import threading
import time
import torch
import unittest
from torch.futures import Future
from torch.testing._internal.common_utils import IS_WINDOWS, TestCase, TemporaryFileName, run_tests
from typing import TypeVar

T = TypeVar("T")

            

Reported by Pylint.

              import time
import torch
import unittest
from torch.futures import Future
from torch.testing._internal.common_utils import IS_WINDOWS, TestCase, TemporaryFileName, run_tests
from typing import TypeVar

T = TypeVar("T")


            

Reported by Pylint.

              import torch
import unittest
from torch.futures import Future
from torch.testing._internal.common_utils import IS_WINDOWS, TestCase, TemporaryFileName, run_tests
from typing import TypeVar

T = TypeVar("T")



            

Reported by Pylint.

              import threading
import time
import torch
import unittest
from torch.futures import Future
from torch.testing._internal.common_utils import IS_WINDOWS, TestCase, TemporaryFileName, run_tests
from typing import TypeVar

T = TypeVar("T")

            

Reported by Pylint.

              import threading
import time
import torch
import unittest
from torch.futures import Future
from torch.testing._internal.common_utils import IS_WINDOWS, TestCase, TemporaryFileName, run_tests
from typing import TypeVar

T = TypeVar("T")

            

Reported by Pylint.

              import unittest
from torch.futures import Future
from torch.testing._internal.common_utils import IS_WINDOWS, TestCase, TemporaryFileName, run_tests
from typing import TypeVar

T = TypeVar("T")


def add_one(fut):

            

Reported by Pylint.

              from torch.testing._internal.common_utils import IS_WINDOWS, TestCase, TemporaryFileName, run_tests
from typing import TypeVar

T = TypeVar("T")


def add_one(fut):
    return fut.wait() + 1


            

Reported by Pylint.

              T = TypeVar("T")


def add_one(fut):
    return fut.wait() + 1


class TestFuture(TestCase):
    def test_set_exception(self) -> None:

            

Reported by Pylint.

                  return fut.wait() + 1


class TestFuture(TestCase):
    def test_set_exception(self) -> None:
        # This test is to ensure errors can propagate across futures.
        error_msg = "Intentional Value Error"
        value_error = ValueError(error_msg)


            

Reported by Pylint.

                  return fut.wait() + 1


class TestFuture(TestCase):
    def test_set_exception(self) -> None:
        # This test is to ensure errors can propagate across futures.
        error_msg = "Intentional Value Error"
        value_error = ValueError(error_msg)


            

Reported by Pylint.

              
import logging

import hypothesis
from hypothesis import given, settings, HealthCheck
import hypothesis.strategies as st
import numpy as np

from caffe2.python import core

            

Reported by Pylint.

              import logging

import hypothesis
from hypothesis import given, settings, HealthCheck
import hypothesis.strategies as st
import numpy as np

from caffe2.python import core
import caffe2.python.hypothesis_test_util as hu

            

Reported by Pylint.

              
import hypothesis
from hypothesis import given, settings, HealthCheck
import hypothesis.strategies as st
import numpy as np

from caffe2.python import core
import caffe2.python.hypothesis_test_util as hu
import caffe2.python.serialized_test.serialized_test_util as serial

            

Reported by Pylint.

              

def wngrad_sparse_test_helper(parent_test, inputs, seq_b, lr, epsilon,
     engine, gc, dc):
    # helper functions for wngrad operator test
    param, grad = inputs
    seq_b = np.array([seq_b, ], dtype=np.float32)
    lr = np.array([lr], dtype=np.float32)


            

Reported by Pylint.

                                           allow_nan=False, allow_infinity=False),
           **hu.gcs_cpu_only)
    @settings(deadline=10000)
    def test_wngrad_dense_base(self, inputs, seq_b, lr, epsilon, gc, dc):
        param, grad = inputs
        seq_b = np.array([seq_b, ], dtype=np.float32)
        lr = np.array([lr], dtype=np.float32)

        op = core.CreateOperator(

            

Reported by Pylint.

                         **hu.gcs_cpu_only)
    @settings(deadline=10000)
    def test_wngrad_dense_output_effective_lr(self, inputs, seq_b,
                                              lr, epsilon, gc, dc):
        param, grad = inputs
        seq_b = np.array([seq_b, ], dtype=np.float32)
        lr = np.array([lr], dtype=np.float32)

        op = core.CreateOperator(

            

Reported by Pylint.

