ArneBinder
commited on
Commit
•
dc0c501
1
Parent(s):
e484998
sort file_names_filtered to get deterministic order of entries
Browse files- dialam2024.py +6 -5
dialam2024.py
CHANGED
@@ -27,7 +27,6 @@ import os
|
|
27 |
import datasets
|
28 |
from datasets import Features, GeneratorBasedBuilder
|
29 |
|
30 |
-
|
31 |
logger = logging.getLogger(__name__)
|
32 |
|
33 |
DATA_URL = "http://dialam.arg.tech/res/files/dataset.zip"
|
@@ -143,13 +142,14 @@ def get_node_id_from_filename(filename: str) -> str:
|
|
143 |
"""Get the ID of a nodeset from a filename."""
|
144 |
|
145 |
return filename.split("nodeset")[1].split(".json")[0]
|
146 |
-
|
147 |
|
148 |
-
class DialAM2024(GeneratorBasedBuilder):
|
149 |
|
|
|
150 |
BUILDER_CONFIGS = [
|
151 |
datasets.BuilderConfig(
|
152 |
-
name="dialam_2024",
|
|
|
|
|
153 |
),
|
154 |
]
|
155 |
|
@@ -202,7 +202,8 @@ class DialAM2024(GeneratorBasedBuilder):
|
|
202 |
data_dir = os.path.abspath(dl_manager.manual_dir)
|
203 |
# collect all json files in the data_dir with glob
|
204 |
file_names = glob.glob(os.path.join(data_dir, "*.json"))
|
205 |
-
|
|
|
206 |
return [
|
207 |
datasets.SplitGenerator(
|
208 |
name=datasets.Split.TRAIN,
|
|
|
27 |
import datasets
|
28 |
from datasets import Features, GeneratorBasedBuilder
|
29 |
|
|
|
30 |
logger = logging.getLogger(__name__)
|
31 |
|
32 |
DATA_URL = "http://dialam.arg.tech/res/files/dataset.zip"
|
|
|
142 |
"""Get the ID of a nodeset from a filename."""
|
143 |
|
144 |
return filename.split("nodeset")[1].split(".json")[0]
|
|
|
145 |
|
|
|
146 |
|
147 |
+
class DialAM2024(GeneratorBasedBuilder):
|
148 |
BUILDER_CONFIGS = [
|
149 |
datasets.BuilderConfig(
|
150 |
+
name="dialam_2024",
|
151 |
+
version=datasets.Version("1.0.0"),
|
152 |
+
description="DialAM-2024 dataset",
|
153 |
),
|
154 |
]
|
155 |
|
|
|
202 |
data_dir = os.path.abspath(dl_manager.manual_dir)
|
203 |
# collect all json files in the data_dir with glob
|
204 |
file_names = glob.glob(os.path.join(data_dir, "*.json"))
|
205 |
+
# filter out blacklisted nodesets and sort to get deterministic order
|
206 |
+
file_names_filtered = sorted([fn for fn in file_names if not is_blacklisted(fn)])
|
207 |
return [
|
208 |
datasets.SplitGenerator(
|
209 |
name=datasets.Split.TRAIN,
|