import unittest
import onnxruntime
import torch

import numpy as np
import io
import itertools
import copy
import os

            

Reported by Pylint.

              import unittest
import onnxruntime
import torch

import numpy as np
import io
import itertools
import copy
import os

            

Reported by Pylint.

              import os
import random

from torch.nn.utils import rnn as rnn_utils
from model_defs.lstm_flattening_result import (LstmFlatteningResultWithSeqLength,
                                               LstmFlatteningResultWithoutSeqLength)
from model_defs.rnn_model_with_packed_sequence import (RnnModelWithPackedSequence,
                                                       RnnModelWithPackedSequenceWithState,
                                                       RnnModelWithPackedSequenceWithoutState)

            

Reported by Pylint.

              from model_defs.rnn_model_with_packed_sequence import (RnnModelWithPackedSequence,
                                                       RnnModelWithPackedSequenceWithState,
                                                       RnnModelWithPackedSequenceWithoutState)
from test_pytorch_common import (skipIfUnsupportedMinOpsetVersion, skipIfUnsupportedOpsetVersion,
                                 skipIfNoLapack, disableScriptTest, skipIfONNXShapeInference,
                                 skipIfUnsupportedMaxOpsetVersion)
from test_pytorch_common import BATCH_SIZE
from test_pytorch_common import RNN_BATCH_SIZE, RNN_SEQUENCE_LENGTH, RNN_INPUT_SIZE, RNN_HIDDEN_SIZE
from typing import List, Tuple, Optional, Dict

            

Reported by Pylint.

              from test_pytorch_common import BATCH_SIZE
from test_pytorch_common import RNN_BATCH_SIZE, RNN_SEQUENCE_LENGTH, RNN_INPUT_SIZE, RNN_HIDDEN_SIZE
from typing import List, Tuple, Optional, Dict
from torch import Tensor
import model_defs.word_language_model as word_language_model

import onnx

import torchvision

            

Reported by Pylint.

              from torch import Tensor
import model_defs.word_language_model as word_language_model

import onnx

import torchvision
from torchvision import ops
from torchvision.models.detection.image_list import ImageList
from torchvision.models.detection.transform import GeneralizedRCNNTransform

            

Reported by Pylint.

              
import onnx

import torchvision
from torchvision import ops
from torchvision.models.detection.image_list import ImageList
from torchvision.models.detection.transform import GeneralizedRCNNTransform
from torchvision.models.detection.rpn import AnchorGenerator, RPNHead, RegionProposalNetwork
from torchvision.models.detection.roi_heads import RoIHeads

            

Reported by Pylint.

              import onnx

import torchvision
from torchvision import ops
from torchvision.models.detection.image_list import ImageList
from torchvision.models.detection.transform import GeneralizedRCNNTransform
from torchvision.models.detection.rpn import AnchorGenerator, RPNHead, RegionProposalNetwork
from torchvision.models.detection.roi_heads import RoIHeads
from torchvision.models.detection.faster_rcnn import FastRCNNPredictor, TwoMLPHead

            

Reported by Pylint.

              
import torchvision
from torchvision import ops
from torchvision.models.detection.image_list import ImageList
from torchvision.models.detection.transform import GeneralizedRCNNTransform
from torchvision.models.detection.rpn import AnchorGenerator, RPNHead, RegionProposalNetwork
from torchvision.models.detection.roi_heads import RoIHeads
from torchvision.models.detection.faster_rcnn import FastRCNNPredictor, TwoMLPHead
from collections import OrderedDict

            

Reported by Pylint.

              import torchvision
from torchvision import ops
from torchvision.models.detection.image_list import ImageList
from torchvision.models.detection.transform import GeneralizedRCNNTransform
from torchvision.models.detection.rpn import AnchorGenerator, RPNHead, RegionProposalNetwork
from torchvision.models.detection.roi_heads import RoIHeads
from torchvision.models.detection.faster_rcnn import FastRCNNPredictor, TwoMLPHead
from collections import OrderedDict


            

Reported by Pylint.

              from operator import mul
from collections import OrderedDict

import torch

# TODO: remove this global setting
# NN tests use double as the default dtype
torch.set_default_dtype(torch.double)


            

Reported by Pylint.

              # NN tests use double as the default dtype
torch.set_default_dtype(torch.double)

from torch._six import inf, nan
import torch.backends.cudnn as cudnn
import torch.nn as nn
import torch.nn.functional as F
import torch.nn.init as init
import torch.nn.utils.rnn as rnn_utils

            

Reported by Pylint.

              torch.set_default_dtype(torch.double)

from torch._six import inf, nan
import torch.backends.cudnn as cudnn
import torch.nn as nn
import torch.nn.functional as F
import torch.nn.init as init
import torch.nn.utils.rnn as rnn_utils
from torch.nn.utils import clip_grad_norm_, clip_grad_value_

            

Reported by Pylint.

              
from torch._six import inf, nan
import torch.backends.cudnn as cudnn
import torch.nn as nn
import torch.nn.functional as F
import torch.nn.init as init
import torch.nn.utils.rnn as rnn_utils
from torch.nn.utils import clip_grad_norm_, clip_grad_value_
import torch.nn.utils.parametrize as parametrize

            

Reported by Pylint.