                         **hu.gcs_cpu_only)
    @settings(deadline=10000)
    def test_wngrad_dense_output_effective_lr_and_update(
            self, inputs, seq_b, lr, epsilon, gc, dc):
        param, grad = inputs
        seq_b = np.abs(np.array([seq_b, ], dtype=np.float32))
        lr = np.array([lr], dtype=np.float32)

        op = core.CreateOperator(

            

Reported by Pylint.

                                           allow_nan=False, allow_infinity=False),
           **hu.gcs_cpu_only)
    @settings(deadline=10000)
    def test_sparse_wngrad_empty(self, inputs, seq_b, lr, epsilon, gc, dc):
        param = inputs[0]
        seq_b = np.array([seq_b, ], dtype=np.float32)
        lr = np.array([lr], dtype=np.float32)

        grad = np.empty(shape=(0,) + param.shape[1:], dtype=np.float32)

            

Reported by Pylint.

                          epsilon=epsilon,
            device_option=gc)

        def ref_sparse(param, seq_b, indices, grad, lr):
            param_out = np.copy(param)
            seq_b_out = np.copy(seq_b)
            return (param_out, seq_b_out)

        print('test_sparse_adagrad_empty with full precision embedding')

            

Reported by Pylint.

                          epsilon=epsilon,
            device_option=gc)

        def ref_sparse(param, seq_b, indices, grad, lr):
            param_out = np.copy(param)
            seq_b_out = np.copy(seq_b)
            return (param_out, seq_b_out)

        print('test_sparse_adagrad_empty with full precision embedding')

            

Reported by Pylint.

                  pyver: Optional[str] = None
    cuda_version: Optional[str] = None
    rocm_version: Optional[str] = None
    # TODO expand this to cover all the USE_* that we want to test for
    #  tesnrorrt, leveldb, lmdb, redis, opencv, mkldnn, ideep, etc.
    # (from https://github.com/pytorch/pytorch/pull/17323#discussion_r259453608)
    is_xla: bool = False
    is_vulkan: bool = False
    is_pure_torch: bool = False

            

Reported by Pylint.

                  def is_test_phase(phase):
        return "test" in phase

    # TODO: Eliminate the special casing for docker paths
    # In the short term, we *will* need to support special casing as docker images are merged for caffe2 and pytorch
    def get_parms(self, for_docker):
        leading = []
        # We just don't run non-important jobs on pull requests;
        # previously we also named them in a way to make it obvious

            

Reported by Pylint.

              
        if Conf.is_test_phase(phase):

            # TODO When merging the caffe2 and pytorch jobs, it might be convenient for a while to make a
            #  caffe2 test job dependent on a pytorch build job. This way we could quickly dedup the repeated
            #  build of pytorch in the caffe2 build job, and just run the caffe2 tests off of a completed
            #  pytorch build job (from https://github.com/pytorch/pytorch/pull/17323#discussion_r259452641)

            dependency_build = self.parent_build or self

            

Reported by Pylint.

                      return {job_name: job_def}


# TODO This is a hack to special case some configs just for the workflow list
class HiddenConf(object):
    def __init__(self, name, parent_build=None, filters=None):
        self.name = name
        self.parent_build = parent_build
        self.filters = filters

            

Reported by Pylint.

                      self.parent_build = parent_build
        self.branch = branch

    def gen_workflow_job(self, phase):
        return {
            "pytorch_doc_push": {
                "name": self.name,
                "branch": self.branch,
                "requires": [self.parent_build],

            

Reported by Pylint.

                          }
        }

# TODO Convert these to graph nodes
def gen_dependent_configs(xenial_parent_config):

    extra_parms = [
        (["multigpu"], "large"),
        (["nogpu", "NO_AVX2"], None),

            

Reported by Pylint.

              
        elif compiler_name == "android":
            android_ndk_version = fc.find_prop("compiler_version")
            # TODO: do we need clang to compile host binaries like protoc?
            parms_list.append("clang5")
            parms_list.append("android-ndk-" + android_ndk_version)
            android_abi = fc.find_prop("android_abi")
            parms_list_ignored_for_docker_image.append(android_abi)
            restrict_phases = ["build"]

            

Reported by Pylint.

                      parallel_backend = fc.find_prop("parallel_backend") or None
        build_only = fc.find_prop("build_only") or False
        shard_test = fc.find_prop("shard_test") or False
        # TODO: fix pure_torch python test packaging issue.
        if shard_test:
            restrict_phases = ["build"] if restrict_phases is None else restrict_phases
            restrict_phases.extend(["test1", "test2"])
        if build_only or is_pure_torch:
            restrict_phases = ["build"]

            

Reported by Pylint.

