Datasets cli improvements

.github/ISSUE_TEMPLATE/bug-report.md

@@ -21,16 +21,9 @@ A clear and concise description of the expected results.
 ## Actual results
 Specify the actual results or traceback.
 
-## Versions
-Paste the output of the following code:
-```python
-import datasets
-import sys
-import platform
-
-print(f"""
-- Datasets: {datasets.__version__}
-- Python: {sys.version}
-- Platform: {platform.platform()}
-""")
-```
+## Environment info
+<!-- You can run the command `datasets-cli env` and copy-and-paste its output below. -->
+- `datasets` version:
+- Platform:
+- Python version:
+- PyArrow version:

src/datasets/commands/__init__.py

@@ -2,7 +2,7 @@
 from argparse import ArgumentParser
 
 
-class BaseTransformersCLICommand(ABC):
+class BaseDatasetsCLICommand(ABC):
     @staticmethod
     @abstractmethod
     def register_subcommand(parser: ArgumentParser):

src/datasets/commands/convert.py

@@ -3,7 +3,7 @@
 import shutil
 from argparse import ArgumentParser, Namespace
 
-from datasets.commands import BaseTransformersCLICommand
+from datasets.commands import BaseDatasetsCLICommand
 from datasets.utils.logging import get_logger
 
 
@@ -42,22 +42,24 @@
 def convert_command_factory(args: Namespace):
     """
     Factory function used to convert a model TF 1.0 checkpoint in a PyTorch checkpoint.
-    :return: ServeCommand
+
+    Returns: ConvertCommand
     """
     return ConvertCommand(args.tfds_path, args.datasets_directory)
 
 
-class ConvertCommand(BaseTransformersCLICommand):
+class ConvertCommand(BaseDatasetsCLICommand):
     @staticmethod
     def register_subcommand(parser: ArgumentParser):
         """
-        Register this command to argparse so it's available for the transformer-cli
-        :param parser: Root parser to register command-specific arguments
-        :return:
+        Register this command to argparse so it's available for the datasets-cli
+
+        Args:
+            parser: Root parser to register command-specific arguments
         """
         train_parser = parser.add_parser(
             "convert",
-            help="CLI tool to convert a (nlp) TensorFlow-Dataset in a HuggingFace-NLP dataset.",
+            help="Convert a TensorFlow Datasets dataset to a HuggingFace Datasets dataset.",
         )
         train_parser.add_argument(
             "--tfds_path",
@@ -66,7 +68,7 @@ def register_subcommand(parser: ArgumentParser):
             help="Path to a TensorFlow Datasets folder to convert or a single tfds file to convert.",
         )
         train_parser.add_argument(
-            "--datasets_directory", type=str, required=True, help="Path to the HuggingFace NLP folder."
+            "--datasets_directory", type=str, required=True, help="Path to the HuggingFace Datasets folder."
         )
         train_parser.set_defaults(func=convert_command_factory)
 

src/datasets/commands/datasets_cli.py

@@ -2,7 +2,6 @@
 from argparse import ArgumentParser
 
 from datasets.commands.convert import ConvertCommand
-from datasets.commands.download import DownloadCommand
 from datasets.commands.dummy_data import DummyDataCommand
 from datasets.commands.env import EnvironmentCommand
 from datasets.commands.run_beam import RunBeamCommand
@@ -17,7 +16,6 @@ def main():
 
     # Register commands
     ConvertCommand.register_subcommand(commands_parser)
-    DownloadCommand.register_subcommand(commands_parser)
     EnvironmentCommand.register_subcommand(commands_parser)
     TestCommand.register_subcommand(commands_parser)
     RunBeamCommand.register_subcommand(commands_parser)

src/datasets/commands/dummy_data.py

@@ -9,7 +9,7 @@
 from typing import Optional
 
 from datasets import config
-from datasets.commands import BaseTransformersCLICommand
+from datasets.commands import BaseDatasetsCLICommand
 from datasets.load import import_main_class, prepare_module
 from datasets.utils import MockDownloadManager
 from datasets.utils.download_manager import DownloadManager
@@ -212,10 +212,10 @@ def compress_autogenerated_dummy_data(self, path_to_dataset):
         shutil.rmtree(base_name)
 
