() const + 0x56 (0x7f9c999475a6 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) frame #4: c10d::ProcessGroupNCCL::WorkNCCL::isCompleted() + 0x70 (0x7f9c99957840 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) frame #5: c10d::ProcessGroupNCCL::watchdogHandler() + 0x782 (0x7f9c999593d2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) frame #6: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x14d (0x7f9c9995afdd in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) frame #7: <unknown function> + 0xef5e4 (0x7f9c89d735e4 in /root/miniconda3/envs/es6d/bin/../lib/libstdc++.so.6) frame #8: <unknown function> + 0x94ac3 (0x7f9d04ad7ac3 in /usr/lib/x86_64-linux-gnu/libc.so.6) frame #9: clone + 0x44 (0x7f9d04b68a04 in /usr/lib/x86_64-linux-gnu/libc.so.6)

terminate called after throwing an instance of 'c10::DistBackendError' [rank1]:[E414 05:10:32.178353149 ProcessGroupNCCL.cpp:1899] [PG ID 0 PG GUID 0(default_pg) Rank 1] Process group watchdog thread terminated with exception: CUDA error: an illegal memory access was encountered CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. For debugging consider passing CUDA_LAUNCH_BLOCKING=1 Compile with TORCH_USE_CUDA_DSA to enable device-side assertions.

Exception raised from c10_cuda_check_implementation at /pytorch/c10/cuda/CUDAException.cpp:43 (most recent call first): frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0x98 (0x7f10d95da5e8 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so) frame #1: c10::detail::torchCheckFail(char const*, char const*, unsigned int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) + 0xe0 (0x7f10d956f4a2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so) frame #2: c10::cuda::c10_cuda_check_implementation(int, char const*, char const*, int, bool) + 0x3c2 (0x7f11447a5422 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10_cuda.so) frame #3: c10d::ProcessGroupNCCL::WorkNCCL::finishedGPUExecutionInternal() const + 0x56 (0x7f10da3475a6 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) frame #4: c10d::ProcessGroupNCCL::WorkNCCL::isCompleted() + 0x70 (0x7f10da357840 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) frame #5: c10d::ProcessGroupNCCL::watchdogHandler() + 0x782 (0x7f10da3593d2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) frame #6: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x14d (0x7f10da35afdd in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) frame #7: <unknown function> + 0xef5e4 (0x7f10ca5735e4 in /root/miniconda3/envs/es6d/bin/../lib/libstdc++.so.6) frame #8: <unknown function> + 0x94ac3 (0x7f11453edac3 in /usr/lib/x86_64-linux-gnu/libc.so.6) frame #9: clone + 0x44 (0x7f114547ea04 in /usr/lib/x86_64-linux-gnu/libc.so.6)

terminate called after throwing an instance of 'c10::DistBackendError' what(): [PG ID 0 PG GUID 0(default_pg) Rank 0] Process group watchdog thread terminated with exception: CUDA error: an illegal memory access was encountered CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. For debugging consider passing CUDA_LAUNCH_BLOCKING=1 Compile with TORCH_USE_CUDA_DSA to enable device-side assertions.

Exception raised from ncclCommWatchdog at /pytorch/torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1905 (most recent call first): frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0x98 (0x7f9d03d785e8 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so) frame #1: <unknown function> + 0xcc7b9e (0x7f9c99929b9e in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) frame #2: <unknown function> + 0x9165ed (0x7f9c995785ed in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) frame #3: <unknown function> + 0xef5e4 (0x7f9c89d735e4 in /root/miniconda3/envs/es6d/bin/../lib/libstdc++.so.6) frame #4: <unknown function> + 0x94ac3 (0x7f9d04ad7ac3 in /usr/lib/x86_64-linux-gnu/libc.so.6) frame #5: clone + 0x44 (0x7f9d04b68a04 in /usr/lib/x86_64-linux-gnu/libc.so.6)

what(): [PG ID 0 PG GUID 0(default_pg) Rank 1] Process group watchdog thread terminated with exception: CUDA error: an illegal memory access was encountered CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. For debugging consider passing CUDA_LAUNCH_BLOCKING=1 Compile with TORCH_USE_CUDA_DSA to enable device-side assertions.

Exception raised from c10_cuda_check_implementation at /pytorch/c10/cuda/CUDAException.cpp:43 (most recent call first): frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0x98 (0x7f10d95da5e8 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so) frame #1: c10::detail::torchCheckFail(char const*, char const*, unsigned int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) + 0xe0 (0x7f10d956f4a2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so) frame #2: c10::cuda::c10_cuda_check_implementation(int, char const*, char const*, int, bool) + 0x3c2 (0x7f11447a5422 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10_cuda.so) frame #3: c10d::ProcessGroupNCCL::WorkNCCL::finishedGPUExecutionInternal() const + 0x56 (0x7f10da3475a6 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) frame #4: c10d::ProcessGroupNCCL::WorkNCCL::isCompleted() + 0x70 (0x7f10da357840 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) frame #5: c10d::ProcessGroupNCCL::watchdogHandler() + 0x782 (0x7f10da3593d2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) frame #6: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x14d (0x7f10da35afdd in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) frame #7: <unknown function> + 0xef5e4 (0x7f10ca5735e4 in /root/miniconda3/envs/es6d/bin/../lib/libstdc++.so.6) frame #8: <unknown function> + 0x94ac3 (0x7f11453edac3 in /usr/lib/x86_64-linux-gnu/libc.so.6) frame #9: clone + 0x44 (0x7f114547ea04 in /usr/lib/x86_64-linux-gnu/libc.so.6)