              
        # run docs builds on "pytorch-linux-xenial-py3.6-gcc5.4". Docs builds
        # should run on a CPU-only build that runs on all PRs.
        # XXX should this be updated to a more modern build? Projects are
        #     beginning to drop python3.6
        if (
            distro_name == "xenial"
            and fc.find_prop("pyver") == "3.6"
            and cuda_version is None

            

Reported by Pylint.

                          and compiler_name == "gcc"
            and fc.find_prop("compiler_version") == "5.4"
        ):
            c.filters = gen_filter_dict(branches_list=r"/.*/",
                                        tags_list=RC_PATTERN)
            c.dependent_tests = gen_docs_configs(c)

        if cuda_version == "10.2" and python_version == "3.6" and not is_libtorch and not is_slow_gradcheck:
            c.dependent_tests = gen_dependent_configs(c)

            

Reported by Pylint.

              import io

import torch
from ._utils import _type, _cuda
from typing import Any, TypeVar, Type

T = TypeVar('T', bound='_StorageBase')
class _StorageBase(object):
    _cdata: Any

            

Reported by Pylint.

                  is_sparse: bool = False

    def __init__(self, *args, **kwargs): ...  # noqa: E704
    def __len__(self) -> int: ...  # noqa: E704
    def __getitem__(self, idx): ...  # noqa: E704
    def copy_(self, source: T) -> T: ...  # noqa: E704
    def size(self) -> int: ...  # noqa: E704
    def type(self, dtype: str = None, non_blocking: bool = False) -> T: ...  # noqa: E704
    def cuda(self, device=None, non_blocking=False, **kwargs) -> T: ...  # noqa: E704

            

Reported by Pylint.

                      if self.is_cuda:
            raise TypeError(f"cannot pin '{self.type()}' only CPU memory can be pinned")
        import torch.cuda
        allocator = torch.cuda._host_allocator()  # type: ignore[attr-defined]
        return type(self)(self.size(), allocator=allocator).copy_(self)

    def share_memory_(self):
        """Moves the storage to shared memory.


            

Reported by Pylint.

                  is_cuda: bool = False
    is_sparse: bool = False

    def __init__(self, *args, **kwargs): ...  # noqa: E704
    def __len__(self) -> int: ...  # noqa: E704
    def __getitem__(self, idx): ...  # noqa: E704
    def copy_(self, source: T) -> T: ...  # noqa: E704
    def size(self) -> int: ...  # noqa: E704
    def type(self, dtype: str = None, non_blocking: bool = False) -> T: ...  # noqa: E704

            

Reported by Pylint.

                  is_cuda: bool = False
    is_sparse: bool = False

    def __init__(self, *args, **kwargs): ...  # noqa: E704
    def __len__(self) -> int: ...  # noqa: E704
    def __getitem__(self, idx): ...  # noqa: E704
    def copy_(self, source: T) -> T: ...  # noqa: E704
    def size(self) -> int: ...  # noqa: E704
    def type(self, dtype: str = None, non_blocking: bool = False) -> T: ...  # noqa: E704

            

Reported by Pylint.

                  def __init__(self, *args, **kwargs): ...  # noqa: E704
    def __len__(self) -> int: ...  # noqa: E704
    def __getitem__(self, idx): ...  # noqa: E704
    def copy_(self, source: T) -> T: ...  # noqa: E704
    def size(self) -> int: ...  # noqa: E704
    def type(self, dtype: str = None, non_blocking: bool = False) -> T: ...  # noqa: E704
    def cuda(self, device=None, non_blocking=False, **kwargs) -> T: ...  # noqa: E704
    def element_size(self) -> int: ...  # noqa: E704
    def get_device(self) -> int: ...  # noqa: E704

            

Reported by Pylint.

                  def __getitem__(self, idx): ...  # noqa: E704
    def copy_(self, source: T) -> T: ...  # noqa: E704
    def size(self) -> int: ...  # noqa: E704
    def type(self, dtype: str = None, non_blocking: bool = False) -> T: ...  # noqa: E704
    def cuda(self, device=None, non_blocking=False, **kwargs) -> T: ...  # noqa: E704
    def element_size(self) -> int: ...  # noqa: E704
    def get_device(self) -> int: ...  # noqa: E704

    # Defined in torch/csrc/generic/StorageSharing.cpp

            

Reported by Pylint.

