Datasets:
Kurt Stolle
commited on
Commit
·
3940cc9
1
Parent(s):
0bb73c7
Cleanup documentation
Browse files
Makefile
CHANGED
@@ -1,6 +1,6 @@
|
|
1 |
.PHONY: prepare build clean
|
2 |
|
3 |
-
all: prepare build
|
4 |
|
5 |
prepare:
|
6 |
mkdir -p downloads
|
@@ -8,7 +8,4 @@ prepare:
|
|
8 |
|
9 |
build:
|
10 |
mkdir -p shards
|
11 |
-
python scripts/build.py manifest.csv data shards
|
12 |
-
|
13 |
-
clean:
|
14 |
-
rm -vr downloads shards
|
|
|
1 |
.PHONY: prepare build clean
|
2 |
|
3 |
+
all: prepare build clean
|
4 |
|
5 |
prepare:
|
6 |
mkdir -p downloads
|
|
|
8 |
|
9 |
build:
|
10 |
mkdir -p shards
|
11 |
+
python scripts/build.py manifest.csv data shards
|
|
|
|
|
|
README.md
CHANGED
@@ -78,48 +78,31 @@ The remaining data should be downloaded from official sources using the provided
|
|
78 |
|
79 |
## Preparation
|
80 |
|
81 |
-
|
82 |
```bash
|
83 |
git clone https://huggingface.co/datasets/khwstolle/csvps && cd csvps
|
84 |
```
|
85 |
|
86 |
-
|
87 |
```bash
|
88 |
python -m pip install -r requirements.txt
|
89 |
```
|
90 |
|
91 |
-
|
92 |
Note that this may prompt your [Cityscapes account](https://cityscapes-dataset.com/login/) login credentials.
|
93 |
```bash
|
94 |
make prepare
|
95 |
```
|
96 |
|
97 |
-
|
98 |
-
```bash
|
99 |
-
make clean
|
100 |
-
```
|
101 |
-
|
102 |
-
## Usage
|
103 |
-
|
104 |
-
To convert the `train`, `val` and `test` directories into a `tar` archive, run the following command:
|
105 |
|
106 |
```bash
|
107 |
-
make
|
108 |
-
```
|
109 |
-
|
110 |
-
Subsequently, the dataset can be loaded using the `webdataset` library:
|
111 |
-
|
112 |
-
```python
|
113 |
-
import webdataset as wds
|
114 |
-
|
115 |
-
dataset = wds.WebDataset("shards/csvps-train.tar")
|
116 |
```
|
117 |
|
|
|
118 |
|
119 |
-
|
120 |
-
|
121 |
-
Please refer to the Cityscapes license for more details.
|
122 |
-
|
123 |
|
124 |
## Citation
|
125 |
|
|
|
78 |
|
79 |
## Preparation
|
80 |
|
81 |
+
1. Clone this dataset repository.
|
82 |
```bash
|
83 |
git clone https://huggingface.co/datasets/khwstolle/csvps && cd csvps
|
84 |
```
|
85 |
|
86 |
+
2. Install the [Cityscapes developer kit](https://github.com/mcordts/cityscapesScripts) and build dependencies using `pip`.
|
87 |
```bash
|
88 |
python -m pip install -r requirements.txt
|
89 |
```
|
90 |
|
91 |
+
3. Run the preparation script provided in this repository.
|
92 |
Note that this may prompt your [Cityscapes account](https://cityscapes-dataset.com/login/) login credentials.
|
93 |
```bash
|
94 |
make prepare
|
95 |
```
|
96 |
|
97 |
+
4. To convert the `train`, `val` and `test` directories into a `tar` archive for use with [WebDataset](https://github.com/webdataset/webdataset), run the following command:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
98 |
|
99 |
```bash
|
100 |
+
make build
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
101 |
```
|
102 |
|
103 |
+
## Usage
|
104 |
|
105 |
+
See `examples.ipynb` for instructions.
|
|
|
|
|
|
|
106 |
|
107 |
## Citation
|
108 |
|