[Examples] Replicates the new --log_level feature to all trainer-based pytorch (#12359)

* added log_level

* fix comment

* fixed log_level

* Trigger CI

* Unfied logging

* simplified args for log_level

Author: bhadreshpsavani

examples/pytorch/language-modeling/run_clm.py

@@ -28,6 +28,7 @@
 from dataclasses import dataclass, field
 from typing import Optional
 
+import datasets
 from datasets import load_dataset
 
 import transformers
@@ -203,18 +204,19 @@ def main():
         datefmt="%m/%d/%Y %H:%M:%S",
         handlers=[logging.StreamHandler(sys.stdout)],
     )
-    logger.setLevel(logging.INFO if training_args.should_log else logging.WARN)
+
+    log_level = training_args.get_process_log_level()
+    logger.setLevel(log_level)
+    datasets.utils.logging.set_verbosity(log_level)
+    transformers.utils.logging.set_verbosity(log_level)
+    transformers.utils.logging.enable_default_handler()
+    transformers.utils.logging.enable_explicit_format()
 
     # Log on each process the small summary:
     logger.warning(
         f"Process rank: {training_args.local_rank}, device: {training_args.device}, n_gpu: {training_args.n_gpu}"
         + f"distributed training: {bool(training_args.local_rank != -1)}, 16-bits training: {training_args.fp16}"
     )
-    # Set the verbosity to info of the Transformers logger (on main process only):
-    if training_args.should_log:
-        transformers.utils.logging.set_verbosity_info()
-        transformers.utils.logging.enable_default_handler()
-        transformers.utils.logging.enable_explicit_format()
     logger.info(f"Training/evaluation parameters {training_args}")
 
     # Detecting last checkpoint.
@@ -246,15 +248,17 @@ def main():
     # download the dataset.
     if data_args.dataset_name is not None:
         # Downloading and loading a dataset from the hub.
-        datasets = load_dataset(data_args.dataset_name, data_args.dataset_config_name, cache_dir=model_args.cache_dir)
-        if "validation" not in datasets.keys():
-            datasets["validation"] = load_dataset(
+        raw_datasets = load_dataset(
+            data_args.dataset_name, data_args.dataset_config_name, cache_dir=model_args.cache_dir
+        )
+        if "validation" not in raw_datasets.keys():
+            raw_datasets["validation"] = load_dataset(
                 data_args.dataset_name,
                 data_args.dataset_config_name,
                 split=f"train[:{data_args.validation_split_percentage}%]",
                 cache_dir=model_args.cache_dir,
             )
-            datasets["train"] = load_dataset(
+            raw_datasets["train"] = load_dataset(
                 data_args.dataset_name,
                 data_args.dataset_config_name,
                 split=f"train[{data_args.validation_split_percentage}%:]",
@@ -273,7 +277,7 @@ def main():
         )
         if extension == "txt":
             extension = "text"
-        datasets = load_dataset(extension, data_files=data_files, cache_dir=model_args.cache_dir)
+        raw_datasets = load_dataset(extension, data_files=data_files, cache_dir=model_args.cache_dir)
     # See more about loading any type of standard or custom dataset (from files, python dict, pandas DataFrame, etc) at
     # https://huggingface.co/docs/datasets/loading_datasets.html.
 
@@ -334,9 +338,9 @@ def main():
     # Preprocessing the datasets.
     # First we tokenize all the texts.
     if training_args.do_train:
-        column_names = datasets["train"].column_names
+        column_names = raw_datasets["train"].column_names
     else:
-        column_names = datasets["validation"].column_names
+        column_names = raw_datasets["validation"].column_names
     text_column_name = "text" if "text" in column_names else column_names[0]
 
     # since this will be pickled to avoid _LazyModule error in Hasher force logger loading before tokenize_function
@@ -352,7 +356,7 @@ def tokenize_function(examples):
             )
         return output
 