Exception raised from ncclCommWatchdog at /pytorch/torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1905 (most recent call first): frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0x98 (0x7f10d95da5e8 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so) frame #1: <unknown function> + 0xcc7b9e (0x7f10da329b9e in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) frame #2: <unknown function> + 0x9165ed (0x7f10d9f785ed in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) frame #3: <unknown function> + 0xef5e4 (0x7f10ca5735e4 in /root/miniconda3/envs/es6d/bin/../lib/libstdc++.so.6) frame #4: <unknown function> + 0x94ac3 (0x7f11453edac3 in /usr/lib/x86_64-linux-gnu/libc.so.6) frame #5: clone + 0x44 (0x7f114547ea04 in /usr/lib/x86_64-linux-gnu/libc.so.6)

W0414 05:10:33.084000 4905 site-packages/torch/multiprocessing/spawn.py:169] Terminating process 4957 via signal SIGTERM Traceback (most recent call last): File "/root/Documents/ES6D/AI/Example_edit.py", line 61, in <module> mp.spawn(main, args=(world_size, 10, 500, 8), nprocs=world_size) File "/root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/multiprocessing/spawn.py", line 340, in spawn return start_processes(fn, args, nprocs, join, daemon, start_method="spawn") File "/root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/multiprocessing/spawn.py", line 296, in start_processes while not context.join(): File "/root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/multiprocessing/spawn.py", line 196, in join raise ProcessExitedException( torch.multiprocessing.spawn.ProcessExitedException: process 1 terminated with signal SIGABRT

Fix Action

Fix / Workaround

CPU: Architecture: x86_64 CPU op-mode(s): 32-bit, 64-bit Address sizes: 52 bits physical, 57 bits virtual Byte Order: Little Endian CPU(s): 208 On-line CPU(s) list: 0-207 Vendor ID: GenuineIntel Model name: Intel(R) Xeon(R) Platinum 8470Q CPU family: 6 Model: 143 Thread(s) per core: 2 Core(s) per socket: 52 Socket(s): 2 Stepping: 8 Frequency boost: enabled CPU max MHz: 2101.0000 CPU min MHz: 800.0000 BogoMIPS: 4200.00 Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf tsc_known_freq pni pclmulqdq dtes64 ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault epb cat_l3 cat_l2 cdp_l3 invpcid_single intel_ppin cdp_l2 ssbd mba ibrs ibpb stibp ibrs_enhanced tpr_shadow vnmi flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb intel_pt avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local split_lock_detect avx_vnni avx512_bf16 wbnoinvd dtherm ida arat pln pts avx512vbmi umip pku ospke waitpkg avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg tme avx512_vpopcntdq la57 rdpid bus_lock_detect cldemote movdiri movdir64b enqcmd fsrm md_clear serialize tsxldtrk pconfig arch_lbr amx_bf16 avx512_fp16 amx_tile amx_int8 flush_l1d arch_capabilities Virtualization: VT-x L1d cache: 4.9 MiB (104 instances) L1i cache: 3.3 MiB (104 instances) L2 cache: 208 MiB (104 instances) L3 cache: 210 MiB (2 instances) NUMA node(s): 2 NUMA node0 CPU(s): 0-51,104-155 NUMA node1 CPU(s): 52-103,156-207 Vulnerability Itlb multihit: Not affected Vulnerability L1tf: Not affected Vulnerability Mds: Not affected Vulnerability Meltdown: Not affected Vulnerability Mmio stale data: Not affected Vulnerability Retbleed: 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 IBRS, IBPB conditional, RSB filling, PBRSB-eIBRS SW sequence Vulnerability Srbds: Not affected Vulnerability Tsx async abort: Not affected

Code Example

import sys
import os
sys.path.insert(0, os.getcwd())
import torch
import torch.nn.functional as F
from torch.utils.data import Dataset, DataLoader

import torch.multiprocessing as mp
from torch.utils.data.distributed import DistributedSampler
from torch.nn.parallel import DistributedDataParallel as DDP
from torch.distributed import init_process_group, destroy_process_group
from models import ES6D as pose_net

def ddp_setup(rank, world_size):
    """
    Args:
        rank: Unique identifier of each process
        world_size: Total number of processes
    """
    os.environ["MASTER_ADDR"] = "localhost"
    os.environ["MASTER_PORT"] = "12355"
    torch.cuda.set_device(rank)
    init_process_group(backend="nccl", rank=rank, world_size=world_size)

class Trainer:
    def __init__(
        self,
        model: torch.nn.Module,
        optimizer: torch.optim.Optimizer,
        gpu_id: int,
        save_every: int,
    ) -> None:
        self.gpu_id = gpu_id
        self.model = model.to(gpu_id)
        self.optimizer = optimizer
        self.save_every = save_every
        self.model = DDP(self.model, device_ids=[gpu_id])
    def myprint(self):
        print("init finished")


def load_train_objs():
    train_set = 1
    model = torch.nn.Linear(20, 1)  # load your model
    # model = torch.nn.Conv2d(2048, 640, kernel_size=1)
    optimizer = torch.optim.SGD(model.parameters(), lr=1e-3)
    return train_set, model, optimizer


def main(rank: int, world_size: int, save_every: int, total_epochs: int, batch_size: int):
    ddp_setup(rank, world_size)
    dataset, model, optimizer = load_train_objs()
    trainer = Trainer(model, optimizer, rank, save_every)
    trainer.myprint()
    destroy_process_group()


if __name__ == "__main__":
    world_size = torch.cuda.device_count()
    print("world size ", world_size)
    mp.spawn(main, args=(world_size, 10, 500, 8), nprocs=world_size)