              from torch._six import inf, nan
import torch.backends.cudnn as cudnn
import torch.nn as nn
import torch.nn.functional as F
import torch.nn.init as init
import torch.nn.utils.rnn as rnn_utils
from torch.nn.utils import clip_grad_norm_, clip_grad_value_
import torch.nn.utils.parametrize as parametrize
import torch.nn.utils.prune as prune

            

Reported by Pylint.

              import torch.backends.cudnn as cudnn
import torch.nn as nn
import torch.nn.functional as F
import torch.nn.init as init
import torch.nn.utils.rnn as rnn_utils
from torch.nn.utils import clip_grad_norm_, clip_grad_value_
import torch.nn.utils.parametrize as parametrize
import torch.nn.utils.prune as prune
from torch.nn.utils import parameters_to_vector, vector_to_parameters

            

Reported by Pylint.

              import torch.nn as nn
import torch.nn.functional as F
import torch.nn.init as init
import torch.nn.utils.rnn as rnn_utils
from torch.nn.utils import clip_grad_norm_, clip_grad_value_
import torch.nn.utils.parametrize as parametrize
import torch.nn.utils.prune as prune
from torch.nn.utils import parameters_to_vector, vector_to_parameters
from torch.nn import Parameter

            

Reported by Pylint.

              import torch.nn.functional as F
import torch.nn.init as init
import torch.nn.utils.rnn as rnn_utils
from torch.nn.utils import clip_grad_norm_, clip_grad_value_
import torch.nn.utils.parametrize as parametrize
import torch.nn.utils.prune as prune
from torch.nn.utils import parameters_to_vector, vector_to_parameters
from torch.nn import Parameter
from torch.nn.parameter import UninitializedParameter, UninitializedBuffer

            

Reported by Pylint.

              import torch.nn.init as init
import torch.nn.utils.rnn as rnn_utils
from torch.nn.utils import clip_grad_norm_, clip_grad_value_
import torch.nn.utils.parametrize as parametrize
import torch.nn.utils.prune as prune
from torch.nn.utils import parameters_to_vector, vector_to_parameters
from torch.nn import Parameter
from torch.nn.parameter import UninitializedParameter, UninitializedBuffer
from torch.nn.parallel._functions import Broadcast

            

Reported by Pylint.

              import torch.nn.utils.rnn as rnn_utils
from torch.nn.utils import clip_grad_norm_, clip_grad_value_
import torch.nn.utils.parametrize as parametrize
import torch.nn.utils.prune as prune
from torch.nn.utils import parameters_to_vector, vector_to_parameters
from torch.nn import Parameter
from torch.nn.parameter import UninitializedParameter, UninitializedBuffer
from torch.nn.parallel._functions import Broadcast
from torch.testing import get_all_fp_dtypes

            

Reported by Pylint.

              from itertools import product, permutations
from operator import mul
from functools import reduce, partial
import torch

from torch import nn
from torch._six import inf, nan
from torch.autograd.function import once_differentiable
from torch.autograd.profiler import (profile, record_function, emit_nvtx)

            

Reported by Pylint.

              from functools import reduce, partial
import torch

from torch import nn
from torch._six import inf, nan
from torch.autograd.function import once_differentiable
from torch.autograd.profiler import (profile, record_function, emit_nvtx)
from torch.autograd.profiler_util import (_format_time, EventList, FunctionEvent, FunctionEventAvg)
import torch.autograd.functional as autogradF

            

Reported by Pylint.

              import torch

from torch import nn
from torch._six import inf, nan
from torch.autograd.function import once_differentiable
from torch.autograd.profiler import (profile, record_function, emit_nvtx)
from torch.autograd.profiler_util import (_format_time, EventList, FunctionEvent, FunctionEventAvg)
import torch.autograd.functional as autogradF
from torch.utils.checkpoint import checkpoint

            

Reported by Pylint.

              
from torch import nn
from torch._six import inf, nan
from torch.autograd.function import once_differentiable
from torch.autograd.profiler import (profile, record_function, emit_nvtx)
from torch.autograd.profiler_util import (_format_time, EventList, FunctionEvent, FunctionEventAvg)
import torch.autograd.functional as autogradF
from torch.utils.checkpoint import checkpoint
from torch.testing._internal.common_cuda import TEST_CUDA

            

Reported by Pylint.

              from torch import nn
from torch._six import inf, nan
from torch.autograd.function import once_differentiable
from torch.autograd.profiler import (profile, record_function, emit_nvtx)
from torch.autograd.profiler_util import (_format_time, EventList, FunctionEvent, FunctionEventAvg)
import torch.autograd.functional as autogradF
from torch.utils.checkpoint import checkpoint
from torch.testing._internal.common_cuda import TEST_CUDA
from torch.testing._internal.common_utils import (TestCase, run_tests, skipIfNoLapack,

            

Reported by Pylint.

              from torch._six import inf, nan
from torch.autograd.function import once_differentiable
from torch.autograd.profiler import (profile, record_function, emit_nvtx)
from torch.autograd.profiler_util import (_format_time, EventList, FunctionEvent, FunctionEventAvg)
import torch.autograd.functional as autogradF
from torch.utils.checkpoint import checkpoint
from torch.testing._internal.common_cuda import TEST_CUDA
from torch.testing._internal.common_utils import (TestCase, run_tests, skipIfNoLapack,
                                                  suppress_warnings, slowTest,

