yunjey

import os
from PIL import Image

# Attribute
f = open('./list_attr_celeba.txt', 'r')
text = f.readlines()

# Number of images
text[0] 

yunjey

import torch
import numpy as np

# Hyper-parameters
vocab_size = 10
batch_size = 3
seq_length = 4

# Generate random indices of range [0, vocab_size)
word_indices = torch.from_numpy(np.random.randint(low=0, high=vocab_size, size=(batch_size, seq_length))).view(-1)

yunjey

# First, you should install flickrapi
# pip install flickrapi

import flickrapi
import urllib
from PIL import Image

# Flickr api access key 
flickr=flickrapi.FlickrAPI('c6a2c45591d4973ff525042472446ca2', '202ffe6f387ce29b', cache=True)

yunjey

# pip install icrawler

from icrawler.builtin import GoogleImageCrawler
from datetime import date

google_crawler = GoogleImageCrawler(parser_threads=2, downloader_threads=3,
                                    storage={'root_dir': './husky'})        # directory where images are downloaded

google_crawler.crawl(keyword='Siberian husky', max_num=1000,   # max_num should be equal or less than 1000
                     date_min=date(2012, 3, 1), date_max=date(2012, 6, 1),  # you can change date to get more than 1000 images

yunjey

def contextual_loss(x, y, h=0.5):
    """Computes contextual loss between x and y.
    
    Args:
      x: features of shape (N, C, H, W).
      y: features of shape (N, C, H, W).
      
    Returns:
      cx_loss = contextual loss between x and y (Eq (1) in the paper)
    """

yunjey

from torch.nn.utils import spectral_norm
import torch.nn.functional as F
import torch.nn as nn
import torch


class NonLocalBlock(nn.Module):
    """Non-local block."""
    def __init__(self, conv_dim):
        super(NonLocalBlock, self).__init__()

yunjey

find . -type d -print0 | while read -d '' -r dir; do
    files=("$dir"/*)
    printf "%5d files in directory %s\n" "${#files[@]}" "$dir"
done

yunjey

import os
from PIL import Image
import argparse
import time
import datetime


def main(config):
    src_dir = config.src_dir
    filenames = os.listdir(src_dir)

yunjey

import torch


x1 = torch.tensor(1.).requires_grad_(True)
x2 = torch.tensor(2.).requires_grad_(True)
w = torch.tensor(3.).requires_grad_(False)

y1 = w * x1
loss1 = torch.mean((y1-1)**2)
loss1.backward()

yunjey

{
    "tench": 0,
    "goldfish": 1,
    "great white shark": 2,
    "tiger shark": 3,
    "hammerhead": 4,
    "electric ray": 5,
    "stingray": 6,
    "cock": 7,
    "hen": 8,