                  def __getitem__(self, idx): ...  # noqa: E704
    def copy_(self, source: T) -> T: ...  # noqa: E704
    def size(self) -> int: ...  # noqa: E704
    def type(self, dtype: str = None, non_blocking: bool = False) -> T: ...  # noqa: E704
    def cuda(self, device=None, non_blocking=False, **kwargs) -> T: ...  # noqa: E704
    def element_size(self) -> int: ...  # noqa: E704
    def get_device(self) -> int: ...  # noqa: E704

    # Defined in torch/csrc/generic/StorageSharing.cpp

            

Reported by Pylint.

                  def copy_(self, source: T) -> T: ...  # noqa: E704
    def size(self) -> int: ...  # noqa: E704
    def type(self, dtype: str = None, non_blocking: bool = False) -> T: ...  # noqa: E704
    def cuda(self, device=None, non_blocking=False, **kwargs) -> T: ...  # noqa: E704
    def element_size(self) -> int: ...  # noqa: E704
    def get_device(self) -> int: ...  # noqa: E704

    # Defined in torch/csrc/generic/StorageSharing.cpp
    def _share_filename_(self): ...  # noqa: E704

            

Reported by Pylint.

                  def copy_(self, source: T) -> T: ...  # noqa: E704
    def size(self) -> int: ...  # noqa: E704
    def type(self, dtype: str = None, non_blocking: bool = False) -> T: ...  # noqa: E704
    def cuda(self, device=None, non_blocking=False, **kwargs) -> T: ...  # noqa: E704
    def element_size(self) -> int: ...  # noqa: E704
    def get_device(self) -> int: ...  # noqa: E704

    # Defined in torch/csrc/generic/StorageSharing.cpp
    def _share_filename_(self): ...  # noqa: E704

            

Reported by Pylint.

              

from caffe2.python import core, dyndep
from hypothesis import given

import caffe2.python.hypothesis_test_util as hu
import hypothesis.strategies as st
import numpy as np


            

Reported by Pylint.

              from hypothesis import given

import caffe2.python.hypothesis_test_util as hu
import hypothesis.strategies as st
import numpy as np


class TestATen(hu.HypothesisTestCase):


            

Reported by Pylint.

              


from caffe2.python import core, dyndep
from hypothesis import given

import caffe2.python.hypothesis_test_util as hu
import hypothesis.strategies as st
import numpy as np

            

Reported by Pylint.

              class TestATen(hu.HypothesisTestCase):

    @given(inputs=hu.tensors(n=2), **hu.gcs)
    def test_add(self, inputs, gc, dc):
        op = core.CreateOperator(
            "ATen",
            ["X", "Y"],
            ["Z"],
            operator="add")

            

Reported by Pylint.

                      self.assertReferenceChecks(gc, op, inputs, ref)

    @given(inputs=hu.tensors(n=2, dtype=np.float16), **hu.gcs_gpu_only)
    def test_add_half(self, inputs, gc, dc):
        op = core.CreateOperator(
            "ATen",
            ["X", "Y"],
            ["Z"],
            operator="add")

            

Reported by Pylint.

                      self.assertReferenceChecks(gc, op, inputs, ref)

    @given(inputs=hu.tensors(n=1), **hu.gcs)
    def test_pow(self, inputs, gc, dc):
        op = core.CreateOperator(
            "ATen",
            ["S"],
            ["Z"],
            operator="pow", exponent=2.0)

            

Reported by Pylint.

                      self.assertReferenceChecks(gc, op, inputs, ref)

    @given(x=st.integers(min_value=2, max_value=8), **hu.gcs)
    def test_sort(self, x, gc, dc):
        inputs = [np.random.permutation(x)]
        op = core.CreateOperator(
            "ATen",
            ["S"],
            ["Z", "I"],

            

Reported by Pylint.

                      self.assertReferenceChecks(gc, op, inputs, ref)

    @given(inputs=hu.tensors(n=1), **hu.gcs)
    def test_sum(self, inputs, gc, dc):
        op = core.CreateOperator(
            "ATen",
            ["S"],
            ["Z"],
            operator="sum")

            

Reported by Pylint.

                      self.assertReferenceChecks(gc, op, inputs, ref)

    @given(**hu.gcs)
    def test_index_uint8(self, gc, dc):
        # Indexing with uint8 is deprecated, but we need to provide backward compatibility for some old models exported through ONNX
        op = core.CreateOperator(
            "ATen",
            ['self', 'mask'],
            ["Z"],

            

Reported by Pylint.

                      self.assertReferenceChecks(gc, op, [tensor, mask], ref)

    @given(**hu.gcs)
    def test_index_put(self, gc, dc):
        op = core.CreateOperator(
            "ATen",
            ['self', 'indices', 'values'],
            ["Z"],
            operator="index_put")

            

Reported by Pylint.

