Catill520 commited on
Commit
5f10b58
·
verified ·
1 Parent(s): b7943f7

Upload mc_dataset.py with huggingface_hub

Browse files
Files changed (1) hide show
  1. mc_dataset.py +51 -0
mc_dataset.py ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import datasets
2
+ import pyarrow.parquet as pq
3
+
4
+ class TrajectoryDataset(datasets.GeneratorBasedBuilder):
5
+ def _info(self):
6
+ return datasets.DatasetInfo(
7
+ description="minecraft image-action data for world model",
8
+ features=datasets.Features({
9
+ "UID": datasets.Value("string"),
10
+ "Frame_base64": datasets.Value("string"),
11
+ "Action": datasets.Sequence(datasets.Value("int32")),
12
+ }),
13
+ )
14
+
15
+
16
+ def _split_generators(self, dl_manager):
17
+
18
+ base_url = "https://huggingface.co/datasets/Catill520/MC_data/resolve/main/data/"
19
+ sub_paths = [
20
+ "find-cave-Jul-28",
21
+ ]
22
+
23
+ data_files = []
24
+ for sub_path in sub_paths:
25
+ subpath_url = f"{base_url}{sub_path}/"
26
+
27
+ files = dl_manager.list_files(subpath_url)
28
+ for file in files:
29
+ data_files.append(f"{subpath_url}{file}")
30
+
31
+
32
+ downloaded_files = dl_manager.download(data_files)
33
+
34
+
35
+ return [
36
+ datasets.SplitGenerator(name="all_data", gen_kwargs={"filepaths": downloaded_files}),
37
+ ]
38
+
39
+
40
+ def _generate_examples(self, filepaths):
41
+ for filepath in filepaths:
42
+
43
+ table = pq.read_table(filepath)
44
+ df = table.to_pandas()
45
+ for idx, row in df.iterrows():
46
+ yield idx, {
47
+ "UID": row["UID"],
48
+ "Frame_base64": row["Frame_base64"],
49
+ "Action": row["Action"],
50
+ }
51
+