-    tokenized_datasets = datasets.map(
+    tokenized_datasets = raw_datasets.map(
         tokenize_function,
         batched=True,
         num_proc=data_args.preprocessing_num_workers,

examples/pytorch/language-modeling/run_mlm.py

@@ -28,6 +28,7 @@
 from dataclasses import dataclass, field
 from typing import Optional
 
+import datasets
 from datasets import load_dataset
 
 import transformers
@@ -212,18 +213,20 @@ def main():
         datefmt="%m/%d/%Y %H:%M:%S",
         handlers=[logging.StreamHandler(sys.stdout)],
     )
-    logger.setLevel(logging.INFO if training_args.should_log else logging.WARN)
+
+    log_level = training_args.get_process_log_level()
+    logger.setLevel(log_level)
+    datasets.utils.logging.set_verbosity(log_level)
+    transformers.utils.logging.set_verbosity(log_level)
+    transformers.utils.logging.enable_default_handler()
+    transformers.utils.logging.enable_explicit_format()
 
     # Log on each process the small summary:
     logger.warning(
         f"Process rank: {training_args.local_rank}, device: {training_args.device}, n_gpu: {training_args.n_gpu}"
         + f"distributed training: {bool(training_args.local_rank != -1)}, 16-bits training: {training_args.fp16}"
     )
     # Set the verbosity to info of the Transformers logger (on main process only):
-    if training_args.should_log:
-        transformers.utils.logging.set_verbosity_info()
-        transformers.utils.logging.enable_default_handler()
-        transformers.utils.logging.enable_explicit_format()
     logger.info(f"Training/evaluation parameters {training_args}")
 
     # Detecting last checkpoint.
@@ -255,15 +258,17 @@ def main():
     # download the dataset.
     if data_args.dataset_name is not None:
         # Downloading and loading a dataset from the hub.
-        datasets = load_dataset(data_args.dataset_name, data_args.dataset_config_name, cache_dir=model_args.cache_dir)
-        if "validation" not in datasets.keys():
-            datasets["validation"] = load_dataset(
+        raw_datasets = load_dataset(
+            data_args.dataset_name, data_args.dataset_config_name, cache_dir=model_args.cache_dir
+        )
+        if "validation" not in raw_datasets.keys():
+            raw_datasets["validation"] = load_dataset(
                 data_args.dataset_name,
                 data_args.dataset_config_name,
                 split=f"train[:{data_args.validation_split_percentage}%]",
                 cache_dir=model_args.cache_dir,
             )
-            datasets["train"] = load_dataset(
+            raw_datasets["train"] = load_dataset(
                 data_args.dataset_name,
                 data_args.dataset_config_name,
                 split=f"train[{data_args.validation_split_percentage}%:]",
@@ -278,7 +283,7 @@ def main():
         extension = data_args.train_file.split(".")[-1]
         if extension == "txt":
             extension = "text"
-        datasets = load_dataset(extension, data_files=data_files, cache_dir=model_args.cache_dir)
+        raw_datasets = load_dataset(extension, data_files=data_files, cache_dir=model_args.cache_dir)
     # See more about loading any type of standard or custom dataset (from files, python dict, pandas DataFrame, etc) at
     # https://huggingface.co/docs/datasets/loading_datasets.html.
 
@@ -337,9 +342,9 @@ def main():
     # Preprocessing the datasets.
     # First we tokenize all the texts.
     if training_args.do_train:
-        column_names = datasets["train"].column_names
+        column_names = raw_datasets["train"].column_names
     else:
-        column_names = datasets["validation"].column_names
+        column_names = raw_datasets["validation"].column_names
     text_column_name = "text" if "text" in column_names else column_names[0]
 
     if data_args.max_seq_length is None:
@@ -377,7 +382,7 @@ def tokenize_function(examples):
                 return_special_tokens_mask=True,
             )
 
-        tokenized_datasets = datasets.map(
+        tokenized_datasets = raw_datasets.map(
             tokenize_function,
             batched=True,
             num_proc=data_args.preprocessing_num_workers,
@@ -392,7 +397,7 @@ def tokenize_function(examples):
         def tokenize_function(examples):
             return tokenizer(examples[text_column_name], return_special_tokens_mask=True)
 
