Function _queue_management_worker has a Cognitive Complexity of 40 (exceeds 5 allowed). Consider refactoring.

Open

def _queue_management_worker(executor_reference,
                             processes,
                             pending_work_items,
                             work_ids_queue,
                             call_queue,

• Read up
  Read up

File localGpuExecutor.py has 410 lines of code (exceeds 250 allowed). Consider refactoring.

Open

import os
#import torch.multiprocessing as multiprocessing
from concurrent import futures
import torch
torch.multiprocessing.set_start_method("spawn",force="True")#.get_context('spawn')

Function __getattr__ has a Cognitive Complexity of 31 (exceeds 5 allowed). Consider refactoring.

Open

    def __getattr__(self, name):
        try:
            return super().__getattr__(name)
        except AttributeError:
            attr = getattr(self.module, name)

• Read up
  Read up

Function forward has a Cognitive Complexity of 26 (exceeds 5 allowed). Consider refactoring.

Open

    def forward(self, x):
        x_shape = x.shape
        if len(x.shape)==2:
            x=x.view(x.shape[0],x.shape[1],1,1)
        if len(x.shape)==3:

• Read up
  Read up

Function genCur has a Cognitive Complexity of 22 (exceeds 5 allowed). Consider refactoring.

Open

def genCur(func, unique = True, minArgs = None):
    """ Generates a 'curried' version of a function. """
    def g(*myArgs, **myKwArgs):
        def f(*args, **kwArgs):
            if args or kwArgs:                  # some more args!

• Read up
  Read up

Function __getattr__ has a Cognitive Complexity of 21 (exceeds 5 allowed). Consider refactoring.

Open

    def __getattr__(self, name):
        try:
            return super().__getattr__(name)
        except AttributeError:
            attr = getattr(self.module, name)

• Read up
  Read up

File utils.py has 285 lines of code (exceeds 250 allowed). Consider refactoring.

Open

import numpy as np
import torch
import numbers
import time
import torch.nn as nn

Function forward has a Cognitive Complexity of 19 (exceeds 5 allowed). Consider refactoring.

Open

    def forward(self, x):

        N,C,H,W=x.shape
        out_h=(H+2*self.padding[0]-self.kernel_size[0]+1)//self.stride[0]
        out_w=(W+2*self.padding[0]-self.kernel_size[0]+1)//self.stride[1]

• Read up
  Read up

Function logStuff has a Cognitive Complexity of 18 (exceeds 5 allowed). Consider refactoring.

Open

    def logStuff(self, step, minibatch=None):
        metrics = {}
        if minibatch is not None and hasattr(self,'loss'):
            try: metrics['Minibatch_Loss'] = self.loss(minibatch).cpu().data.numpy()
            except (NotImplementedError, TypeError): pass

• Read up
  Read up

Function split_dataset has a Cognitive Complexity of 16 (exceeds 5 allowed). Consider refactoring.

Open

def split_dataset(dataset,splits):
    """ Inputs: A torchvision.dataset DATASET and a dictionary SPLITS
        containing fractions or number of elements for each of the new datasets.
        Allows values (0,1] or (1,N] or -1 to fill with remaining.
        Example {'train':-1,'val':.1} will create a (.9, .1) split of the dataset.

• Read up
  Read up

Function unflatten has a Cognitive Complexity of 15 (exceeds 5 allowed). Consider refactoring.

Open

def unflatten(d,sep='/'):
    """Take a dictionary with keys {'k1/k2/k3':v} to {'k1':{'k2':{'k3':v}}}
        as outputted by flatten """
    out_dict={}
    for k,v in d.items():

• Read up
  Read up

Function makeTrainer has 15 arguments (exceeds 4 allowed). Consider refactoring.

Open

def makeTrainer(*,dataset=CIFAR10,network=layer13,num_epochs=100,

Function __init__ has 13 arguments (exceeds 4 allowed). Consider refactoring.

Open

    def __init__(self, in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1,bias=True, eps=1e-2, n_iter=5, momentum=0.1, mode=4, num_groups=16,debug=False):

Function makeTrainer has 13 arguments (exceeds 4 allowed). Consider refactoring.

Open

def makeTrainer(*,dataset=CIFAR10,network=layer13,num_epochs=200,

Function makeTrainer has 12 arguments (exceeds 4 allowed). Consider refactoring.

Open

def makeTrainer(*,gen=resnetgan.Generator,disc=resnetgan.Discriminator,

Function __init__ has 12 arguments (exceeds 4 allowed). Consider refactoring.

Open

    def __init__(self, in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1,bias=True, eps=1e-2, n_iter=5, momentum=0.1, num_groups=16,sampling_stride=3):

Function makeTrainer has 12 arguments (exceeds 4 allowed). Consider refactoring.

Open

def makeTrainer(*,gen=cgan.Generator,disc=cgan.Discriminator,

Function _find_images_and_annotation has a Cognitive Complexity of 12 (exceeds 5 allowed). Consider refactoring.

Open

def _find_images_and_annotation(root_dir):
    images = {}
    attr = None
    assert os.path.exists(root_dir), "{} not exists".format(root_dir)
    for root, _, fnames in sorted(os.walk(root_dir)):

• Read up
  Read up

Function __call__ has a Cognitive Complexity of 12 (exceeds 5 allowed). Consider refactoring.

Open

    def __call__(self,cfg,i=None):
        try:
            cfg.pop('local_rank',None) #TODO: properly handle distributed
            resume = cfg.pop('resume',False)
            save = cfg.pop('save',False)

• Read up
  Read up

Function _process_worker has a Cognitive Complexity of 11 (exceeds 5 allowed). Consider refactoring.

Open

def _process_worker(call_queue, result_queue,gpu_set,gpu_pool):
    """Evaluates calls from call_queue and places the results in result_queue.
    This worker is run in a separate process.
    Args:
        call_queue: A multiprocessing.Queue of _CallItems that will be read and

• Read up
  Read up