from .DummyData import DummyData

data_map = {
    "DummyData": DummyData
}

            

Reported by Pylint.

              from .DummyData import DummyData

data_map = {
    "DummyData": DummyData
}

            

Reported by Pylint.

              from .DummyModel import DummyModel

model_map = {
    "DummyModel": DummyModel
}

            

Reported by Pylint.

              from .DummyModel import DummyModel

model_map = {
    "DummyModel": DummyModel
}

            

Reported by Pylint.

              from .server import AverageBatchParameterServer, AverageParameterServer

server_map = {
    "AverageParameterServer": AverageParameterServer,
    "AverageBatchParameterServer": AverageBatchParameterServer
}

            

Reported by Pylint.

              from .server import AverageBatchParameterServer, AverageParameterServer

server_map = {
    "AverageParameterServer": AverageParameterServer,
    "AverageBatchParameterServer": AverageBatchParameterServer
}

            

Reported by Pylint.

              import torch.nn as nn


def cel(rank):
    r"""A function that creates a CrossEntropyLoss
    criterion for training.
    Args:
        rank (int): worker rank
    """

            

Reported by Pylint.

              import torch.nn as nn


def cel(rank):
    r"""A function that creates a CrossEntropyLoss
    criterion for training.
    Args:
        rank (int): worker rank
    """

            

Reported by Pylint.

              class BasicHookState:

    def __init__(self, cref, process_group):
        r"""
        A class that holds state information that is needed by the communication hook
        during the training algorithm.
        Args:
            cref (DdpTrainer): reference to the self keyword of the trainer instance
            process_group (ProcessGroup): distributed process group

            

Reported by Pylint.

              class BasicHookState:

    def __init__(self, cref, process_group):
        r"""
        A class that holds state information that is needed by the communication hook
        during the training algorithm.
        Args:
            cref (DdpTrainer): reference to the self keyword of the trainer instance
            process_group (ProcessGroup): distributed process group

            

Reported by Pylint.

              def basic_iteration_step(self, ddp_model, criterion, optimizer, hook_state, epoch, index, batch):
    r"""
    A function that performs an iteration of training.
    Args:
        ddp_model (nn.Module): distributed data parallel model
        criterion (nn.Module): loss function to measure model
        optimizer (optim.Optimizer): updates model parameters
        hook_state (object): ddp communication hook state object
        epoch (int): index of pass through the data

            

Reported by Pylint.

              def basic_iteration_step(self, ddp_model, criterion, optimizer, hook_state, epoch, index, batch):
    r"""
    A function that performs an iteration of training.
    Args:
        ddp_model (nn.Module): distributed data parallel model
        criterion (nn.Module): loss function to measure model
        optimizer (optim.Optimizer): updates model parameters
        hook_state (object): ddp communication hook state object
        epoch (int): index of pass through the data

            

Reported by Pylint.

              def preprocess_dummy_data(rank, data):
    r"""
    A function that moves the data from CPU to GPU
    for DummyData class.
    Args:
        rank (int): worker rank
        data (list): training examples
    """
    for i in range(len(data)):

            

Reported by Pylint.

                      rank (int): worker rank
        data (list): training examples
    """
    for i in range(len(data)):
        data[i][0] = data[i][0].cuda(rank)
        data[i][1] = data[i][1].cuda(rank)
    return data

            

Reported by Pylint.

              from setuptools import setup
from torch.utils.cpp_extension import CppExtension, BuildExtension

setup(name='benchmark_cpp_extension',
      ext_modules=[CppExtension('benchmark_cpp_extension', ['extension.cpp'])],
      cmdclass={'build_ext': BuildExtension})

            

Reported by Pylint.

              from setuptools import setup
from torch.utils.cpp_extension import CppExtension, BuildExtension

setup(name='benchmark_cpp_extension',
      ext_modules=[CppExtension('benchmark_cpp_extension', ['extension.cpp'])],
      cmdclass={'build_ext': BuildExtension})

            

Reported by Pylint.

                          /*atol*/ 1e-08,
            use_equalnan));
      } else {
        EXPECT_TRUE(l[i].toTensor().equal(r[i].toTensor()));
      }
    }
  }
}


            

Reported by FlawFinder.

                        /*atol*/ 1e-08,
          use_equalnan));
    } else {
      EXPECT_TRUE(expect.toTensor().equal(actual.toTensor()));
    }
    return;
  } else if (expect.isTuple()) {
    EXPECT_TRUE(actual.isTuple());
    auto lhs = expect.toTuple()->elements();

            

Reported by FlawFinder.

                  data_i64[i] = kJunkPattern64;
  }
  if (remaining_bytes > 0) {
    memcpy(data_i64 + int64_count, &kJunkPattern64, remaining_bytes);
  }
}

void* alloc_cpu(size_t nbytes) {
  if (nbytes == 0) {

            

Reported by FlawFinder.

                // NOLINTNEXTLINE(cppcoreguidelines-init-variables)
  void* data;
#ifdef __ANDROID__
  data = memalign(gAlignment, nbytes);
#elif defined(_MSC_VER)
  data = _aligned_malloc(nbytes, gAlignment);
#else
  int err = posix_memalign(&data, gAlignment, nbytes);
  if (err != 0) {

            

Reported by FlawFinder.