Spaces:
Sleeping
Sleeping
rzimmerdev
commited on
Commit
β’
6ec3bf6
1
Parent(s):
590d01a
Finished MNIST downloading and caching modules
Browse files- notebooks/dataloader.ipynb +0 -198
- notebooks/dataset.ipynb +187 -0
- src/dataset.py +57 -0
- {datasets β src}/downloader.py +21 -13
notebooks/dataloader.ipynb
DELETED
@@ -1,198 +0,0 @@
|
|
1 |
-
{
|
2 |
-
"cells": [
|
3 |
-
{
|
4 |
-
"cell_type": "code",
|
5 |
-
"execution_count": 1,
|
6 |
-
"metadata": {
|
7 |
-
"collapsed": true,
|
8 |
-
"pycharm": {
|
9 |
-
"name": "#%%\n"
|
10 |
-
}
|
11 |
-
},
|
12 |
-
"outputs": [
|
13 |
-
{
|
14 |
-
"name": "stdout",
|
15 |
-
"output_type": "stream",
|
16 |
-
"text": [
|
17 |
-
"/mnt/c/Users/rzimm/Workspace/data/zero-to-hero\n"
|
18 |
-
]
|
19 |
-
}
|
20 |
-
],
|
21 |
-
"source": [
|
22 |
-
"%cd ..\n",
|
23 |
-
"%load_ext autoreload\n",
|
24 |
-
"%autoreload 2\n",
|
25 |
-
"from datasets import downloader"
|
26 |
-
]
|
27 |
-
},
|
28 |
-
{
|
29 |
-
"cell_type": "code",
|
30 |
-
"execution_count": 56,
|
31 |
-
"outputs": [],
|
32 |
-
"source": [
|
33 |
-
"import pandas as pd\n",
|
34 |
-
"import numpy as np\n",
|
35 |
-
"import random\n",
|
36 |
-
"from glob import glob, escape\n",
|
37 |
-
"import imageio.v2 as imageio"
|
38 |
-
],
|
39 |
-
"metadata": {
|
40 |
-
"collapsed": false,
|
41 |
-
"pycharm": {
|
42 |
-
"name": "#%%\n"
|
43 |
-
}
|
44 |
-
}
|
45 |
-
},
|
46 |
-
{
|
47 |
-
"cell_type": "code",
|
48 |
-
"execution_count": null,
|
49 |
-
"outputs": [],
|
50 |
-
"source": [
|
51 |
-
"# download.download(\"cityscapes\", \"datasets/downloaded\")"
|
52 |
-
],
|
53 |
-
"metadata": {
|
54 |
-
"collapsed": false,
|
55 |
-
"pycharm": {
|
56 |
-
"name": "#%%\n",
|
57 |
-
"is_executing": true
|
58 |
-
}
|
59 |
-
}
|
60 |
-
},
|
61 |
-
{
|
62 |
-
"cell_type": "code",
|
63 |
-
"execution_count": 41,
|
64 |
-
"outputs": [],
|
65 |
-
"source": [
|
66 |
-
"def load_dataset(name=\"gtFine\", path=\"datasets/downloads/\"):\n",
|
67 |
-
" src = path+name\n",
|
68 |
-
" test, train, val = [f\"{src}/{subpath}\" for subpath in [\"test\", \"train\", \"val\"]]\n",
|
69 |
-
"\n",
|
70 |
-
" dataset = {\"test\": glob(test + \"/*/*\"), \"train\": glob(train + \"/*/*\"), \"val\": glob(val + \"/*/*\")}\n",
|
71 |
-
"\n",
|
72 |
-
" return dataset"
|
73 |
-
],
|
74 |
-
"metadata": {
|
75 |
-
"collapsed": false,
|
76 |
-
"pycharm": {
|
77 |
-
"name": "#%%\n"
|
78 |
-
}
|
79 |
-
}
|
80 |
-
},
|
81 |
-
{
|
82 |
-
"cell_type": "code",
|
83 |
-
"execution_count": 44,
|
84 |
-
"outputs": [
|
85 |
-
{
|
86 |
-