---

# python Example_edit.py
/root/Documents/ES6D/lib/transformations.py:1912: UserWarning: failed to import module _transformations
  warnings.warn('failed to import module %s' % name)
world size  2
/root/Documents/ES6D/lib/transformations.py:1912: UserWarning: failed to import module _transformations
  warnings.warn('failed to import module %s' % name)
/root/Documents/ES6D/lib/transformations.py:1912: UserWarning: failed to import module _transformations
  warnings.warn('failed to import module %s' % name)
init finished
init finished

---

import sys
import os
sys.path.insert(0, os.getcwd())
import torch
import torch.nn.functional as F
from torch.utils.data import Dataset, DataLoader

import torch.multiprocessing as mp
from torch.utils.data.distributed import DistributedSampler
from torch.nn.parallel import DistributedDataParallel as DDP
from torch.distributed import init_process_group, destroy_process_group
from models import ES6D as pose_net

def ddp_setup(rank, world_size):
    """
    Args:
        rank: Unique identifier of each process
        world_size: Total number of processes
    """
    os.environ["MASTER_ADDR"] = "localhost"
    os.environ["MASTER_PORT"] = "12355"
    torch.cuda.set_device(rank)
    init_process_group(backend="nccl", rank=rank, world_size=world_size)

class Trainer:
    def __init__(
        self,
        model: torch.nn.Module,
        optimizer: torch.optim.Optimizer,
        gpu_id: int,
        save_every: int,
    ) -> None:
        self.gpu_id = gpu_id
        self.model = model.to(gpu_id)
        self.optimizer = optimizer
        self.save_every = save_every
        self.model = DDP(self.model, device_ids=[gpu_id])
    def myprint(self):
        print("init finished")


def load_train_objs():
    train_set = 1
    # model = torch.nn.Linear(20, 1)  # load your model
    model = torch.nn.Conv2d(2048, 640, kernel_size=1)
    optimizer = torch.optim.SGD(model.parameters(), lr=1e-3)
    return train_set, model, optimizer


def main(rank: int, world_size: int, save_every: int, total_epochs: int, batch_size: int):
    ddp_setup(rank, world_size)
    dataset, model, optimizer = load_train_objs()
    trainer = Trainer(model, optimizer, rank, save_every)
    trainer.myprint()
    destroy_process_group()


if __name__ == "__main__":
    world_size = torch.cuda.device_count()
    print("world size ", world_size)
    mp.spawn(main, args=(world_size, 10, 500, 8), nprocs=world_size)

---

# python Example_edit.py
/root/Documents/ES6D/lib/transformations.py:1912: UserWarning: failed to import module _transformations
  warnings.warn('failed to import module %s' % name)
world size  2
/root/Documents/ES6D/lib/transformations.py:1912: UserWarning: failed to import module _transformations
  warnings.warn('failed to import module %s' % name)
/root/Documents/ES6D/lib/transformations.py:1912: UserWarning: failed to import module _transformations
  warnings.warn('failed to import module %s' % name)
init finished
init finished
[rank0]:[E414 05:10:32.178007984 ProcessGroupNCCL.cpp:1899] [PG ID 0 PG GUID 0(default_pg) Rank 0] Process group watchdog thread terminated with exception: CUDA error: an illegal memory access was encountered
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.

Exception raised from c10_cuda_check_implementation at /pytorch/c10/cuda/CUDAException.cpp:43 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0x98 (0x7f9d03d785e8 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so)
frame #1: c10::detail::torchCheckFail(char const*, char const*, unsigned int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) + 0xe0 (0x7f9d03d0d4a2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so)
frame #2: c10::cuda::c10_cuda_check_implementation(int, char const*, char const*, int, bool) + 0x3c2 (0x7f9d0417a422 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10_cuda.so)
frame #3: c10d::ProcessGroupNCCL::WorkNCCL::finishedGPUExecutionInternal() const + 0x56 (0x7f9c999475a6 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #4: c10d::ProcessGroupNCCL::WorkNCCL::isCompleted() + 0x70 (0x7f9c99957840 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #5: c10d::ProcessGroupNCCL::watchdogHandler() + 0x782 (0x7f9c999593d2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #6: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x14d (0x7f9c9995afdd in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #7: <unknown function> + 0xef5e4 (0x7f9c89d735e4 in /root/miniconda3/envs/es6d/bin/../lib/libstdc++.so.6)
frame #8: <unknown function> + 0x94ac3 (0x7f9d04ad7ac3 in /usr/lib/x86_64-linux-gnu/libc.so.6)
frame #9: clone + 0x44 (0x7f9d04b68a04 in /usr/lib/x86_64-linux-gnu/libc.so.6)

terminate called after throwing an instance of 'c10::DistBackendError'
[rank1]:[E414 05:10:32.178353149 ProcessGroupNCCL.cpp:1899] [PG ID 0 PG GUID 0(default_pg) Rank 1] Process group watchdog thread terminated with exception: CUDA error: an illegal memory access was encountered
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.