 
-class DummyDataCommand(BaseTransformersCLICommand):
+class DummyDataCommand(BaseDatasetsCLICommand):
     @staticmethod
     def register_subcommand(parser: ArgumentParser):
-        test_parser = parser.add_parser("dummy_data")
+        test_parser = parser.add_parser("dummy_data", help="Generate dummy data.")
         test_parser.add_argument("--auto_generate", action="store_true", help="Automatically generate dummy data")
         test_parser.add_argument(
             "--n_lines", type=int, default=5, help="Number of lines or samples to keep when auto-generating dummy data"

src/datasets/commands/env.py

@@ -1,54 +1,31 @@
 import platform
 from argparse import ArgumentParser
 
+import pyarrow
+
 from datasets import __version__ as version
-from datasets import config
-from datasets.commands import BaseTransformersCLICommand
+from datasets.commands import BaseDatasetsCLICommand
 
 
 def info_command_factory(_):
     return EnvironmentCommand()
 
 
-class EnvironmentCommand(BaseTransformersCLICommand):
+class EnvironmentCommand(BaseDatasetsCLICommand):
     @staticmethod
     def register_subcommand(parser: ArgumentParser):
-        download_parser = parser.add_parser("env")
+        download_parser = parser.add_parser("env", help="Print relevant system environment info.")
         download_parser.set_defaults(func=info_command_factory)
 
     def run(self):
-        pt_version = "not installed"
-        pt_cuda_available = "NA"
-        if config.TORCH_AVAILABLE:
-            import torch
-
-            pt_version = torch.__version__
-            pt_cuda_available = torch.cuda.is_available()
-
-        tf_version = "not installed"
-        tf_cuda_available = "NA"
-        if config.TF_AVAILABLE:
-            import tensorflow as tf
-
-            tf_version = tf.__version__
-            try:
-                # deprecated in v2.1
-                tf_cuda_available = tf.test.is_gpu_available()
-            except AttributeError:
-                # returns list of devices, convert to bool
-                tf_cuda_available = bool(tf.config.list_physical_devices("GPU"))
-
         info = {
             "`datasets` version": version,
             "Platform": platform.platform(),
             "Python version": platform.python_version(),
-            "PyTorch version (GPU?)": "{} ({})".format(pt_version, pt_cuda_available),
-            "Tensorflow version (GPU?)": "{} ({})".format(tf_version, tf_cuda_available),
-            "Using GPU in script?": "<fill in>",
-            "Using distributed or parallel set-up in script?": "<fill in>",
+            "PyArrow version": pyarrow.__version__,
         }
 
-        print("\nCopy-and-paste the text below in your GitHub issue and FILL OUT the two last points.\n")
+        print("\nCopy-and-paste the text below in your GitHub issue.\n")
         print(self.format_dict(info))
 
         return info

src/datasets/commands/run_beam.py

@@ -6,7 +6,7 @@
 
 from datasets import config
 from datasets.builder import DatasetBuilder
-from datasets.commands import BaseTransformersCLICommand
+from datasets.commands import BaseDatasetsCLICommand
 from datasets.load import import_main_class, prepare_module
 from datasets.utils.download_manager import DownloadConfig, GenerateMode
 
@@ -25,10 +25,10 @@ def run_beam_command_factory(args):
     )
 
 
-class RunBeamCommand(BaseTransformersCLICommand):
+class RunBeamCommand(BaseDatasetsCLICommand):
     @staticmethod
     def register_subcommand(parser: ArgumentParser):
-        run_beam_parser = parser.add_parser("run_beam")
+        run_beam_parser = parser.add_parser("run_beam", help="Run a Beam dataset processing pipeline.")
         run_beam_parser.add_argument("--name", type=str, default=None, help="Dataset processing name")
         run_beam_parser.add_argument(
             "--cache_dir",

src/datasets/commands/test.py

@@ -6,7 +6,7 @@
 
 import datasets.config
 from datasets.builder import DatasetBuilder
-from datasets.commands import BaseTransformersCLICommand
+from datasets.commands import BaseDatasetsCLICommand
 from datasets.load import import_main_class, prepare_module
 from datasets.utils.download_manager import GenerateMode
 from datasets.utils.filelock import logger as fl_logger
@@ -32,10 +32,10 @@ def test_command_factory(args):
     )
 
 
-class TestCommand(BaseTransformersCLICommand):
+class TestCommand(BaseDatasetsCLICommand):
     @staticmethod
     def register_subcommand(parser: ArgumentParser):
-        test_parser = parser.add_parser("test")
+        test_parser = parser.add_parser("test", help="Test dataset implementation.")
         test_parser.add_argument("--name", type=str, default=None, help="Dataset processing name")
         test_parser.add_argument(
             "--cache_dir",