"data": {
|
87 |
-
"text/plain": "list"
|
88 |
-
},
|
89 |
-
"execution_count": 44,
|
90 |
-
"metadata": {},
|
91 |
-
"output_type": "execute_result"
|
92 |
-
}
|
93 |
-
],
|
94 |
-
"source": [
|
95 |
-
"type(load_dataset()[\"train\"])"
|
96 |
-
],
|
97 |
-
"metadata": {
|
98 |
-
"collapsed": false,
|
99 |
-
"pycharm": {
|
100 |
-
"name": "#%%\n"
|
101 |
-
}
|
102 |
-
}
|
103 |
-
},
|
104 |
-
{
|
105 |
-
"cell_type": "code",
|
106 |
-
"execution_count": 45,
|
107 |
-
"outputs": [],
|
108 |
-
"source": [
|
109 |
-
"a = [1, 2, 3]"
|
110 |
-
],
|
111 |
-
"metadata": {
|
112 |
-
"collapsed": false,
|
113 |
-
"pycharm": {
|
114 |
-
"name": "#%%\n"
|
115 |
-
}
|
116 |
-
}
|
117 |
-
},
|
118 |
-
{
|
119 |
-
"cell_type": "code",
|
120 |
-
"execution_count": 143,
|
121 |
-
"outputs": [],
|
122 |
-
"source": [
|
123 |
-
"class DataLoader:\n",
|
124 |
-
" def __init__(self, data):\n",
|
125 |
-
" self.data = np.array(data)\n",
|
126 |
-
" self.total = len(self.data)\n",
|
127 |
-
" self.__items = self.data\n",
|
128 |
-
" self.__remaining = len(self.data)\n",
|
129 |
-
" def __next__(self, n=1):\n",
|
130 |
-
" if n > self.total:\n",
|
131 |
-
" raise ValueError(f\"Dataset doesn't have enough elements to suffice request of {n} elements.\")\n",
|
132 |
-
" if self.__remaining > 0:\n",
|
133 |
-
" indices = random.sample(range(self.__remaining), n)\n",
|
134 |
-
" sampled = self.__items[indices]\n",
|
135 |
-
" self.__items = np.delete(self.__items, indices)\n",
|
136 |
-
" self.__remaining -= n\n",
|
137 |
-
" return sampled\n",
|
138 |
-
" else:\n",
|
139 |
-
" self.__items = self.data\n",
|
140 |
-
" self.__remaining = len(self.data)\n",
|
141 |
-
" return self.__next__(n)"
|
142 |
-
],
|
143 |
-
"metadata": {
|
144 |
-
"collapsed": false,
|
145 |
-
"pycharm": {
|
146 |
-
"name": "#%%\n"
|
147 |
-
}
|
148 |
-
}
|
149 |
-
},
|
150 |
-
{
|
151 |
-
"cell_type": "code",
|
152 |
-
"execution_count": 144,
|
153 |
-
"outputs": [],
|
154 |
-
"source": [
|
155 |
-
"loader = DataLoader(a)"
|
156 |
-
],
|
157 |
-
"metadata": {
|
158 |
-
"collapsed": false,
|
159 |
-
"pycharm": {
|
160 |
-
"name": "#%%\n"
|
161 |
-
}
|
162 |
-
}
|
163 |
-
},
|
164 |
-
{
|
165 |
-
"cell_type": "code",
|
166 |
-
"execution_count": null,
|
167 |
-
"outputs": [],
|
168 |
-
"source": [],
|
169 |
-
"metadata": {
|
170 |
-
"collapsed": false,
|
171 |
-
"pycharm": {
|
172 |
-
"name": "#%%\n"
|
173 |
-
}
|
174 |
-
}
|
175 |
-
}
|
176 |
-
],
|
177 |
-
"metadata": {
|
178 |
-
"kernelspec": {
|
179 |
-
"display_name": "Python 3",