Exception raised from c10_cuda_check_implementation at /pytorch/c10/cuda/CUDAException.cpp:43 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0x98 (0x7f10d95da5e8 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so)
frame #1: c10::detail::torchCheckFail(char const*, char const*, unsigned int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) + 0xe0 (0x7f10d956f4a2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so)
frame #2: c10::cuda::c10_cuda_check_implementation(int, char const*, char const*, int, bool) + 0x3c2 (0x7f11447a5422 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10_cuda.so)
frame #3: c10d::ProcessGroupNCCL::WorkNCCL::finishedGPUExecutionInternal() const + 0x56 (0x7f10da3475a6 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #4: c10d::ProcessGroupNCCL::WorkNCCL::isCompleted() + 0x70 (0x7f10da357840 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #5: c10d::ProcessGroupNCCL::watchdogHandler() + 0x782 (0x7f10da3593d2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #6: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x14d (0x7f10da35afdd in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #7: <unknown function> + 0xef5e4 (0x7f10ca5735e4 in /root/miniconda3/envs/es6d/bin/../lib/libstdc++.so.6)
frame #8: <unknown function> + 0x94ac3 (0x7f11453edac3 in /usr/lib/x86_64-linux-gnu/libc.so.6)
frame #9: clone + 0x44 (0x7f114547ea04 in /usr/lib/x86_64-linux-gnu/libc.so.6)

terminate called after throwing an instance of 'c10::DistBackendError'
  what():  [PG ID 0 PG GUID 0(default_pg) Rank 0] Process group watchdog thread terminated with exception: CUDA error: an illegal memory access was encountered
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.

Exception raised from c10_cuda_check_implementation at /pytorch/c10/cuda/CUDAException.cpp:43 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0x98 (0x7f9d03d785e8 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so)
frame #1: c10::detail::torchCheckFail(char const*, char const*, unsigned int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) + 0xe0 (0x7f9d03d0d4a2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so)
frame #2: c10::cuda::c10_cuda_check_implementation(int, char const*, char const*, int, bool) + 0x3c2 (0x7f9d0417a422 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10_cuda.so)
frame #3: c10d::ProcessGroupNCCL::WorkNCCL::finishedGPUExecutionInternal() const + 0x56 (0x7f9c999475a6 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #4: c10d::ProcessGroupNCCL::WorkNCCL::isCompleted() + 0x70 (0x7f9c99957840 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #5: c10d::ProcessGroupNCCL::watchdogHandler() + 0x782 (0x7f9c999593d2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #6: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x14d (0x7f9c9995afdd in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #7: <unknown function> + 0xef5e4 (0x7f9c89d735e4 in /root/miniconda3/envs/es6d/bin/../lib/libstdc++.so.6)
frame #8: <unknown function> + 0x94ac3 (0x7f9d04ad7ac3 in /usr/lib/x86_64-linux-gnu/libc.so.6)
frame #9: clone + 0x44 (0x7f9d04b68a04 in /usr/lib/x86_64-linux-gnu/libc.so.6)

Exception raised from ncclCommWatchdog at /pytorch/torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1905 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0x98 (0x7f9d03d785e8 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so)
frame #1: <unknown function> + 0xcc7b9e (0x7f9c99929b9e in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #2: <unknown function> + 0x9165ed (0x7f9c995785ed in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #3: <unknown function> + 0xef5e4 (0x7f9c89d735e4 in /root/miniconda3/envs/es6d/bin/../lib/libstdc++.so.6)
frame #4: <unknown function> + 0x94ac3 (0x7f9d04ad7ac3 in /usr/lib/x86_64-linux-gnu/libc.so.6)
frame #5: clone + 0x44 (0x7f9d04b68a04 in /usr/lib/x86_64-linux-gnu/libc.so.6)

  what():  [PG ID 0 PG GUID 0(default_pg) Rank 1] Process group watchdog thread terminated with exception: CUDA error: an illegal memory access was encountered
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.

Exception raised from c10_cuda_check_implementation at /pytorch/c10/cuda/CUDAException.cpp:43 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0x98 (0x7f10d95da5e8 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so)
frame #1: c10::detail::torchCheckFail(char const*, char const*, unsigned int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) + 0xe0 (0x7f10d956f4a2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so)
frame #2: c10::cuda::c10_cuda_check_implementation(int, char const*, char const*, int, bool) + 0x3c2 (0x7f11447a5422 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10_cuda.so)
frame #3: c10d::ProcessGroupNCCL::WorkNCCL::finishedGPUExecutionInternal() const + 0x56 (0x7f10da3475a6 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #4: c10d::ProcessGroupNCCL::WorkNCCL::isCompleted() + 0x70 (0x7f10da357840 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #5: c10d::ProcessGroupNCCL::watchdogHandler() + 0x782 (0x7f10da3593d2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #6: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x14d (0x7f10da35afdd in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #7: <unknown function> + 0xef5e4 (0x7f10ca5735e4 in /root/miniconda3/envs/es6d/bin/../lib/libstdc++.so.6)
frame #8: <unknown function> + 0x94ac3 (0x7f11453edac3 in /usr/lib/x86_64-linux-gnu/libc.so.6)
frame #9: clone + 0x44 (0x7f114547ea04 in /usr/lib/x86_64-linux-gnu/libc.so.6)

Exception raised from ncclCommWatchdog at /pytorch/torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1905 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0x98 (0x7f10d95da5e8 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so)
frame #1: <unknown function> + 0xcc7b9e (0x7f10da329b9e in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #2: <unknown function> + 0x9165ed (0x7f10d9f785ed in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #3: <unknown function> + 0xef5e4 (0x7f10ca5735e4 in /root/miniconda3/envs/es6d/bin/../lib/libstdc++.so.6)
frame #4: <unknown function> + 0x94ac3 (0x7f11453edac3 in /usr/lib/x86_64-linux-gnu/libc.so.6)
frame #5: clone + 0x44 (0x7f114547ea04 in /usr/lib/x86_64-linux-gnu/libc.so.6)

