Add examples telemetry (#17552)

* Add examples telemetry

* Alternative approach

* Add to all other examples

* Add to templates as well

* Put framework separately

* Same for TensorFlow

examples/tensorflow/language-modeling/run_clm.py

@@ -53,6 +53,7 @@ from transformers import (
     create_optimizer,
     set_seed,
 )
+from transformers.utils import send_example_telemetry
 from transformers.utils.versions import require_version
 
 
@@ -232,6 +233,10 @@ def main():
     else:
         model_args, data_args, training_args = parser.parse_args_into_dataclasses()
 
+    # Sending telemetry. Tracking the example usage helps us better allocate resources to maintain them. The
+    # information sent is the one passed as arguments along with your Python/PyTorch versions.
+    send_example_telemetry("run_clm", model_args, data_args, framework="tensorflow")
+
     # Sanity checks
     if data_args.dataset_name is None and data_args.train_file is None and data_args.validation_file is None:
         raise ValueError("Need either a dataset name or a training/validation file.")

examples/tensorflow/language-modeling/run_mlm.py

@@ -55,6 +55,7 @@ from transformers import (
     create_optimizer,
     set_seed,
 )
+from transformers.utils import send_example_telemetry
 from transformers.utils.versions import require_version
 
 
@@ -242,6 +243,10 @@ def main():
     else:
         model_args, data_args, training_args = parser.parse_args_into_dataclasses()
 
+    # Sending telemetry. Tracking the example usage helps us better allocate resources to maintain them. The
+    # information sent is the one passed as arguments along with your Python/PyTorch versions.
+    send_example_telemetry("run_mlm", model_args, data_args, framework="tensorflow")
+
     # Sanity checks
     if data_args.dataset_name is None and data_args.train_file is None and data_args.validation_file is None:
         raise ValueError("Need either a dataset name or a training/validation file.")

examples/tensorflow/multiple-choice/run_swag.py

@@ -44,7 +44,7 @@ from transformers import (
     set_seed,
 )
 from transformers.tokenization_utils_base import PreTrainedTokenizerBase
-from transformers.utils import PaddingStrategy, check_min_version
+from transformers.utils import PaddingStrategy, check_min_version, send_example_telemetry
 
 
 # Will error if the minimal version of Transformers is not installed. Remove at your own risks.
@@ -246,6 +246,10 @@ def main():
     else:
         model_args, data_args, training_args = parser.parse_args_into_dataclasses()
 
+    # Sending telemetry. Tracking the example usage helps us better allocate resources to maintain them. The
+    # information sent is the one passed as arguments along with your Python/PyTorch versions.
+    send_example_telemetry("run_swag", model_args, data_args, framework="tensorflow")
+
     output_dir = Path(training_args.output_dir)
     output_dir.mkdir(parents=True, exist_ok=True)
     # endregion

examples/tensorflow/question-answering/run_qa.py

@@ -41,7 +41,7 @@ from transformers import (
     TFTrainingArguments,
     set_seed,
 )
-from transformers.utils import CONFIG_NAME, TF2_WEIGHTS_NAME, check_min_version
+from transformers.utils import CONFIG_NAME, TF2_WEIGHTS_NAME, check_min_version, send_example_telemetry
 from utils_qa import postprocess_qa_predictions
 
 
@@ -242,6 +242,10 @@ def main():
     else:
         model_args, data_args, training_args = parser.parse_args_into_dataclasses()
 
+    # Sending telemetry. Tracking the example usage helps us better allocate resources to maintain them. The
+    # information sent is the one passed as arguments along with your Python/PyTorch versions.
+    send_example_telemetry("run_qa", model_args, data_args, framework="tensorflow")
+
     output_dir = Path(training_args.output_dir)
     output_dir.mkdir(parents=True, exist_ok=True)
     # endregion

examples/tensorflow/summarization/run_summarization.py

@@ -44,7 +44,7 @@ from transformers import (
     set_seed,
 )
 from transformers.trainer_utils import get_last_checkpoint
-from transformers.utils import check_min_version, is_offline_mode
+from transformers.utils import check_min_version, is_offline_mode, send_example_telemetry
 from transformers.utils.versions import require_version
 