-        tokenized_datasets = datasets.map(
+        tokenized_datasets = raw_datasets.map(
             tokenize_function,
             batched=True,
             num_proc=data_args.preprocessing_num_workers,

examples/pytorch/language-modeling/run_plm.py

@@ -25,6 +25,7 @@
 from dataclasses import dataclass, field
 from typing import Optional
 
+import datasets
 from datasets import load_dataset
 
 import transformers
@@ -209,18 +210,19 @@ def main():
         datefmt="%m/%d/%Y %H:%M:%S",
         handlers=[logging.StreamHandler(sys.stdout)],
     )
-    logger.setLevel(logging.INFO if training_args.should_log else logging.WARN)
+
+    log_level = training_args.get_process_log_level()
+    logger.setLevel(log_level)
+    datasets.utils.logging.set_verbosity(log_level)
+    transformers.utils.logging.set_verbosity(log_level)
+    transformers.utils.logging.enable_default_handler()
+    transformers.utils.logging.enable_explicit_format()
 
     # Log on each process the small summary:
     logger.warning(
         f"Process rank: {training_args.local_rank}, device: {training_args.device}, n_gpu: {training_args.n_gpu}"
         + f"distributed training: {bool(training_args.local_rank != -1)}, 16-bits training: {training_args.fp16}"
     )
-    # Set the verbosity to info of the Transformers logger (on main process only):
-    if training_args.should_log:
-        transformers.utils.logging.set_verbosity_info()
-        transformers.utils.logging.enable_default_handler()
-        transformers.utils.logging.enable_explicit_format()
     logger.info(f"Training/evaluation parameters {training_args}")
 
     # Detecting last checkpoint.
@@ -252,15 +254,17 @@ def main():
     # download the dataset.
     if data_args.dataset_name is not None:
         # Downloading and loading a dataset from the hub.
-        datasets = load_dataset(data_args.dataset_name, data_args.dataset_config_name, cache_dir=model_args.cache_dir)
-        if "validation" not in datasets.keys():
-            datasets["validation"] = load_dataset(
+        raw_datasets = load_dataset(
+            data_args.dataset_name, data_args.dataset_config_name, cache_dir=model_args.cache_dir
+        )
+        if "validation" not in raw_datasets.keys():
+            raw_datasets["validation"] = load_dataset(
                 data_args.dataset_name,
                 data_args.dataset_config_name,
                 split=f"train[:{data_args.validation_split_percentage}%]",
                 cache_dir=model_args.cache_dir,
             )
-            datasets["train"] = load_dataset(
+            raw_datasets["train"] = load_dataset(
                 data_args.dataset_name,
                 data_args.dataset_config_name,
                 split=f"train[{data_args.validation_split_percentage}%:]",
@@ -275,7 +279,7 @@ def main():
         extension = data_args.train_file.split(".")[-1]
         if extension == "txt":
             extension = "text"
-        datasets = load_dataset(extension, data_files=data_files, cache_dir=model_args.cache_dir)
+        raw_datasets = load_dataset(extension, data_files=data_files, cache_dir=model_args.cache_dir)
     # See more about loading any type of standard or custom dataset (from files, python dict, pandas DataFrame, etc) at
     # https://huggingface.co/docs/datasets/loading_datasets.html.
 
@@ -334,9 +338,9 @@ def main():
     # Preprocessing the datasets.
     # First we tokenize all the texts.
     if training_args.do_train:
-        column_names = datasets["train"].column_names
+        column_names = raw_datasets["train"].column_names
     else:
-        column_names = datasets["validation"].column_names
+        column_names = raw_datasets["validation"].column_names
     text_column_name = "text" if "text" in column_names else column_names[0]
 
     if data_args.max_seq_length > tokenizer.model_max_length:
@@ -355,7 +359,7 @@ def tokenize_function(examples):
             examples["text"] = [line for line in examples["text"] if len(line) > 0 and not line.isspace()]
             return tokenizer(examples["text"], padding=padding, truncation=True, max_length=max_seq_length)
 