W0414 05:10:33.084000 4905 site-packages/torch/multiprocessing/spawn.py:169] Terminating process 4957 via signal SIGTERM
Traceback (most recent call last):
  File "/root/Documents/ES6D/AI/Example_edit.py", line 61, in <module>
    mp.spawn(main, args=(world_size, 10, 500, 8), nprocs=world_size)
  File "/root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/multiprocessing/spawn.py", line 340, in spawn
    return start_processes(fn, args, nprocs, join, daemon, start_method="spawn")
  File "/root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/multiprocessing/spawn.py", line 296, in start_processes
    while not context.join():
  File "/root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/multiprocessing/spawn.py", line 196, in join
    raise ProcessExitedException(
torch.multiprocessing.spawn.ProcessExitedException: process 1 terminated with signal SIGABRT

RAW_BUFFERClick to expand / collapse

🐛 Describe the bug

OS: ubuntu 22.04 LTS GPU: NVIDIA RTX PRO 6000 & NVIDIA RTX PRO 5090 Env: Container When I ran this script:

import sys
import os
sys.path.insert(0, os.getcwd())
import torch
import torch.nn.functional as F
from torch.utils.data import Dataset, DataLoader

import torch.multiprocessing as mp
from torch.utils.data.distributed import DistributedSampler
from torch.nn.parallel import DistributedDataParallel as DDP
from torch.distributed import init_process_group, destroy_process_group
from models import ES6D as pose_net

def ddp_setup(rank, world_size):
    """
    Args:
        rank: Unique identifier of each process
        world_size: Total number of processes
    """
    os.environ["MASTER_ADDR"] = "localhost"
    os.environ["MASTER_PORT"] = "12355"
    torch.cuda.set_device(rank)
    init_process_group(backend="nccl", rank=rank, world_size=world_size)

class Trainer:
    def __init__(
        self,
        model: torch.nn.Module,
        optimizer: torch.optim.Optimizer,
        gpu_id: int,
        save_every: int,
    ) -> None:
        self.gpu_id = gpu_id
        self.model = model.to(gpu_id)
        self.optimizer = optimizer
        self.save_every = save_every
        self.model = DDP(self.model, device_ids=[gpu_id])
    def myprint(self):
        print("init finished")


def load_train_objs():
    train_set = 1
    model = torch.nn.Linear(20, 1)  # load your model
    # model = torch.nn.Conv2d(2048, 640, kernel_size=1)
    optimizer = torch.optim.SGD(model.parameters(), lr=1e-3)
    return train_set, model, optimizer


def main(rank: int, world_size: int, save_every: int, total_epochs: int, batch_size: int):
    ddp_setup(rank, world_size)
    dataset, model, optimizer = load_train_objs()
    trainer = Trainer(model, optimizer, rank, save_every)
    trainer.myprint()
    destroy_process_group()


if __name__ == "__main__":
    world_size = torch.cuda.device_count()
    print("world size ", world_size)
    mp.spawn(main, args=(world_size, 10, 500, 8), nprocs=world_size)

It ran as expected:

# python Example_edit.py
/root/Documents/ES6D/lib/transformations.py:1912: UserWarning: failed to import module _transformations
  warnings.warn('failed to import module %s' % name)
world size  2
/root/Documents/ES6D/lib/transformations.py:1912: UserWarning: failed to import module _transformations
  warnings.warn('failed to import module %s' % name)
/root/Documents/ES6D/lib/transformations.py:1912: UserWarning: failed to import module _transformations
  warnings.warn('failed to import module %s' % name)
init finished
init finished

But when I changed model = torch.nn.Linear(20, 1) to model = torch.nn.Conv2d(2048, 640, kernel_size=1):

import sys
import os
sys.path.insert(0, os.getcwd())
import torch
import torch.nn.functional as F
from torch.utils.data import Dataset, DataLoader

import torch.multiprocessing as mp
from torch.utils.data.distributed import DistributedSampler
from torch.nn.parallel import DistributedDataParallel as DDP
from torch.distributed import init_process_group, destroy_process_group
from models import ES6D as pose_net

def ddp_setup(rank, world_size):
    """
    Args:
        rank: Unique identifier of each process
        world_size: Total number of processes
    """
    os.environ["MASTER_ADDR"] = "localhost"
    os.environ["MASTER_PORT"] = "12355"
    torch.cuda.set_device(rank)
    init_process_group(backend="nccl", rank=rank, world_size=world_size)

class Trainer:
    def __init__(
        self,
        model: torch.nn.Module,
        optimizer: torch.optim.Optimizer,
        gpu_id: int,
        save_every: int,
    ) -> None:
        self.gpu_id = gpu_id
        self.model = model.to(gpu_id)
        self.optimizer = optimizer
        self.save_every = save_every
        self.model = DDP(self.model, device_ids=[gpu_id])
    def myprint(self):
        print("init finished")


def load_train_objs():
    train_set = 1
    # model = torch.nn.Linear(20, 1)  # load your model
    model = torch.nn.Conv2d(2048, 640, kernel_size=1)
    optimizer = torch.optim.SGD(model.parameters(), lr=1e-3)
    return train_set, model, optimizer


def main(rank: int, world_size: int, save_every: int, total_epochs: int, batch_size: int):
    ddp_setup(rank, world_size)
    dataset, model, optimizer = load_train_objs()
    trainer = Trainer(model, optimizer, rank, save_every)
    trainer.myprint()
    destroy_process_group()


if __name__ == "__main__":
    world_size = torch.cuda.device_count()
    print("world size ", world_size)
    mp.spawn(main, args=(world_size, 10, 500, 8), nprocs=world_size)

It got this error:

# python Example_edit.py
/root/Documents/ES6D/lib/transformations.py:1912: UserWarning: failed to import module _transformations
  warnings.warn('failed to import module %s' % name)
world size  2
/root/Documents/ES6D/lib/transformations.py:1912: UserWarning: failed to import module _transformations
  warnings.warn('failed to import module %s' % name)
/root/Documents/ES6D/lib/transformations.py:1912: UserWarning: failed to import module _transformations
  warnings.warn('failed to import module %s' % name)
init finished
init finished
[rank0]:[E414 05:10:32.178007984 ProcessGroupNCCL.cpp:1899] [PG ID 0 PG GUID 0(default_pg) Rank 0] Process group watchdog thread terminated with exception: CUDA error: an illegal memory access was encountered
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.