            

Reported by Pylint.

              from torch.autograd.function import once_differentiable
from torch.autograd.profiler import (profile, record_function, emit_nvtx)
from torch.autograd.profiler_util import (_format_time, EventList, FunctionEvent, FunctionEventAvg)
import torch.autograd.functional as autogradF
from torch.utils.checkpoint import checkpoint
from torch.testing._internal.common_cuda import TEST_CUDA
from torch.testing._internal.common_utils import (TestCase, run_tests, skipIfNoLapack,
                                                  suppress_warnings, slowTest,
                                                  load_tests,

            

Reported by Pylint.

              from torch.autograd.profiler import (profile, record_function, emit_nvtx)
from torch.autograd.profiler_util import (_format_time, EventList, FunctionEvent, FunctionEventAvg)
import torch.autograd.functional as autogradF
from torch.utils.checkpoint import checkpoint
from torch.testing._internal.common_cuda import TEST_CUDA
from torch.testing._internal.common_utils import (TestCase, run_tests, skipIfNoLapack,
                                                  suppress_warnings, slowTest,
                                                  load_tests,
                                                  IS_WINDOWS, IS_MACOS, CudaMemoryLeakCheck,

            

Reported by Pylint.

              from torch.autograd.profiler_util import (_format_time, EventList, FunctionEvent, FunctionEventAvg)
import torch.autograd.functional as autogradF
from torch.utils.checkpoint import checkpoint
from torch.testing._internal.common_cuda import TEST_CUDA
from torch.testing._internal.common_utils import (TestCase, run_tests, skipIfNoLapack,
                                                  suppress_warnings, slowTest,
                                                  load_tests,
                                                  IS_WINDOWS, IS_MACOS, CudaMemoryLeakCheck,
                                                  TEST_WITH_ROCM, disable_gc,

            

Reported by Pylint.

              import torch.autograd.functional as autogradF
from torch.utils.checkpoint import checkpoint
from torch.testing._internal.common_cuda import TEST_CUDA
from torch.testing._internal.common_utils import (TestCase, run_tests, skipIfNoLapack,
                                                  suppress_warnings, slowTest,
                                                  load_tests,
                                                  IS_WINDOWS, IS_MACOS, CudaMemoryLeakCheck,
                                                  TEST_WITH_ROCM, disable_gc,
                                                  gradcheck, gradgradcheck, make_tensor)

            

Reported by Pylint.

                   floating_and_complex_types, floating_and_complex_types_and,
     all_types_and_complex_and, all_types_and, all_types_and_complex,
     integral_types_and, all_types, double_types)
from .._core import _dispatch_dtypes
from torch.testing._internal.common_device_type import \
    (onlyOnCPUAndCUDA, skipCUDAIfNoMagma, skipCUDAIfNoMagmaAndNoCusolver, skipCUDAIfNoCusolver,
     skipCPUIfNoLapack, skipCPUIfNoFFT, skipCUDAIfRocm, precisionOverride, toleranceOverride, tol)
from torch.testing._internal.common_cuda import CUDA11OrLater, SM53OrLater, SM60OrLater
from torch.testing._internal.common_utils import \

            

Reported by Pylint.

              from setuptools import distutils

if TEST_SCIPY:
    import scipy.special


class DecorateInfo(object):
    """Describes which test, or type of tests, should be wrapped in the given
       decorators when testing an operator. Any test that matches all provided

            

Reported by Pylint.

                      else:
            backward_dtypes = self.backward_dtypes

        allowed_backward_dtypes = floating_and_complex_types_and(torch.bfloat16, torch.float16)
        return set(allowed_backward_dtypes).intersection(backward_dtypes)

    def supports_complex_autograd(self, device_type):
        if device_type == 'cpu':
            return any(dtype.is_complex for dtype in self.backward_dtypesIfCPU)

            

Reported by Pylint.

                      else:
            backward_dtypes = self.backward_dtypes

        allowed_backward_dtypes = floating_and_complex_types_and(torch.bfloat16, torch.float16)
        return set(allowed_backward_dtypes).intersection(backward_dtypes)

    def supports_complex_autograd(self, device_type):
        if device_type == 'cpu':
            return any(dtype.is_complex for dtype in self.backward_dtypesIfCPU)

            

Reported by Pylint.

                               dtypesIfCUDA=None,
                 dtypesIfROCM=None,
                 default_test_dtypes=(
                     torch.uint8, torch.long, torch.half, torch.bfloat16,
                     torch.float32, torch.cfloat),  # dtypes which tests check by default
                 domain=(None, None),  # the [low, high) domain of the function
                 handles_large_floats=True,  # whether the op correctly handles large float values (like 1e20)
                 handles_extremals=True,  # whether the op correctly handles extremal values (like inf)
                 handles_complex_extremals=True,  # whether the op correct handles complex extremals (like inf -infj)

            

Reported by Pylint.

