Datasets:

sihaochen
/

propsegment

Tasks:

Text Classification

Token Classification

Text Generation

Modalities:

Languages:

Size:

ArXiv:

Tags:

google-research-datasets

Libraries:

License:

Dataset card Data Studio Files Files and versions

schen149 commited on May 26, 2023

Commit

0a2c2eb

·

1 Parent(s): bb57a96

updating dataset loader

Files changed (1) hide show

propsegment.py +5 -7

propsegment.py CHANGED Viewed

@@ -129,7 +129,8 @@ class PropSegment(datasets.GeneratorBasedBuilder):
     def _split_generators(self, dl_manager):
         config_name = self.config.name
         urls = _URLS[config_name]
-        data_dir = dl_manager.download_and_extract(urls)
         file_prefix = _CONFIG_TO_FILENAME[config_name]
         return [
@@ -137,8 +138,7 @@ class PropSegment(datasets.GeneratorBasedBuilder):
                 name=datasets.Split.TRAIN,
                 # These kwargs will be passed to _generate_examples
                 gen_kwargs={
-                    "filepath": os.path.join(
-                        data_dir, "{}.train.jsonl".format(file_prefix)),
                     "split": "train",
                 },
             ),
@@ -146,8 +146,7 @@ class PropSegment(datasets.GeneratorBasedBuilder):
                 name=datasets.Split.VALIDATION,
                 # These kwargs will be passed to _generate_examples
                 gen_kwargs={
-                    "filepath": os.path.join(
-                        data_dir, "{}.dev.jsonl".format(file_prefix)),
                     "split": "dev",
                 },
             ),
@@ -155,8 +154,7 @@ class PropSegment(datasets.GeneratorBasedBuilder):
                 name=datasets.Split.TEST,
                 # These kwargs will be passed to _generate_examples
                 gen_kwargs={
-                    "filepath": os.path.join(
-                        data_dir, "{}.test.jsonl".format(file_prefix)),
                     "split": "test"
                 },
             ),

     def _split_generators(self, dl_manager):
         config_name = self.config.name
         urls = _URLS[config_name]
+        data_dir = dl_manager.download(urls)
         file_prefix = _CONFIG_TO_FILENAME[config_name]
         return [
                 name=datasets.Split.TRAIN,
                 # These kwargs will be passed to _generate_examples
                 gen_kwargs={
+                    "filepath": data_dir["train"],
                     "split": "train",
                 },
             ),
                 name=datasets.Split.VALIDATION,
                 # These kwargs will be passed to _generate_examples
                 gen_kwargs={
+                    "filepath": data_dir["dev"],
                     "split": "dev",
                 },
             ),
                 name=datasets.Split.TEST,
                 # These kwargs will be passed to _generate_examples
                 gen_kwargs={
+                    "filepath": data_dir["test"],
                     "split": "test"
                 },
             ),