-        tokenized_datasets = datasets.map(
+        tokenized_datasets = raw_datasets.map(
             tokenize_function,
             batched=True,
             num_proc=data_args.preprocessing_num_workers,
@@ -368,7 +372,7 @@ def tokenize_function(examples):
         def tokenize_function(examples):
             return tokenizer(examples[text_column_name])
 
-        tokenized_datasets = datasets.map(
+        tokenized_datasets = raw_datasets.map(
             tokenize_function,
             batched=True,
             num_proc=data_args.preprocessing_num_workers,

examples/pytorch/multiple-choice/run_swag.py

@@ -24,6 +24,7 @@
 from dataclasses import dataclass, field
 from typing import Optional, Union
 
+import datasets
 import numpy as np
 import torch
 from datasets import load_dataset
@@ -220,18 +221,18 @@ def main():
         datefmt="%m/%d/%Y %H:%M:%S",
         handlers=[logging.StreamHandler(sys.stdout)],
     )
-    logger.setLevel(logging.INFO if training_args.should_log else logging.WARN)
+    log_level = training_args.get_process_log_level()
+    logger.setLevel(log_level)
+    datasets.utils.logging.set_verbosity(log_level)
+    transformers.utils.logging.set_verbosity(log_level)
+    transformers.utils.logging.enable_default_handler()
+    transformers.utils.logging.enable_explicit_format()
 
     # Log on each process the small summary:
     logger.warning(
         f"Process rank: {training_args.local_rank}, device: {training_args.device}, n_gpu: {training_args.n_gpu}"
         + f"distributed training: {bool(training_args.local_rank != -1)}, 16-bits training: {training_args.fp16}"
     )
-    # Set the verbosity to info of the Transformers logger (on main process only):
-    if training_args.should_log:
-        transformers.utils.logging.set_verbosity_info()
-        transformers.utils.logging.enable_default_handler()
-        transformers.utils.logging.enable_explicit_format()
     logger.info(f"Training/evaluation parameters {training_args}")
 
     # Detecting last checkpoint.
@@ -268,10 +269,10 @@ def main():
         if data_args.validation_file is not None:
             data_files["validation"] = data_args.validation_file
         extension = data_args.train_file.split(".")[-1]
-        datasets = load_dataset(extension, data_files=data_files, cache_dir=model_args.cache_dir)
+        raw_datasets = load_dataset(extension, data_files=data_files, cache_dir=model_args.cache_dir)
     else:
         # Downloading and loading the swag dataset from the hub.
-        datasets = load_dataset("swag", "regular", cache_dir=model_args.cache_dir)
+        raw_datasets = load_dataset("swag", "regular", cache_dir=model_args.cache_dir)
     # See more about loading any type of standard or custom dataset (from files, python dict, pandas DataFrame, etc) at
     # https://huggingface.co/docs/datasets/loading_datasets.html.
 
@@ -347,9 +348,9 @@ def preprocess_function(examples):
         return {k: [v[i : i + 4] for i in range(0, len(v), 4)] for k, v in tokenized_examples.items()}
 
     if training_args.do_train:
-        if "train" not in datasets:
+        if "train" not in raw_datasets:
             raise ValueError("--do_train requires a train dataset")
-        train_dataset = datasets["train"]
+        train_dataset = raw_datasets["train"]
         if data_args.max_train_samples is not None:
             train_dataset = train_dataset.select(range(data_args.max_train_samples))
         train_dataset = train_dataset.map(
@@ -360,9 +361,9 @@ def preprocess_function(examples):
         )
 
     if training_args.do_eval:
-        if "validation" not in datasets:
+        if "validation" not in raw_datasets:
             raise ValueError("--do_eval requires a validation dataset")
-        eval_dataset = datasets["validation"]
+        eval_dataset = raw_datasets["validation"]
         if data_args.max_eval_samples is not None:
             eval_dataset = eval_dataset.select(range(data_args.max_eval_samples))
         eval_dataset = eval_dataset.map(