|
180 |
-
"language": "python",
|
181 |
-
"name": "python3"
|
182 |
-
},
|
183 |
-
"language_info": {
|
184 |
-
"codemirror_mode": {
|
185 |
-
"name": "ipython",
|
186 |
-
"version": 2
|
187 |
-
},
|
188 |
-
"file_extension": ".py",
|
189 |
-
"mimetype": "text/x-python",
|
190 |
-
"name": "python",
|
191 |
-
"nbconvert_exporter": "python",
|
192 |
-
"pygments_lexer": "ipython2",
|
193 |
-
"version": "2.7.6"
|
194 |
-
}
|
195 |
-
},
|
196 |
-
"nbformat": 4,
|
197 |
-
"nbformat_minor": 0
|
198 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
notebooks/dataset.ipynb
ADDED
@@ -0,0 +1,187 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cells": [
|
3 |
+
{
|
4 |
+
"cell_type": "code",
|
5 |
+
"execution_count": 2,
|
6 |
+
"outputs": [],
|
7 |
+
"source": [
|
8 |
+
"import os\n",
|
9 |
+
"import gzip"
|
10 |
+
],
|
11 |
+
"metadata": {
|
12 |
+
"collapsed": false,
|
13 |
+
"pycharm": {
|
14 |
+
"name": "#%%\n"
|
15 |
+
}
|
16 |
+
}
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"cell_type": "code",
|
20 |
+
"execution_count": 3,
|
21 |
+
"outputs": [
|
22 |
+
{
|
23 |
+
"name": "stderr",
|
24 |
+
"output_type": "stream",
|
25 |
+
"text": [
|
26 |
+
"/home/rzimmerdev/conda/envs/data/lib/python3.9/site-packages/_distutils_hack/__init__.py:33: UserWarning: Setuptools is replacing distutils.\n",
|
27 |
+
" warnings.warn(\"Setuptools is replacing distutils.\")\n"
|
28 |
+
]
|
29 |
+
}
|
30 |
+
],
|
31 |
+
"source": [
|
32 |
+
"from src.downloader import download_dataset\n",
|
33 |
+
"download_dataset(\"mnist\", \"../datasets/mnist\")"
|
34 |
+
],
|
35 |
+
"metadata": {
|
36 |
+
"collapsed": false,
|
37 |
+
"pycharm": {
|
38 |
+
"name": "#%%\n"
|
39 |
+
}
|
40 |
+
}
|
41 |
+
},
|
42 |
+
{
|
43 |
+
"cell_type": "code",
|
44 |
+
"execution_count": 4,
|
45 |
+
"outputs": [
|
46 |
+
{
|
47 |
+
"data": {
|
48 |
+
"text/plain": "b'\\x00\\x00\\x08\\x03\\x00\\x00\\xea`\\x00\\x00\\x00\\x1c\\x00\\x00\\x00\\x1c'"
|
49 |
+
},
|
50 |
+
"execution_count": 4,
|
51 |
+
"metadata": {},
|
52 |
+
"output_type": "execute_result"
|
53 |
+
}
|
54 |
+
],
|
55 |
+
"source": [
|
56 |
+
"f = gzip.open(\"../datasets/mnist/\" + os.listdir(\"../datasets/mnist/\")[0], 'r')\n",
|
57 |
+
"f.read(16)"
|
58 |
+
],
|
59 |
+
"metadata": {
|
60 |
+
"collapsed": false,
|
61 |
+
"pycharm": {
|
62 |
+
"name": "#%%\n"
|
63 |
+
}
|
64 |
+
}
|
65 |
+
},
|
66 |
+
{
|
67 |
+
"cell_type": "code",
|
68 |
+
"execution_count": 5,
|
69 |
+
"outputs": [],
|
70 |
+
"source": [
|
71 |
+
"import numpy as np\n",