 
@@ -348,6 +348,10 @@ def main():
         model_args, data_args, training_args = parser.parse_json_file(json_file=os.path.abspath(sys.argv[1]))
     else:
         model_args, data_args, training_args = parser.parse_args_into_dataclasses()
+
+    # Sending telemetry. Tracking the example usage helps us better allocate resources to maintain them. The
+    # information sent is the one passed as arguments along with your Python/PyTorch versions.
+    send_example_telemetry("run_summarization", model_args, data_args, framework="tensorflow")
     # endregion
 
     # region Logging

examples/tensorflow/text-classification/run_glue.py

@@ -39,7 +39,7 @@ from transformers import (
     set_seed,
 )
 from transformers.trainer_utils import get_last_checkpoint, is_main_process
-from transformers.utils import check_min_version
+from transformers.utils import check_min_version, send_example_telemetry
 
 
 # region Helper functions
@@ -206,6 +206,10 @@ def main():
     else:
         model_args, data_args, training_args = parser.parse_args_into_dataclasses()
 
+    # Sending telemetry. Tracking the example usage helps us better allocate resources to maintain them. The
+    # information sent is the one passed as arguments along with your Python/PyTorch versions.
+    send_example_telemetry("run_glue", model_args, data_args, framework="tensorflow")
+
     if not (training_args.do_train or training_args.do_eval or training_args.do_predict):
         exit("Must specify at least one of --do_train, --do_eval or --do_predict!")
     # endregion

examples/tensorflow/text-classification/run_text_classification.py

@@ -37,7 +37,7 @@ from transformers import (
     TFTrainingArguments,
     set_seed,
 )
-from transformers.utils import CONFIG_NAME, TF2_WEIGHTS_NAME
+from transformers.utils import CONFIG_NAME, TF2_WEIGHTS_NAME, send_example_telemetry
 
 
 os.environ["TF_CPP_MIN_LOG_LEVEL"] = "1"  # Reduce the amount of console output from TF
@@ -196,6 +196,11 @@ def main():
         model_args, data_args, training_args = parser.parse_json_file(json_file=os.path.abspath(sys.argv[1]))
     else:
         model_args, data_args, training_args = parser.parse_args_into_dataclasses()
+
+    # Sending telemetry. Tracking the example usage helps us better allocate resources to maintain them. The
+    # information sent is the one passed as arguments along with your Python/PyTorch versions.
+    send_example_telemetry("run_text_classification", model_args, data_args, framework="tensorflow")
+
     output_dir = Path(training_args.output_dir)
     output_dir.mkdir(parents=True, exist_ok=True)
     # endregion

examples/tensorflow/token-classification/run_ner.py

@@ -41,6 +41,7 @@ from transformers import (
     create_optimizer,
     set_seed,
 )
+from transformers.utils import send_example_telemetry
 from transformers.utils.versions import require_version
 
 
@@ -252,6 +253,10 @@ def main():
     # region Argument Parsing
     parser = HfArgumentParser((ModelArguments, DataTrainingArguments, TFTrainingArguments))
     model_args, data_args, training_args = parser.parse_args_into_dataclasses()
+
+    # Sending telemetry. Tracking the example usage helps us better allocate resources to maintain them. The
+    # information sent is the one passed as arguments along with your Python/PyTorch versions.
+    send_example_telemetry("run_ner", model_args, data_args, framework="tensorflow")
     # endregion
 
     # region Setup logging

examples/tensorflow/translation/run_translation.py

@@ -47,7 +47,7 @@ from transformers import (
     set_seed,
 )
 from transformers.trainer_utils import get_last_checkpoint
-from transformers.utils import check_min_version
+from transformers.utils import check_min_version, send_example_telemetry
 from transformers.utils.versions import require_version
 
 
@@ -318,6 +318,10 @@ def main():
         model_args, data_args, training_args = parser.parse_json_file(json_file=os.path.abspath(sys.argv[1]))
     else:
         model_args, data_args, training_args = parser.parse_args_into_dataclasses()
+
+    # Sending telemetry. Tracking the example usage helps us better allocate resources to maintain them. The
+    # information sent is the one passed as arguments along with your Python/PyTorch versions.
+    send_example_telemetry("run_translation", model_args, data_args, framework="tensorflow")
     # endregion
 
     # region Logging