Exception raised from c10_cuda_check_implementation at /pytorch/c10/cuda/CUDAException.cpp:43 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0x98 (0x7f9d03d785e8 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so)
frame #1: c10::detail::torchCheckFail(char const*, char const*, unsigned int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) + 0xe0 (0x7f9d03d0d4a2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so)
frame #2: c10::cuda::c10_cuda_check_implementation(int, char const*, char const*, int, bool) + 0x3c2 (0x7f9d0417a422 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10_cuda.so)
frame #3: c10d::ProcessGroupNCCL::WorkNCCL::finishedGPUExecutionInternal() const + 0x56 (0x7f9c999475a6 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #4: c10d::ProcessGroupNCCL::WorkNCCL::isCompleted() + 0x70 (0x7f9c99957840 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #5: c10d::ProcessGroupNCCL::watchdogHandler() + 0x782 (0x7f9c999593d2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #6: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x14d (0x7f9c9995afdd in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #7: <unknown function> + 0xef5e4 (0x7f9c89d735e4 in /root/miniconda3/envs/es6d/bin/../lib/libstdc++.so.6)
frame #8: <unknown function> + 0x94ac3 (0x7f9d04ad7ac3 in /usr/lib/x86_64-linux-gnu/libc.so.6)
frame #9: clone + 0x44 (0x7f9d04b68a04 in /usr/lib/x86_64-linux-gnu/libc.so.6)

terminate called after throwing an instance of 'c10::DistBackendError'
[rank1]:[E414 05:10:32.178353149 ProcessGroupNCCL.cpp:1899] [PG ID 0 PG GUID 0(default_pg) Rank 1] Process group watchdog thread terminated with exception: CUDA error: an illegal memory access was encountered
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.

Exception raised from c10_cuda_check_implementation at /pytorch/c10/cuda/CUDAException.cpp:43 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0x98 (0x7f10d95da5e8 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so)
frame #1: c10::detail::torchCheckFail(char const*, char const*, unsigned int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) + 0xe0 (0x7f10d956f4a2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so)
frame #2: c10::cuda::c10_cuda_check_implementation(int, char const*, char const*, int, bool) + 0x3c2 (0x7f11447a5422 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10_cuda.so)
frame #3: c10d::ProcessGroupNCCL::WorkNCCL::finishedGPUExecutionInternal() const + 0x56 (0x7f10da3475a6 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #4: c10d::ProcessGroupNCCL::WorkNCCL::isCompleted() + 0x70 (0x7f10da357840 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #5: c10d::ProcessGroupNCCL::watchdogHandler() + 0x782 (0x7f10da3593d2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #6: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x14d (0x7f10da35afdd in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #7: <unknown function> + 0xef5e4 (0x7f10ca5735e4 in /root/miniconda3/envs/es6d/bin/../lib/libstdc++.so.6)
frame #8: <unknown function> + 0x94ac3 (0x7f11453edac3 in /usr/lib/x86_64-linux-gnu/libc.so.6)
frame #9: clone + 0x44 (0x7f114547ea04 in /usr/lib/x86_64-linux-gnu/libc.so.6)

terminate called after throwing an instance of 'c10::DistBackendError'
  what():  [PG ID 0 PG GUID 0(default_pg) Rank 0] Process group watchdog thread terminated with exception: CUDA error: an illegal memory access was encountered
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.

Exception raised from c10_cuda_check_implementation at /pytorch/c10/cuda/CUDAException.cpp:43 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0x98 (0x7f9d03d785e8 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so)
frame #1: c10::detail::torchCheckFail(char const*, char const*, unsigned int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) + 0xe0 (0x7f9d03d0d4a2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so)
frame #2: c10::cuda::c10_cuda_check_implementation(int, char const*, char const*, int, bool) + 0x3c2 (0x7f9d0417a422 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10_cuda.so)
frame #3: c10d::ProcessGroupNCCL::WorkNCCL::finishedGPUExecutionInternal() const + 0x56 (0x7f9c999475a6 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #4: c10d::ProcessGroupNCCL::WorkNCCL::isCompleted() + 0x70 (0x7f9c99957840 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #5: c10d::ProcessGroupNCCL::watchdogHandler() + 0x782 (0x7f9c999593d2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #6: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x14d (0x7f9c9995afdd in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #7: <unknown function> + 0xef5e4 (0x7f9c89d735e4 in /root/miniconda3/envs/es6d/bin/../lib/libstdc++.so.6)
frame #8: <unknown function> + 0x94ac3 (0x7f9d04ad7ac3 in /usr/lib/x86_64-linux-gnu/libc.so.6)
frame #9: clone + 0x44 (0x7f9d04b68a04 in /usr/lib/x86_64-linux-gnu/libc.so.6)

Exception raised from ncclCommWatchdog at /pytorch/torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1905 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0x98 (0x7f9d03d785e8 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so)
frame #1: <unknown function> + 0xcc7b9e (0x7f9c99929b9e in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #2: <unknown function> + 0x9165ed (0x7f9c995785ed in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #3: <unknown function> + 0xef5e4 (0x7f9c89d735e4 in /root/miniconda3/envs/es6d/bin/../lib/libstdc++.so.6)
frame #4: <unknown function> + 0x94ac3 (0x7f9d04ad7ac3 in /usr/lib/x86_64-linux-gnu/libc.so.6)
frame #5: clone + 0x44 (0x7f9d04b68a04 in /usr/lib/x86_64-linux-gnu/libc.so.6)

  what():  [PG ID 0 PG GUID 0(default_pg) Rank 1] Process group watchdog thread terminated with exception: CUDA error: an illegal memory access was encountered
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.