|
72 |
+
"from torch.utils.data import DataLoader, Dataset"
|
73 |
+
],
|
74 |
+
"metadata": {
|
75 |
+
"collapsed": false,
|
76 |
+
"pycharm": {
|
77 |
+
"name": "#%%\n"
|
78 |
+
}
|
79 |
+
}
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"cell_type": "code",
|
83 |
+
"execution_count": 6,
|
84 |
+
"outputs": [],
|
85 |
+
"source": [
|
86 |
+
"class DatasetMNIST(Dataset):\n",
|
87 |
+
" def __init__(self, images, labels):\n",
|
88 |
+
" with gzip.open(images, 'r') as f:\n",
|
89 |
+
" f.read(4)\n",
|
90 |
+
" self.total = int.from_bytes(f.read(4), 'big')\n",
|
91 |
+
" rows = int.from_bytes(f.read(4), 'big')\n",
|
92 |
+
" columns = int.from_bytes(f.read(4), 'big')\n",
|
93 |
+
"\n",
|
94 |
+
" image_data = f.read()\n",
|
95 |
+
" images = np.frombuffer(image_data, dtype=np.uint8)\\\n",
|
96 |
+
" .reshape((self.total, rows, columns))\n",
|
97 |
+
" self.images = images\n",
|
98 |
+
" with gzip.open(labels, 'r') as f:\n",
|
99 |
+
" f.read(4)\n",
|
100 |
+
" total = int.from_bytes(f.read(4), 'big')\n",
|
101 |
+
"\n",
|
102 |
+
" label_data = f.read()\n",
|
103 |
+
" labels = np.frombuffer(label_data, dtype=np.uint8)\n",
|
104 |
+
" self.labels = labels\n",
|
105 |
+
" self.data = list(zip(self.images, self.labels))\n",
|
106 |
+
" def __getitem__(self, n):\n",
|
107 |
+
" if n > self.total:\n",
|
108 |
+
" raise ValueError(f\"Dataset doesn't have enough elements to suffice request of {n} elements.\")\n",
|
109 |
+
" return self.data[n]\n",
|
110 |
+
"\n",
|
111 |
+
" def __len__(self):\n",
|
112 |
+
" return len(self.data)"
|
113 |
+
],
|
114 |
+
"metadata": {
|
115 |
+
"collapsed": false,
|
116 |
+
"pycharm": {
|
117 |
+
"name": "#%%\n"
|
118 |
+
}
|
119 |
+
}
|
120 |
+
},
|
121 |
+
{
|
122 |
+
"cell_type": "code",
|
123 |
+
"execution_count": 7,
|
124 |
+
"outputs": [],
|
125 |
+
"source": [
|
126 |
+
"dataset_dir = \"../datasets/mnist/\"\n",
|
127 |
+
"loader = DatasetMNIST(dataset_dir + \"train_images\", dataset_dir + \"train_labels\")"
|
128 |
+
],
|
129 |
+
"metadata": {
|
130 |
+
"collapsed": false,
|
131 |
+
"pycharm": {
|
132 |
+
"name": "#%%\n"
|
133 |
+
}
|
134 |
+
}
|
135 |
+
},
|
136 |
+
{
|
137 |
+
"cell_type": "code",
|
138 |
+
"execution_count": 8,
|
139 |
+
"outputs": [
|
140 |
+
{
|
141 |
+
"data": {
|
142 |
+
"text/plain": "<Figure size 432x288 with 1 Axes>",
|
143 |
+
"image/png": "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\n"
|
144 |
+
},
|
145 |
+
"metadata": {
|
146 |
+
"needs_background": "light"
|
147 |
+
},
|
148 |
+
