t-vi

t-vi

t-vi

import torch
import pickle
import math
print ("git revision:", torch.__version__)
with open('/tmp/1.0.0a0-058c128_timings.pkl', "rb") as f:
    reference_timings_dict = pickle.load(f)

print ("""
input shape = (bs, channels) + features
mode = (training/eval)-(forward+backward)/forward

t-vi

csrc = """
#include <torch/extension.h>
#include <THC/THCDeviceUtils.cuh>
#include <THC/THCGeneral.h>
#include "ATen/ATen.h"
#include "ATen/AccumulateType.h"
#include "ATen/cuda/CUDAContext.h"
using namespace at;

t-vi

from typing import List, Tuple, Optional, Union, Any, ContextManager, Callable, overload

import builtins
import math
import pickle


class dtype: ...
_dtype = dtype

t-vi

import torch
from PIL import Image
from torch.utils.data import DataLoader
import torchvision
from torchvision import transforms, datasets
from torch.autograd import Variable
import torch.nn as nn
import torch.optim
import torch.backends.cudnn as cudnn; cudnn.benchmark = True

t-vi

import torch
indices = torch.LongTensor([[1,1,1],
                            [2,1,1]]) # must be two dimensional with one row per dimension
values = torch.arange(1,4)
size = torch.Size((3,3))
a = torch.sparse.FloatTensor(indices, values, size)
b = torch.eye(3)
b += a
print (b)

t-vi

t-vi

import torch
from torch.autograd import Variable

def linear_with_sumsq(inp, weight, bias=None):
    def provide_sumsq(inp,w,b):
        def _h(i):
            if not hasattr(w, 'grad_sumsq'):
                w.grad_sumsq = 0
            w.grad_sumsq += ((i**2).t().matmul(inp**2))*i.size(0)
            if b is not None:

t-vi

from timeit import default_timer as time
import numpy as np
from numba import cuda
import os
os.environ['NUMBAPRO_LIBDEVICE']='/usr/lib/nvidia-cuda-toolkit/libdevice/'
os.environ['NUMBAPRO_NVVM']='/usr/lib/x86_64-linux-gnu/libnvvm.so.3.1.0'

import numpy
import torch
import ctypes