                  """
    return ProfilerAction.RECORD

def tensorboard_trace_handler(dir_name: str, worker_name: Optional[str] = None, use_gzip: bool = False):
    """
    Outputs tracing files to directory of ``dir_name``, then that directory can be
    directly delivered to tensorboard as logdir.
    ``worker_name`` should be unique for each worker in distributed scenario,
    it will be set to '[hostname]_[pid]' by default.

            

Reported by Pylint.

                  ``worker_name`` should be unique for each worker in distributed scenario,
    it will be set to '[hostname]_[pid]' by default.
    """
    import os
    import socket
    import time

    def handler_fn(prof) -> None:
        nonlocal worker_name

            

Reported by Pylint.

                  ``worker_name`` should be unique for each worker in distributed scenario,
    it will be set to '[hostname]_[pid]' by default.
    """
    import os
    import socket
    import time

    def handler_fn(prof) -> None:
        nonlocal worker_name

            

Reported by Pylint.

                  import socket
    import time

    def handler_fn(prof) -> None:
        nonlocal worker_name
        if not os.path.isdir(dir_name):
            try:
                os.makedirs(dir_name, exist_ok=True)
            except Exception:

            

Reported by Pylint.

                          try:
                os.makedirs(dir_name, exist_ok=True)
            except Exception:
                raise RuntimeError("Can't create directory: " + dir_name)
        if not worker_name:
            worker_name = "{}_{}".format(socket.gethostname(), str(os.getpid()))
        file_name = "{}.{}.pt.trace.json".format(worker_name, int(time.time() * 1000))
        if use_gzip:
            file_name = file_name + '.gz'

            

Reported by Pylint.

                  This, in turn, results in including CUDA time in the profiler table output,
    but not in the JSON trace.
    """
    return torch.autograd._supported_activities()


class profile(object):
    """Profiler context manager.


            

Reported by Pylint.

                          self,
            *,
            activities: Optional[Iterable[ProfilerActivity]] = None,
            schedule: Optional[Callable[[int], ProfilerAction]] = None,
            on_trace_ready: Optional[Callable[..., Any]] = None,
            record_shapes: bool = False,
            profile_memory: bool = False,
            with_stack: bool = False,
            with_flops: bool = False,

            

Reported by Pylint.

                      into the trace file
        """
        wrapped_value = "\"" + value.replace('"', '\\"') + "\""
        torch.autograd._add_metadata_json(key, wrapped_value)

    def add_metadata_json(self, key: str, value: str):
        """
        Adds a user defined metadata with a string key and a valid json value
        into the trace file

            

Reported by Pylint.

                      Adds a user defined metadata with a string key and a valid json value
        into the trace file
        """
        torch.autograd._add_metadata_json(key, value)

    def _get_distributed_info(self):
        import torch.distributed as dist
        if not dist.is_available() or not dist.is_initialized():
            return None

            

Reported by Pylint.

                          with_modules=self.with_modules,
            use_kineto=True,
        )
        self.profiler._prepare_trace()

    def _start_trace(self):
        assert self.profiler is not None
        self.profiler._start_trace()


            

Reported by Pylint.

              import math
import torch
from ..optimizer import Optimizer
from collections import defaultdict

class AdamW(Optimizer):
    r"""Implements AdamW algorithm.

    The original Adam algorithm was proposed in `Adam: A Method for Stochastic Optimization`_.

            

Reported by Pylint.

                              if len(state) == 0:
                    state['step'] = 0
                    # Exponential moving average of gradient values
                    state['exp_avg'] = torch.zeros_like(p, memory_format=torch.preserve_format)
                    # Exponential moving average of squared gradient values
                    state['exp_avg_sq'] = torch.zeros_like(p, memory_format=torch.preserve_format)
                    if amsgrad:
                        # Maintains max of all exp. moving avg. of sq. grad. values
                        state['max_exp_avg_sq'] = torch.zeros_like(p, memory_format=torch.preserve_format)

            

Reported by Pylint.