"output_type": "display_data"
|
149 |
+
}
|
150 |
+
],
|
151 |
+
"source": [
|
152 |
+
"import matplotlib.pyplot as plt\n",
|
153 |
+
"X, y = loader[4]\n",
|
154 |
+
"plt.imshow(X, cmap=\"gray\")\n",
|
155 |
+
"plt.title(label=\"Annotated label: \" + str(y))\n",
|
156 |
+
"plt.show()"
|
157 |
+
],
|
158 |
+
"metadata": {
|
159 |
+
"collapsed": false,
|
160 |
+
"pycharm": {
|
161 |
+
"name": "#%%\n"
|
162 |
+
}
|
163 |
+
}
|
164 |
+
}
|
165 |
+
],
|
166 |
+
"metadata": {
|
167 |
+
"kernelspec": {
|
168 |
+
"display_name": "Python 3",
|
169 |
+
"language": "python",
|
170 |
+
"name": "python3"
|
171 |
+
},
|
172 |
+
"language_info": {
|
173 |
+
"codemirror_mode": {
|
174 |
+
"name": "ipython",
|
175 |
+
"version": 2
|
176 |
+
},
|
177 |
+
"file_extension": ".py",
|
178 |
+
"mimetype": "text/x-python",
|
179 |
+
"name": "python",
|
180 |
+
"nbconvert_exporter": "python",
|
181 |
+
"pygments_lexer": "ipython2",
|
182 |
+
"version": "2.7.6"
|
183 |
+
}
|
184 |
+
},
|
185 |
+
"nbformat": 4,
|
186 |
+
"nbformat_minor": 0
|
187 |
+
}
|
src/dataset.py
ADDED
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
#!/usr/bin/env python
|
2 |
+
# coding: utf-8
|
3 |
+
import gzip
|
4 |
+
|
5 |
+
from src.downloader import download_dataset
|
6 |
+
|
7 |
+
import numpy as np
|
8 |
+
from torch.utils.data import Dataset
|
9 |
+
|
10 |
+
|
11 |
+
def load_mnist(download_dir):
|
12 |
+
download_dataset("mnist", download_dir)
|
13 |
+
|
14 |
+
return {"train": (download_dir + "train_images", download_dir + "train_labels"),
|
15 |
+
"test": (download_dir + "test_images", download_dir + "test_labels")}
|
16 |
+
|
17 |
+
|
18 |
+
class DatasetMNIST(Dataset):
|
19 |
+
def __init__(self, images, labels):
|
20 |
+
with gzip.open(images, 'r') as f:
|
21 |
+
f.read(4)
|
22 |
+
self.total = int.from_bytes(f.read(4), 'big')
|
23 |
+
rows = int.from_bytes(f.read(4), 'big')
|
24 |
+
columns = int.from_bytes(f.read(4), 'big')
|
25 |
+
|
26 |
+
image_data = f.read()
|
27 |
+
images = np.frombuffer(image_data, dtype=np.uint8).reshape((self.total, rows, columns))
|
28 |
+
self.images = images
|
29 |
+
with gzip.open(labels, 'r') as f:
|
30 |
+
f.read(8)
|
31 |
+
|
32 |
+
label_data = f.read()
|
33 |
+
labels = np.frombuffer(label_data, dtype=np.uint8)
|
34 |
+
self.labels = labels
|
35 |
+
self.data = list(zip(self.images, self.labels))
|
36 |
+
|
37 |
+
def __getitem__(self, n):
|
38 |
+
if n > self.total:
|
39 |
+
raise ValueError(f"Dataset doesn't have enough elements to suffice request of {n} elements.")