                               dtypesIfCUDA=None,
                 dtypesIfROCM=None,
                 default_test_dtypes=(
                     torch.uint8, torch.long, torch.half, torch.bfloat16,
                     torch.float32, torch.cfloat),  # dtypes which tests check by default
                 domain=(None, None),  # the [low, high) domain of the function
                 handles_large_floats=True,  # whether the op correctly handles large float values (like 1e20)
                 handles_extremals=True,  # whether the op correctly handles extremal values (like inf)
                 handles_complex_extremals=True,  # whether the op correct handles complex extremals (like inf -infj)

            

Reported by Pylint.

                               dtypesIfCUDA=None,
                 dtypesIfROCM=None,
                 default_test_dtypes=(
                     torch.uint8, torch.long, torch.half, torch.bfloat16,
                     torch.float32, torch.cfloat),  # dtypes which tests check by default
                 domain=(None, None),  # the [low, high) domain of the function
                 handles_large_floats=True,  # whether the op correctly handles large float values (like 1e20)
                 handles_extremals=True,  # whether the op correctly handles extremal values (like inf)
                 handles_complex_extremals=True,  # whether the op correct handles complex extremals (like inf -infj)

            

Reported by Pylint.

                               dtypesIfCUDA=None,
                 dtypesIfROCM=None,
                 default_test_dtypes=(
                     torch.uint8, torch.long, torch.half, torch.bfloat16,
                     torch.float32, torch.cfloat),  # dtypes which tests check by default
                 domain=(None, None),  # the [low, high) domain of the function
                 handles_large_floats=True,  # whether the op correctly handles large float values (like 1e20)
                 handles_extremals=True,  # whether the op correctly handles extremal values (like inf)
                 handles_complex_extremals=True,  # whether the op correct handles complex extremals (like inf -infj)

            

Reported by Pylint.

                               dtypesIfROCM=None,
                 default_test_dtypes=(
                     torch.uint8, torch.long, torch.half, torch.bfloat16,
                     torch.float32, torch.cfloat),  # dtypes which tests check by default
                 domain=(None, None),  # the [low, high) domain of the function
                 handles_large_floats=True,  # whether the op correctly handles large float values (like 1e20)
                 handles_extremals=True,  # whether the op correctly handles extremal values (like inf)
                 handles_complex_extremals=True,  # whether the op correct handles complex extremals (like inf -infj)
                 supports_complex_to_float=False,  # op supports casting from complex input to real output safely eg. angle

            

Reported by Pylint.

                               dtypesIfROCM=None,
                 default_test_dtypes=(
                     torch.uint8, torch.long, torch.half, torch.bfloat16,
                     torch.float32, torch.cfloat),  # dtypes which tests check by default
                 domain=(None, None),  # the [low, high) domain of the function
                 handles_large_floats=True,  # whether the op correctly handles large float values (like 1e20)
                 handles_extremals=True,  # whether the op correctly handles extremal values (like inf)
                 handles_complex_extremals=True,  # whether the op correct handles complex extremals (like inf -infj)
                 supports_complex_to_float=False,  # op supports casting from complex input to real output safely eg. angle

            

Reported by Pylint.

              

def foo_add():
    return torch.add(torch.ones(1), torch.ones(1))

def udf_with_torch_ops(device=-1, use_record_function=False):
    device_ctx = contextlib.suppress() if device == -1 else torch.cuda.device(device)
    record_function_ctx = (
        torch.autograd.profiler.record_function("##forward##")

            

Reported by Pylint.

              

def foo_add():
    return torch.add(torch.ones(1), torch.ones(1))

def udf_with_torch_ops(device=-1, use_record_function=False):
    device_ctx = contextlib.suppress() if device == -1 else torch.cuda.device(device)
    record_function_ctx = (
        torch.autograd.profiler.record_function("##forward##")

            

Reported by Pylint.

              

def foo_add():
    return torch.add(torch.ones(1), torch.ones(1))

def udf_with_torch_ops(device=-1, use_record_function=False):
    device_ctx = contextlib.suppress() if device == -1 else torch.cuda.device(device)
    record_function_ctx = (
        torch.autograd.profiler.record_function("##forward##")

            

Reported by Pylint.

                      else contextlib.suppress()
    )
    with device_ctx, record_function_ctx:
        t1, t2 = torch.ones(1), torch.ones(1)
        t = torch.add(t1, t2)
        t = torch.mul(t, t)
        t = t.relu()
        t = t.sigmoid()


            

Reported by Pylint.

                      else contextlib.suppress()
    )
    with device_ctx, record_function_ctx:
        t1, t2 = torch.ones(1), torch.ones(1)
        t = torch.add(t1, t2)
        t = torch.mul(t, t)
        t = t.relu()
        t = t.sigmoid()


            

Reported by Pylint.