                              if len(state) == 0:
                    state['step'] = 0
                    # Exponential moving average of gradient values
                    state['exp_avg'] = torch.zeros_like(p, memory_format=torch.preserve_format)
                    # Exponential moving average of squared gradient values
                    state['exp_avg_sq'] = torch.zeros_like(p, memory_format=torch.preserve_format)
                    if amsgrad:
                        # Maintains max of all exp. moving avg. of sq. grad. values
                        state['max_exp_avg_sq'] = torch.zeros_like(p, memory_format=torch.preserve_format)

            

Reported by Pylint.

                                  # Exponential moving average of gradient values
                    state['exp_avg'] = torch.zeros_like(p, memory_format=torch.preserve_format)
                    # Exponential moving average of squared gradient values
                    state['exp_avg_sq'] = torch.zeros_like(p, memory_format=torch.preserve_format)
                    if amsgrad:
                        # Maintains max of all exp. moving avg. of sq. grad. values
                        state['max_exp_avg_sq'] = torch.zeros_like(p, memory_format=torch.preserve_format)

                exp_avg.append(state['exp_avg'])

            

Reported by Pylint.

                                  # Exponential moving average of gradient values
                    state['exp_avg'] = torch.zeros_like(p, memory_format=torch.preserve_format)
                    # Exponential moving average of squared gradient values
                    state['exp_avg_sq'] = torch.zeros_like(p, memory_format=torch.preserve_format)
                    if amsgrad:
                        # Maintains max of all exp. moving avg. of sq. grad. values
                        state['max_exp_avg_sq'] = torch.zeros_like(p, memory_format=torch.preserve_format)

                exp_avg.append(state['exp_avg'])

            

Reported by Pylint.

                                  state['exp_avg_sq'] = torch.zeros_like(p, memory_format=torch.preserve_format)
                    if amsgrad:
                        # Maintains max of all exp. moving avg. of sq. grad. values
                        state['max_exp_avg_sq'] = torch.zeros_like(p, memory_format=torch.preserve_format)

                exp_avg.append(state['exp_avg'])
                exp_avg_sq.append(state['exp_avg_sq'])

                if amsgrad:

            

Reported by Pylint.

                                  state['exp_avg_sq'] = torch.zeros_like(p, memory_format=torch.preserve_format)
                    if amsgrad:
                        # Maintains max of all exp. moving avg. of sq. grad. values
                        state['max_exp_avg_sq'] = torch.zeros_like(p, memory_format=torch.preserve_format)

                exp_avg.append(state['exp_avg'])
                exp_avg_sq.append(state['exp_avg_sq'])

                if amsgrad:

            

Reported by Pylint.

                          #
            # Decay the first and second moment running average coefficient
            #
            torch._foreach_mul_(exp_avg, beta1)
            torch._foreach_add_(exp_avg, grads, alpha=1 - beta1)

            torch._foreach_mul_(exp_avg_sq, beta2)
            torch._foreach_addcmul_(exp_avg_sq, grads, grads, 1 - beta2)


            

Reported by Pylint.

                          # Decay the first and second moment running average coefficient
            #
            torch._foreach_mul_(exp_avg, beta1)
            torch._foreach_add_(exp_avg, grads, alpha=1 - beta1)

            torch._foreach_mul_(exp_avg_sq, beta2)
            torch._foreach_addcmul_(exp_avg_sq, grads, grads, 1 - beta2)

            if amsgrad:

            

Reported by Pylint.

                          torch._foreach_mul_(exp_avg, beta1)
            torch._foreach_add_(exp_avg, grads, alpha=1 - beta1)

            torch._foreach_mul_(exp_avg_sq, beta2)
            torch._foreach_addcmul_(exp_avg_sq, grads, grads, 1 - beta2)

            if amsgrad:
                # Maintains the maximum of all 2nd moment running avg. till now
                max_exp_avg_sq = torch._foreach_maximum(max_exp_avg_sq, exp_avg_sq)

            

Reported by Pylint.

              from queue import Queue
import threading

import pytest
import torch
from torch import nn

from torch.distributed.pipeline.sync.checkpoint import enable_checkpointing, enable_recomputing
from torch.distributed.pipeline.sync.microbatch import Batch

            

Reported by Pylint.

              import threading

import pytest
import torch
from torch import nn

from torch.distributed.pipeline.sync.checkpoint import enable_checkpointing, enable_recomputing
from torch.distributed.pipeline.sync.microbatch import Batch
from torch.distributed.pipeline.sync.skip import pop, skippable, stash

            

Reported by Pylint.