Exception raised from c10_cuda_check_implementation at /pytorch/c10/cuda/CUDAException.cpp:43 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0x98 (0x7f10d95da5e8 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so)
frame #1: c10::detail::torchCheckFail(char const*, char const*, unsigned int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) + 0xe0 (0x7f10d956f4a2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so)
frame #2: c10::cuda::c10_cuda_check_implementation(int, char const*, char const*, int, bool) + 0x3c2 (0x7f11447a5422 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10_cuda.so)
frame #3: c10d::ProcessGroupNCCL::WorkNCCL::finishedGPUExecutionInternal() const + 0x56 (0x7f10da3475a6 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #4: c10d::ProcessGroupNCCL::WorkNCCL::isCompleted() + 0x70 (0x7f10da357840 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #5: c10d::ProcessGroupNCCL::watchdogHandler() + 0x782 (0x7f10da3593d2 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #6: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x14d (0x7f10da35afdd in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #7: <unknown function> + 0xef5e4 (0x7f10ca5735e4 in /root/miniconda3/envs/es6d/bin/../lib/libstdc++.so.6)
frame #8: <unknown function> + 0x94ac3 (0x7f11453edac3 in /usr/lib/x86_64-linux-gnu/libc.so.6)
frame #9: clone + 0x44 (0x7f114547ea04 in /usr/lib/x86_64-linux-gnu/libc.so.6)

Exception raised from ncclCommWatchdog at /pytorch/torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1905 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0x98 (0x7f10d95da5e8 in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libc10.so)
frame #1: <unknown function> + 0xcc7b9e (0x7f10da329b9e in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #2: <unknown function> + 0x9165ed (0x7f10d9f785ed in /root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so)
frame #3: <unknown function> + 0xef5e4 (0x7f10ca5735e4 in /root/miniconda3/envs/es6d/bin/../lib/libstdc++.so.6)
frame #4: <unknown function> + 0x94ac3 (0x7f11453edac3 in /usr/lib/x86_64-linux-gnu/libc.so.6)
frame #5: clone + 0x44 (0x7f114547ea04 in /usr/lib/x86_64-linux-gnu/libc.so.6)

W0414 05:10:33.084000 4905 site-packages/torch/multiprocessing/spawn.py:169] Terminating process 4957 via signal SIGTERM
Traceback (most recent call last):
  File "/root/Documents/ES6D/AI/Example_edit.py", line 61, in <module>
    mp.spawn(main, args=(world_size, 10, 500, 8), nprocs=world_size)
  File "/root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/multiprocessing/spawn.py", line 340, in spawn
    return start_processes(fn, args, nprocs, join, daemon, start_method="spawn")
  File "/root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/multiprocessing/spawn.py", line 296, in start_processes
    while not context.join():
  File "/root/miniconda3/envs/es6d/lib/python3.10/site-packages/torch/multiprocessing/spawn.py", line 196, in join
    raise ProcessExitedException(
torch.multiprocessing.spawn.ProcessExitedException: process 1 terminated with signal SIGABRT

This issue happened on a cloud GPU server. When I executed it on my local machine (torch1.7), it didn't happen.

Versions

Collecting environment information... PyTorch version: 2.7.1+cu128 Is debug build: False CUDA used to build PyTorch: 12.8 ROCM used to build PyTorch: N/A

OS: Ubuntu 22.04.5 LTS (x86_64) GCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 Clang version: Could not collect CMake version: version 3.22.1 Libc version: glibc-2.35

Python version: 3.10.20 | packaged by conda-forge | (main, Mar 5 2026, 16:42:22) [GCC 14.3.0] (64-bit runtime) Python platform: Linux-5.15.0-78-generic-x86_64-with-glibc2.35 Is CUDA available: True CUDA runtime version: Could not collect CUDA_MODULE_LOADING set to: LAZY GPU models and configuration: GPU 0: NVIDIA RTX PRO 6000 Blackwell Server Edition GPU 1: NVIDIA RTX PRO 6000 Blackwell Server Edition

Nvidia driver version: 595.58.03 cuDNN version: Probably one of the following: /usr/lib/x86_64-linux-gnu/libcudnn.so.9.8.0 /usr/lib/x86_64-linux-gnu/libcudnn_adv.so.9.8.0 /usr/lib/x86_64-linux-gnu/libcudnn_cnn.so.9.8.0 /usr/lib/x86_64-linux-gnu/libcudnn_engines_precompiled.so.9.8.0 /usr/lib/x86_64-linux-gnu/libcudnn_engines_runtime_compiled.so.9.8.0 /usr/lib/x86_64-linux-gnu/libcudnn_graph.so.9.8.0 /usr/lib/x86_64-linux-gnu/libcudnn_heuristic.so.9.8.0 /usr/lib/x86_64-linux-gnu/libcudnn_ops.so.9.8.0 Is XPU available: False HIP runtime version: N/A MIOpen runtime version: N/A Is XNNPACK available: True Caching allocator config: N/A