                  )
    with device_ctx, record_function_ctx:
        t1, t2 = torch.ones(1), torch.ones(1)
        t = torch.add(t1, t2)
        t = torch.mul(t, t)
        t = t.relu()
        t = t.sigmoid()

# Events (operator invocations) that are expected to be ran as part of the above

            

Reported by Pylint.

                  with device_ctx, record_function_ctx:
        t1, t2 = torch.ones(1), torch.ones(1)
        t = torch.add(t1, t2)
        t = torch.mul(t, t)
        t = t.relu()
        t = t.sigmoid()

# Events (operator invocations) that are expected to be ran as part of the above
# function.

            

Reported by Pylint.

              
    def __getstate__(self):
        (pickled_python_udf, tensors) = _internal_rpc_pickler.serialize(
            PythonUDF(my_tensor_function, (torch.ones(2, 2), torch.ones(2, 2)), None)
        )
        return (pickled_python_udf, tensors)

    def __setstate__(self, obj):
        python_udf = _internal_rpc_pickler.deserialize(obj[0], obj[1])

            

Reported by Pylint.

              
    def __getstate__(self):
        (pickled_python_udf, tensors) = _internal_rpc_pickler.serialize(
            PythonUDF(my_tensor_function, (torch.ones(2, 2), torch.ones(2, 2)), None)
        )
        return (pickled_python_udf, tensors)

    def __setstate__(self, obj):
        python_udf = _internal_rpc_pickler.deserialize(obj[0], obj[1])

            

Reported by Pylint.

              
    def my_slow_method(self, my_tensor_arg):
        time.sleep(5)
        return torch.add(self.a, my_tensor_arg)


def _call_method_on_rref(method, rref, *args, **kwargs):
    return method(rref.local_value(), *args, **kwargs)


            

Reported by Pylint.

              # -*- coding: utf-8 -*-
import torch
import numpy as np

import unittest
import itertools
import warnings
import math
from math import inf, nan, isnan

            

Reported by Pylint.

              from itertools import product
from functools import reduce

from torch.testing._internal.common_utils import \
    (TestCase, run_tests, TEST_SCIPY, IS_MACOS, IS_WINDOWS, slowTest,
     TEST_WITH_ASAN, make_tensor, TEST_WITH_ROCM, IS_FBCODE, IS_REMOTE_GPU,
     iter_indices, gradcheck, gradgradcheck)
from torch.testing._internal.common_device_type import \
    (instantiate_device_type_tests, dtypes,

            

Reported by Pylint.

                  (TestCase, run_tests, TEST_SCIPY, IS_MACOS, IS_WINDOWS, slowTest,
     TEST_WITH_ASAN, make_tensor, TEST_WITH_ROCM, IS_FBCODE, IS_REMOTE_GPU,
     iter_indices, gradcheck, gradgradcheck)
from torch.testing._internal.common_device_type import \
    (instantiate_device_type_tests, dtypes,
     onlyCPU, skipCUDAIf, skipCUDAIfNoMagma, skipCPUIfNoLapack, precisionOverride,
     skipCUDAIfNoMagmaAndNoCusolver, skipCUDAIfRocm, onlyOnCPUAndCUDA, dtypesIfCUDA,
     onlyCUDA, skipCUDAVersionIn, skipMeta, skipCUDAIfNoCusolver)
from torch.testing import floating_and_complex_types, floating_types, all_types

            

Reported by Pylint.

                   onlyCPU, skipCUDAIf, skipCUDAIfNoMagma, skipCPUIfNoLapack, precisionOverride,
     skipCUDAIfNoMagmaAndNoCusolver, skipCUDAIfRocm, onlyOnCPUAndCUDA, dtypesIfCUDA,
     onlyCUDA, skipCUDAVersionIn, skipMeta, skipCUDAIfNoCusolver)
from torch.testing import floating_and_complex_types, floating_types, all_types
from torch.testing._internal.common_cuda import SM53OrLater, tf32_on_and_off, CUDA11OrLater, CUDA9
from torch.distributions.binomial import Binomial

# Protects against includes accidentally setting the default dtype
# NOTE: jit_metaprogramming_utils sets the default dtype to double!

            

Reported by Pylint.

                   skipCUDAIfNoMagmaAndNoCusolver, skipCUDAIfRocm, onlyOnCPUAndCUDA, dtypesIfCUDA,
     onlyCUDA, skipCUDAVersionIn, skipMeta, skipCUDAIfNoCusolver)
from torch.testing import floating_and_complex_types, floating_types, all_types
from torch.testing._internal.common_cuda import SM53OrLater, tf32_on_and_off, CUDA11OrLater, CUDA9
from torch.distributions.binomial import Binomial

# Protects against includes accidentally setting the default dtype
# NOTE: jit_metaprogramming_utils sets the default dtype to double!
torch.set_default_dtype(torch.float32)

            

Reported by Pylint.

                   onlyCUDA, skipCUDAVersionIn, skipMeta, skipCUDAIfNoCusolver)
from torch.testing import floating_and_complex_types, floating_types, all_types
from torch.testing._internal.common_cuda import SM53OrLater, tf32_on_and_off, CUDA11OrLater, CUDA9
from torch.distributions.binomial import Binomial

# Protects against includes accidentally setting the default dtype
# NOTE: jit_metaprogramming_utils sets the default dtype to double!
torch.set_default_dtype(torch.float32)
assert torch.get_default_dtype() is torch.float32

            

Reported by Pylint.

              assert torch.get_default_dtype() is torch.float32

if TEST_SCIPY:
    import scipy

class TestLinalg(TestCase):
    def setUp(self):
        super(self.__class__, self).setUp()
        torch.backends.cuda.matmul.allow_tf32 = False

            

Reported by Pylint.

              
class TestLinalg(TestCase):
    def setUp(self):
        super(self.__class__, self).setUp()
        torch.backends.cuda.matmul.allow_tf32 = False

    def tearDown(self):
        torch.backends.cuda.matmul.allow_tf32 = True
        super(self.__class__, self).tearDown()

            

Reported by Pylint.

              
    def tearDown(self):
        torch.backends.cuda.matmul.allow_tf32 = True
        super(self.__class__, self).tearDown()

    exact_dtype = True

    @dtypes(torch.float, torch.cfloat)
    @precisionOverride({torch.float: 1e-06, torch.cfloat: 1e-06})

            

Reported by Pylint.