              
import pytest
import torch
from torch import nn

from torch.distributed.pipeline.sync.checkpoint import enable_checkpointing, enable_recomputing
from torch.distributed.pipeline.sync.microbatch import Batch
from torch.distributed.pipeline.sync.skip import pop, skippable, stash
from torch.distributed.pipeline.sync.skip.layout import SkipLayout

            

Reported by Pylint.

              import torch
from torch import nn

from torch.distributed.pipeline.sync.checkpoint import enable_checkpointing, enable_recomputing
from torch.distributed.pipeline.sync.microbatch import Batch
from torch.distributed.pipeline.sync.skip import pop, skippable, stash
from torch.distributed.pipeline.sync.skip.layout import SkipLayout
from torch.distributed.pipeline.sync.skip.tracker import SkipTracker, SkipTrackerThroughPotals, current_skip_tracker


            

Reported by Pylint.

              from torch import nn

from torch.distributed.pipeline.sync.checkpoint import enable_checkpointing, enable_recomputing
from torch.distributed.pipeline.sync.microbatch import Batch
from torch.distributed.pipeline.sync.skip import pop, skippable, stash
from torch.distributed.pipeline.sync.skip.layout import SkipLayout
from torch.distributed.pipeline.sync.skip.tracker import SkipTracker, SkipTrackerThroughPotals, current_skip_tracker



            

Reported by Pylint.

              
from torch.distributed.pipeline.sync.checkpoint import enable_checkpointing, enable_recomputing
from torch.distributed.pipeline.sync.microbatch import Batch
from torch.distributed.pipeline.sync.skip import pop, skippable, stash
from torch.distributed.pipeline.sync.skip.layout import SkipLayout
from torch.distributed.pipeline.sync.skip.tracker import SkipTracker, SkipTrackerThroughPotals, current_skip_tracker


def test_default_skip_tracker():

            

Reported by Pylint.

              from torch.distributed.pipeline.sync.checkpoint import enable_checkpointing, enable_recomputing
from torch.distributed.pipeline.sync.microbatch import Batch
from torch.distributed.pipeline.sync.skip import pop, skippable, stash
from torch.distributed.pipeline.sync.skip.layout import SkipLayout
from torch.distributed.pipeline.sync.skip.tracker import SkipTracker, SkipTrackerThroughPotals, current_skip_tracker


def test_default_skip_tracker():
    q = Queue()

            

Reported by Pylint.

              from torch.distributed.pipeline.sync.microbatch import Batch
from torch.distributed.pipeline.sync.skip import pop, skippable, stash
from torch.distributed.pipeline.sync.skip.layout import SkipLayout
from torch.distributed.pipeline.sync.skip.tracker import SkipTracker, SkipTrackerThroughPotals, current_skip_tracker


def test_default_skip_tracker():
    q = Queue()


            

Reported by Pylint.

              def test_default_skip_tracker_by_data_parallel():
    @skippable(stash=["foo"])
    class Stash(nn.Module):
        def forward(self, input):
            yield stash("foo", input)
            return input * 2  # noqa: B901

    @skippable(pop=["foo"])
    class Pop(nn.Module):

            

Reported by Pylint.

              
    @skippable(pop=["foo"])
    class Pop(nn.Module):
        def forward(self, input):
            foo = yield pop("foo")
            return foo

    model = nn.Sequential(Stash(), Pop())
    model = nn.DataParallel(model, device_ids=[0, 0], output_device=0)

            

Reported by Pylint.

              import numpy as np

from caffe2.python import core, workspace
from hypothesis import given, settings
import caffe2.python.hypothesis_test_util as hu
import caffe2.python.serialized_test.serialized_test_util as serial
import hypothesis.strategies as st
import hypothesis.extra.numpy as hnp


            

Reported by Pylint.

              from hypothesis import given, settings
import caffe2.python.hypothesis_test_util as hu
import caffe2.python.serialized_test.serialized_test_util as serial
import hypothesis.strategies as st
import hypothesis.extra.numpy as hnp

# Basic implementation of gather for axis == 0, shich is lookup of indices
# in the outer dimension. Keeping it for reference here, although is similar
# to more general function below.

            

Reported by Pylint.

              import caffe2.python.hypothesis_test_util as hu
import caffe2.python.serialized_test.serialized_test_util as serial
import hypothesis.strategies as st
import hypothesis.extra.numpy as hnp

# Basic implementation of gather for axis == 0, shich is lookup of indices
# in the outer dimension. Keeping it for reference here, although is similar
# to more general function below.
def ref_gather_axis0():

            

Reported by Pylint.