Versions of relevant libraries: [pip3] numpy==2.2.6 [pip3] nvidia-cublas-cu12==12.8.3.14 [pip3] nvidia-cuda-cupti-cu12==12.8.57 [pip3] nvidia-cuda-nvrtc-cu12==12.8.61 [pip3] nvidia-cuda-runtime-cu12==12.8.57 [pip3] nvidia-cudnn-cu12==9.7.1.26 [pip3] nvidia-cufft-cu12==11.3.3.41 [pip3] nvidia-curand-cu12==10.3.9.55 [pip3] nvidia-cusolver-cu12==11.7.2.55 [pip3] nvidia-cusparse-cu12==12.5.7.53 [pip3] nvidia-cusparselt-cu12==0.6.3 [pip3] nvidia-nccl-cu12==2.26.2 [pip3] nvidia-nvjitlink-cu12==12.8.61 [pip3] nvidia-nvtx-cu12==12.8.55 [pip3] pytorch-msssim==1.0.0 [pip3] torch==2.7.1+cu128 [pip3] torchaudio==2.11.0+cu128 [pip3] torchvision==0.22.1+cu128 [pip3] triton==3.3.1 [conda] libopenvino-pytorch-frontend 2026.0.0 hecca717_1 conda-forge [conda] numpy 2.2.6 py310hefbff90_0 conda-forge [conda] nvidia-cublas-cu12 12.8.3.14 pypi_0 pypi [conda] nvidia-cuda-cupti-cu12 12.8.57 pypi_0 pypi [conda] nvidia-cuda-nvrtc-cu12 12.8.61 pypi_0 pypi [conda] nvidia-cuda-runtime-cu12 12.8.57 pypi_0 pypi [conda] nvidia-cudnn-cu12 9.7.1.26 pypi_0 pypi [conda] nvidia-cufft-cu12 11.3.3.41 pypi_0 pypi [conda] nvidia-curand-cu12 10.3.9.55 pypi_0 pypi [conda] nvidia-cusolver-cu12 11.7.2.55 pypi_0 pypi [conda] nvidia-cusparse-cu12 12.5.7.53 pypi_0 pypi [conda] nvidia-cusparselt-cu12 0.6.3 pypi_0 pypi [conda] nvidia-nccl-cu12 2.26.2 pypi_0 pypi [conda] nvidia-nvjitlink-cu12 12.8.61 pypi_0 pypi [conda] nvidia-nvtx-cu12 12.8.55 pypi_0 pypi [conda] pytorch-msssim 1.0.0 pypi_0 pypi [conda] tbb 2022.3.0 hb700be7_2 conda-forge [conda] torch 2.7.1+cu128 pypi_0 pypi [conda] torchaudio 2.11.0+cu128 pypi_0 pypi [conda] torchvision 0.22.1+cu128 pypi_0 pypi [conda] triton 3.3.1 pypi_0 pypi

cc @awgu @wanchaol @fegin @fduwjj @wz337 @wconstab @d4l3k @pragupta @msaroufim @dcci @aditvenk @xmfan @weifengpy

extent analysis

TL;DR

The issue is likely due to a CUDA error caused by an illegal memory access when using torch.nn.Conv2d with distributed data parallelism, and setting CUDA_LAUNCH_BLOCKING=1 may help with debugging.

Guidance

The error message suggests a CUDA error, which may be related to the specific GPU architecture or driver version.
Setting CUDA_LAUNCH_BLOCKING=1 as suggested in the error message may help with debugging by making CUDA kernel launches synchronous.
The issue may be specific to the torch.nn.Conv2d layer, as the error does not occur with torch.nn.Linear.
Checking the GPU memory usage and ensuring that the model fits within the available memory may also be helpful.
Consider updating the NVIDIA driver to the latest version, as the current version (595.58.03) may have known issues.

Example

No specific code example is provided, as the issue is likely related to the interaction between PyTorch, CUDA, and the specific GPU architecture.

Notes

The issue is specific to the cloud GPU server and does not occur on the local machine with PyTorch 1.7, suggesting a potential issue with the PyTorch version or the cloud environment.
The error message suggests a CUDA error, but the exact cause is unclear without further debugging.

Recommendation

Apply the workaround by setting CUDA_LAUNCH_BLOCKING=1 to help with debugging, and consider updating the NVIDIA driver to the latest version. If the issue persists, further debugging and potentially updating PyTorch to a newer version may be necessary.

Vote matrix · Quick signals

Works

Did the solution work? Tap to confirm.

Easy Fix

Was it a quick fix?

Time Saver

Did it save you time?

Blocking

Was it severely blocking?

Common Issue

Are others likely hitting this too?

Flaky / Intermittent

Is it intermittent?

Verified / Reproducible

Can you reproduce it reliably?

#api #container setup #orchestration issue #cache issue #memory leak

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Data

Security

Network

Code

UI/UX

Text

System

Multimedia

Protocol

API

Engineering

pytorch - 💡(How to fix) Fix DistributedDataParallel + Conv2d = CUDA error: an illegal memory access was encountered [1 comments, 1 participants]

Recommended Tools

GitHub issue graph ai analysis

Error Message

python Example_edit.py

Fix Action

Fix / Workaround

Code Example

🐛 Describe the bug

Versions

extent analysis

TL;DR

Guidance

Example

Notes

Recommendation

Still need to ship something?

TRENDING

pytorch - 💡(How to fix) Fix DistributedDataParallel + Conv2d = CUDA error: an illegal memory access was encountered [1 comments, 1 participants]

Recommended Tools

GitHub issue graph ai analysis

Error Message

python Example_edit.py

Fix Action

Fix / Workaround

Code Example

🐛 Describe the bug

Versions

extent analysis

TL;DR

Guidance

Example

Notes

Recommendation

Still need to ship something?

RELATED_DISCOVERY

TRENDING