                  @skipCPUIfNoLapack
    @dtypes(torch.float, torch.double, torch.cfloat, torch.cdouble)
    def test_linalg_lstsq(self, device, dtype):
        from torch.testing._internal.common_utils import random_well_conditioned_matrix
        if self.device_type == 'cpu':
            drivers = ('gels', 'gelsy', 'gelsd', 'gelss', None)
        else:
            drivers = ('gels', None)


            

Reported by Pylint.

              # -*- coding: utf-8 -*-
import torch
import torch.utils.data
import numpy as np

import contextlib
import gc
import io
import inspect

            

Reported by Pylint.

              # -*- coding: utf-8 -*-
import torch
import torch.utils.data
import numpy as np

import contextlib
import gc
import io
import inspect

            

Reported by Pylint.

              import subprocess
import weakref
import sys
from torch.utils.dlpack import from_dlpack, to_dlpack
from torch._six import inf, nan, string_classes
from itertools import product, combinations, permutations
from functools import partial
from torch import multiprocessing as mp
from torch.testing._internal.common_utils import (

            

Reported by Pylint.

              import weakref
import sys
from torch.utils.dlpack import from_dlpack, to_dlpack
from torch._six import inf, nan, string_classes
from itertools import product, combinations, permutations
from functools import partial
from torch import multiprocessing as mp
from torch.testing._internal.common_utils import (
    TestCase, TEST_WITH_ROCM, run_tests,

            

Reported by Pylint.

              from torch._six import inf, nan, string_classes
from itertools import product, combinations, permutations
from functools import partial
from torch import multiprocessing as mp
from torch.testing._internal.common_utils import (
    TestCase, TEST_WITH_ROCM, run_tests,
    IS_WINDOWS, IS_FILESYSTEM_UTF8_ENCODING, NO_MULTIPROCESSING_SPAWN,
    do_test_dtypes, IS_SANDCASTLE, IS_FBCODE, IS_REMOTE_GPU, load_tests, slowTest,
    skipCUDAMemoryLeakCheckIf, BytesIOContext, noarchTest,

            

Reported by Pylint.

              from itertools import product, combinations, permutations
from functools import partial
from torch import multiprocessing as mp
from torch.testing._internal.common_utils import (
    TestCase, TEST_WITH_ROCM, run_tests,
    IS_WINDOWS, IS_FILESYSTEM_UTF8_ENCODING, NO_MULTIPROCESSING_SPAWN,
    do_test_dtypes, IS_SANDCASTLE, IS_FBCODE, IS_REMOTE_GPU, load_tests, slowTest,
    skipCUDAMemoryLeakCheckIf, BytesIOContext, noarchTest,
    skipIfRocm, skipIfNoSciPy, TemporaryFileName, TemporaryDirectoryName,

            

Reported by Pylint.

                  skipIfRocm, skipIfNoSciPy, TemporaryFileName, TemporaryDirectoryName,
    wrapDeterministicFlagAPITest, DeterministicGuard, CudaSyncGuard, make_tensor)
from multiprocessing.reduction import ForkingPickler
from torch.testing._internal.common_device_type import (
    instantiate_device_type_tests,
    skipCUDAVersionIn,
    onlyCUDA, onlyCPU,
    dtypes, dtypesIfCUDA, dtypesIfCPU, deviceCountAtLeast,
    skipMeta,

            

Reported by Pylint.

                  PYTORCH_CUDA_MEMCHECK, largeTensorTest, onlyOnCPUAndCUDA,
    expectedAlertNondeterministic)
from typing import Dict, List, Tuple
import torch.backends.quantized
import torch.testing._internal.data
from torch.testing._internal.common_cuda import tf32_on_and_off, tf32_is_not_fp32

# Protects against includes accidentally setting the default dtype
assert torch.get_default_dtype() is torch.float32

            

Reported by Pylint.

                  expectedAlertNondeterministic)
from typing import Dict, List, Tuple
import torch.backends.quantized
import torch.testing._internal.data
from torch.testing._internal.common_cuda import tf32_on_and_off, tf32_is_not_fp32

# Protects against includes accidentally setting the default dtype
assert torch.get_default_dtype() is torch.float32


            

Reported by Pylint.

              from typing import Dict, List, Tuple
import torch.backends.quantized
import torch.testing._internal.data
from torch.testing._internal.common_cuda import tf32_on_and_off, tf32_is_not_fp32

# Protects against includes accidentally setting the default dtype
assert torch.get_default_dtype() is torch.float32

# load_tests from torch.testing._internal.common_utils is used to automatically filter tests for

            

Reported by Pylint.