                  @given(inputs=_inputs(),
           **hu.gcs)
    @settings(deadline=10000)
    def test_batch_gather_ops(self, inputs, gc, dc):
        data, ind = inputs
        op = core.CreateOperator(
            'BatchGather',
            ['data', 'ind'],
            ['output'])

            

Reported by Pylint.

                         index_num=st.integers(0, 5000),
           **hu.gcs)
    @settings(deadline=10000)
    def test_batch_gather_ops(self, rows_num, cols_num, index_num, gc, dc):
        data = np.random.random((rows_num, cols_num)).astype(np.float32)
        ind = np.random.randint(rows_num, size=(index_num, )).astype('int32')

        net = core.Net("bench")


            

Reported by Pylint.

                         index_num=st.integers(0, 5000),
           **hu.gcs)
    @settings(deadline=10000)
    def test_batch_gather_ops(self, rows_num, cols_num, index_num, gc, dc):
        data = np.random.random((rows_num, cols_num)).astype(np.float32)
        ind = np.random.randint(rows_num, size=(index_num, )).astype('int32')

        net = core.Net("bench")


            

Reported by Pylint.

              



import numpy as np

from caffe2.python import core, workspace
from hypothesis import given, settings
import caffe2.python.hypothesis_test_util as hu

            

Reported by Pylint.

              # Basic implementation of gather for axis == 0, shich is lookup of indices
# in the outer dimension. Keeping it for reference here, although is similar
# to more general function below.
def ref_gather_axis0():
    def inner(data, ind):
        if ind.size == 0 or data.shape[0] == 0:
            return [np.zeros((0, 10, 20)).astype(np.float32)]
        output = [data[i] for i in ind]
        return [output]

            

Reported by Pylint.

              
# Returns axis-based lookup. We just use numpy take() which handles different
# axis values as we want.
def ref_gather(axis):
    def inner(data, ind):
        if ind.size == 0 or data.shape[axis] == 0:
            shape = list(data.shape)
            shape[0] = 0
            return [np.zeros(tuple(shape)).astype(np.float32)]

            

Reported by Pylint.

                  return inner

# Gather(..., match_outer==True)
def ref_gather_match_outer(axis=1):
    def inner(data, ind):
        if ind.size == 0 or data.shape[axis] == 0:
            shape = list(data.shape)
            shape[0] = 0
            return [np.zeros(tuple(shape)).astype(np.float32)]

            

Reported by Pylint.

              import sys

from itertools import product
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.testing import FileCheck
import unittest


            

Reported by Pylint.

              
from itertools import product
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.testing import FileCheck
import unittest

try:

            

Reported by Pylint.

              from itertools import product
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.testing import FileCheck
import unittest

try:
    import torchvision

            

Reported by Pylint.

              import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.testing import FileCheck
import unittest

try:
    import torchvision
    HAS_TORCHVISION = True

            

Reported by Pylint.

              # Make the helper files in test/ importable
pytorch_test_dir = os.path.dirname(os.path.dirname(os.path.realpath(__file__)))
sys.path.append(pytorch_test_dir)
from torch.testing._internal.jit_utils import JitTestCase

if __name__ == '__main__':
    raise RuntimeError("This test file is not meant to be run directly, use:\n\n"
                       "\tpython test/test_jit.py TESTNAME\n\n"
                       "instead.")

            

Reported by Pylint.

                          torch.sigmoid_,
            torch.tanh_,
        ]:
            def test_basic(x):
                y = x + 1
                activation(y)
                return y

            fn = torch.jit.script(test_basic)

            

Reported by Pylint.

                      for activation in activations:
            def test_basic(x):
                y = x + 1
                z = activation(y)
                return z

            fn = torch.jit.script(test_basic)
            self.run_pass("inline", fn.graph)
            self.run_pass("constant_propagation", fn.graph)

            

Reported by Pylint.

                      for activation in activations:
            def test_basic(x):
                y = x + 1
                activation(y, inplace=True)
                return y

            fn = torch.jit.script(test_basic)
            self.run_pass("inline", fn.graph)
            self.run_pass("constant_propagation", fn.graph)

            

Reported by Pylint.

              import os
import sys

from itertools import product
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.testing import FileCheck
import unittest

            

Reported by Pylint.

              import torch.nn as nn
import torch.nn.functional as F
from torch.testing import FileCheck
import unittest

try:
    import torchvision
    HAS_TORCHVISION = True
except ImportError:

            

Reported by Pylint.