alperyeg

from torch.nn.functional import one_hot
import torch


def enkf_lstsq(ens, model_out, obs, gamma, batch_s, ensemble_size):
    for i in range(batch_s):
        g_tmp = model_out[:, :, i]
        Cpp = torch.tensordot(
            (g_tmp - g_tmp.mean(0)), (g_tmp - g_tmp.mean(0)), dims=([0], [0])) / ensemble_size
        Cup = torch.tensordot(

alperyeg

import torch
import torch.nn as nn
import torch.nn.functional as F
import numpy as np

from torch import optim
from torchvision import datasets, transforms

torch.manual_seed(1)
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")

alperyeg

from sklearn.datasets import fetch_openml
from sklearn.utils import check_random_state
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler


def fetch_data(test_size=10000, randomize=False, standardize=True):
    X, y = fetch_openml('mnist_784', version=1, return_X_y=True)
    if randomize:
        random_state = check_random_state(0)

alperyeg

Timer unit: 1e-06 s

Total time: 10.4797 s
Function: _cch_memory at line 397

Line #      Hits         Time  Per Hit   % Time  Line Contents
==============================================================
   397                                               @profile
   398                                               def _cch_memory(st_1, st_2, win, mode, norm, border_corr, binary, kern):
   399