                  HAS_TORCHVISION = False

if sys.platform == "win32":
    import msvcrt
else:
    import fcntl


class Foo:

            

Reported by Pylint.

                  def __eq__(self, other):
        def eq(value, other):
            if isinstance(value, torch.Tensor):
                return torch.equal(value, other)
            return value == other

        for attr, value in self.__dict__.items():
            other_value = other.__dict__[attr]
            if not eq(value, other_value):

            

Reported by Pylint.

              f = Foo(10)
f.bar = 1

foo_cpu_tensor = Foo(torch.randn(3, 3))


COLLECTIVES_OBJECT_TEST_LIST = [
    {"key1": 3, "key2": 4, "key3": {"nested": True}},
    f,

            

Reported by Pylint.

              

class TestNamedTupleInput_1(NamedTuple):
    a: torch.tensor
    b: torch.tensor


skipIfNoTorchVision = sandcastle_skip_if(not HAS_TORCHVISION, "no torchvision")


            

Reported by Pylint.

              
class TestNamedTupleInput_1(NamedTuple):
    a: torch.tensor
    b: torch.tensor


skipIfNoTorchVision = sandcastle_skip_if(not HAS_TORCHVISION, "no torchvision")

BACKEND = os.environ["BACKEND"]

            

Reported by Pylint.

              class DDPUnevenTestInput(NamedTuple):
    name: str
    model: nn.Module
    inp: Union[torch.tensor, tuple]
    sync_interval: int
    throw_on_early_termination: bool = False
    hook: Callable = None
    state: Any = None


            

Reported by Pylint.

                      self.fc3 = nn.Linear(50, 4, bias=False)
        self.relu = nn.ReLU()
        self.no_grad_param = nn.Parameter(
            torch.tensor([2, 2]).long(), requires_grad=False
        )

    def forward(self, x):
        x = self.relu(self.fc1(x))
        x = self.relu(self.fc2(x))

            

Reported by Pylint.

              class Task(nn.Module):
    def __init__(self):
        super().__init__()
        self.p = nn.Parameter(torch.ones(2, 2))

    def forward(self, x):
        return self.p + x



            

Reported by Pylint.

                      self.fc2 = nn.Linear(40, 4, bias=False)

    def forward(self, x):
        x = torch.reshape(self.fc1(x), (-1, 4, 10))
        x = self.bn(x)
        x = torch.reshape(x, (-1, 40))
        x = self.fc2(x)
        return F.softmax(x, dim=1)


            

Reported by Pylint.

                  def forward(self, x):
        x = torch.reshape(self.fc1(x), (-1, 4, 10))
        x = self.bn(x)
        x = torch.reshape(x, (-1, 40))
        x = self.fc2(x)
        return F.softmax(x, dim=1)


class UnusedParamTwoLinLayerNet(nn.Module):

            

Reported by Pylint.

                      # Integer division does trunction rounding
        div = g.op("Div", self, other)
        # Division is negative if: self < 0 != other < 0
        zero = g.op("Constant", value_t=torch.tensor(0, dtype=torch.int64))
        negative = g.op("Xor",
                        g.op("Less", self, zero),
                        g.op("Less", other, zero))

        # For negative numbers with self % other != 0, subtract 1 to round down instead of up

            

Reported by Pylint.

                      # Integer division does trunction rounding
        div = g.op("Div", self, other)
        # Division is negative if: self < 0 != other < 0
        zero = g.op("Constant", value_t=torch.tensor(0, dtype=torch.int64))
        negative = g.op("Xor",
                        g.op("Less", self, zero),
                        g.op("Less", other, zero))

        # For negative numbers with self % other != 0, subtract 1 to round down instead of up

            

Reported by Pylint.

                      fixup_mask = g.op("And", negative,
                          g.op("Not", g.op("Equal", mod, zero)))

        one = g.op("Constant", value_t=torch.tensor(1, dtype=torch.int64))
        fixup = g.op("Sub", div, one)
        return g.op("Where", fixup_mask, fixup, div)


def floor_divide(g, self, other):

            

Reported by Pylint.

                      fixup_mask = g.op("And", negative,
                          g.op("Not", g.op("Equal", mod, zero)))

        one = g.op("Constant", value_t=torch.tensor(1, dtype=torch.int64))
        fixup = g.op("Sub", div, one)
        return g.op("Where", fixup_mask, fixup, div)


def floor_divide(g, self, other):

            

Reported by Pylint.

              
    # Case 4: neither is floating
    # Casts both inputs to the default scalar type
    scalar_type = torch.get_default_dtype()
    onnx_scalar_type = sym_help.cast_pytorch_to_onnx["Float"]
    assert scalar_type is torch.float or scalar_type is torch.double
    if torch.get_default_dtype() is torch.double:
        onnx_scalar_type = sym_help.cast_pytorch_to_onnx["Double"]


            

Reported by Pylint.

                  # Casts both inputs to the default scalar type
    scalar_type = torch.get_default_dtype()
    onnx_scalar_type = sym_help.cast_pytorch_to_onnx["Float"]
    assert scalar_type is torch.float or scalar_type is torch.double
    if torch.get_default_dtype() is torch.double:
        onnx_scalar_type = sym_help.cast_pytorch_to_onnx["Double"]

    self = g.op("Cast", self, to_i=onnx_scalar_type)
    other = g.op("Cast", other, to_i=onnx_scalar_type)

            

Reported by Pylint.