|
40 |
+
return self.data[n]
|
41 |
+
|
42 |
+
def __len__(self):
|
43 |
+
return len(self.data)
|
44 |
+
|
45 |
+
|
46 |
+
if __name__ == "__main__":
|
47 |
+
download_dir = "../downloads/mnist/"
|
48 |
+
mnist = load_mnist(download_dir)
|
49 |
+
|
50 |
+
dataset = DatasetMNIST(*mnist["train"])
|
51 |
+
|
52 |
+
import matplotlib.pyplot as plt
|
53 |
+
|
54 |
+
X, y = dataset[4]
|
55 |
+
plt.imshow(X, cmap="gray")
|
56 |
+
plt.title(label="Annotated label: " + str(y))
|
57 |
+
plt.show()
|
{datasets β src}/downloader.py
RENAMED
@@ -4,26 +4,38 @@
|
|
4 |
# To learn more about the dataset, access:
|
5 |
# https://www.cityscapes-dataset.com/
|
6 |
import os
|
7 |
-
import sys
|
8 |
import pip
|
|
|
9 |
|
10 |
|
11 |
-
|
12 |
-
|
13 |
-
pass
|
14 |
-
|
15 |
-
|
16 |
-
def download(name='cityscapes', path='datasets/downloads'):
|
17 |
-
"""Select one of the available and implemented datasets to download:
|
18 |
name=any(['cityscapes', 'camvid', 'labelme'])
|
19 |
"""
|
20 |
if name == 'cityscapes':
|
21 |
download_cityscapes(path)
|
|
|
|
|
22 |
else:
|
23 |
raise NotImplementedError
|
24 |
|
25 |
|
26 |
-
def
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
27 |
if hasattr(pip, 'main'):
|
28 |
pip.main(['install', 'cityscapesscripts'])
|
29 |
else:
|
@@ -36,7 +48,3 @@ def download_cityscapes(path='datasets/downloads'):
|
|
36 |
print("Invalid dataset name. Please try again.")
|
37 |
ds_name = input()
|
38 |
os.system(f"csDownload {ds_name} -d {path}/{ds_name}")
|
39 |
-
|
40 |
-
|
41 |
-
if __name__ == "__main__":
|
42 |
-
main()
|
|
|
4 |
# To learn more about the dataset, access:
|
5 |
# https://www.cityscapes-dataset.com/
|
6 |
import os
|
|
|
7 |
import pip
|
8 |
+
from urllib.request import urlretrieve
|
9 |
|
10 |
|
11 |
+
def download_dataset(name='cityscapes', path='downloads/downloads'):
|
12 |
+
"""Select one of the available and implemented downloads to download:
|
|
|
|
|
|
|
|
|
|
|
13 |
name=any(['cityscapes', 'camvid', 'labelme'])
|
14 |
"""
|
15 |
if name == 'cityscapes':
|
16 |
download_cityscapes(path)
|
17 |
+
elif name == "mnist":
|
18 |
+
pass
|
19 |
else:
|
20 |
raise NotImplementedError
|
21 |
|
22 |
|
23 |
+
def download_mnist(path="downloads/mnist"):
|
24 |
+
remote_files = {"train_images": "http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz",
|
25 |
+
"train_labels": "http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz",
|
26 |
+
"test_images": "http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz",
|
27 |
+
"test_labels": "http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz"}
|
28 |
+
if not os.path.exists(path):
|
29 |
+
os.makedirs(path)
|
30 |
+
|
31 |
+
for file in remote_files.keys():
|
32 |
+
if os.path.exists(path + "/" + file):
|
33 |
+
continue
|
34 |
+
|
35 |
+
urlretrieve(remote_files[file], path + "/" + file)
|
36 |
+
|
37 |
+
|
38 |
+
def download_cityscapes(path='downloads/cityscapes'):
|
39 |
if hasattr(pip, 'main'):
|
40 |
pip.main(['install', 'cityscapesscripts'])
|
41 |
else:
|
|
|
48 |
print("Invalid dataset name. Please try again.")
|
49 |
ds_name = input()
|
50 |
os.system(f"csDownload {ds_name} -d {path}/{ds_name}")
|
|
|
|
|
|
|
|