                  # Casts both inputs to the default scalar type
    scalar_type = torch.get_default_dtype()
    onnx_scalar_type = sym_help.cast_pytorch_to_onnx["Float"]
    assert scalar_type is torch.float or scalar_type is torch.double
    if torch.get_default_dtype() is torch.double:
        onnx_scalar_type = sym_help.cast_pytorch_to_onnx["Double"]

    self = g.op("Cast", self, to_i=onnx_scalar_type)
    other = g.op("Cast", other, to_i=onnx_scalar_type)

            

Reported by Pylint.

                  scalar_type = torch.get_default_dtype()
    onnx_scalar_type = sym_help.cast_pytorch_to_onnx["Float"]
    assert scalar_type is torch.float or scalar_type is torch.double
    if torch.get_default_dtype() is torch.double:
        onnx_scalar_type = sym_help.cast_pytorch_to_onnx["Double"]

    self = g.op("Cast", self, to_i=onnx_scalar_type)
    other = g.op("Cast", other, to_i=onnx_scalar_type)
    return g.op("Div", self, other)

            

Reported by Pylint.

                  scalar_type = torch.get_default_dtype()
    onnx_scalar_type = sym_help.cast_pytorch_to_onnx["Float"]
    assert scalar_type is torch.float or scalar_type is torch.double
    if torch.get_default_dtype() is torch.double:
        onnx_scalar_type = sym_help.cast_pytorch_to_onnx["Double"]

    self = g.op("Cast", self, to_i=onnx_scalar_type)
    other = g.op("Cast", other, to_i=onnx_scalar_type)
    return g.op("Div", self, other)

            

Reported by Pylint.

              

def reciprocal(g, self):
    return g.op("Div", torch.ones(1), self)


@parse_args("v", "i")
def cat(g, tensor_list, dim):
    tensors = sym_help._unpack_list(tensor_list)

            

Reported by Pylint.

              import operator
import random

import torch
from torch import _VF
import torch.jit
import torch.nn.functional as F
from torch.nn.modules.utils import _single, _pair


            

Reported by Pylint.

              import random

import torch
from torch import _VF
import torch.jit
import torch.nn.functional as F
from torch.nn.modules.utils import _single, _pair

from hypothesis import settings, HealthCheck

            

Reported by Pylint.

              
import torch
from torch import _VF
import torch.jit
import torch.nn.functional as F
from torch.nn.modules.utils import _single, _pair

from hypothesis import settings, HealthCheck
from hypothesis import assume, given, note

            

Reported by Pylint.

              import torch
from torch import _VF
import torch.jit
import torch.nn.functional as F
from torch.nn.modules.utils import _single, _pair

from hypothesis import settings, HealthCheck
from hypothesis import assume, given, note
from hypothesis import strategies as st

            

Reported by Pylint.

              from torch import _VF
import torch.jit
import torch.nn.functional as F
from torch.nn.modules.utils import _single, _pair

from hypothesis import settings, HealthCheck
from hypothesis import assume, given, note
from hypothesis import strategies as st
import torch.testing._internal.hypothesis_utils as hu

            

Reported by Pylint.

              import torch.nn.functional as F
from torch.nn.modules.utils import _single, _pair

from hypothesis import settings, HealthCheck
from hypothesis import assume, given, note
from hypothesis import strategies as st
import torch.testing._internal.hypothesis_utils as hu
hu.assert_deadline_disabled()


            

Reported by Pylint.

              from torch.nn.modules.utils import _single, _pair

from hypothesis import settings, HealthCheck
from hypothesis import assume, given, note
from hypothesis import strategies as st
import torch.testing._internal.hypothesis_utils as hu
hu.assert_deadline_disabled()

from torch.testing._internal.common_utils import TestCase

            

Reported by Pylint.

              
from hypothesis import settings, HealthCheck
from hypothesis import assume, given, note
from hypothesis import strategies as st
import torch.testing._internal.hypothesis_utils as hu
hu.assert_deadline_disabled()

from torch.testing._internal.common_utils import TestCase
from torch.testing._internal.common_utils import IS_PPC, TEST_WITH_UBSAN, IS_MACOS

            

Reported by Pylint.

              from hypothesis import settings, HealthCheck
from hypothesis import assume, given, note
from hypothesis import strategies as st
import torch.testing._internal.hypothesis_utils as hu
hu.assert_deadline_disabled()

from torch.testing._internal.common_utils import TestCase
from torch.testing._internal.common_utils import IS_PPC, TEST_WITH_UBSAN, IS_MACOS
from torch.testing._internal.common_quantization import skipIfNoFBGEMM, skipIfNoQNNPACK

            

Reported by Pylint.

              import torch.testing._internal.hypothesis_utils as hu
hu.assert_deadline_disabled()

from torch.testing._internal.common_utils import TestCase
from torch.testing._internal.common_utils import IS_PPC, TEST_WITH_UBSAN, IS_MACOS
from torch.testing._internal.common_quantization import skipIfNoFBGEMM, skipIfNoQNNPACK
from torch.testing._internal.common_quantized import _quantize, _dequantize, _calculate_dynamic_qparams, \
    override_quantized_engine, supported_qengines, override_qengines, _snr
from torch.testing._internal.common_quantized import qengine_is_qnnpack

            

Reported by Pylint.