Add missing whitespace to multiline strings (#13916)

src/transformers/benchmark/benchmark_args.py

@@ -69,7 +69,7 @@ class PyTorchBenchmarkArguments(BenchmarkArguments):
         default="O1",
         metadata={
             "help": (
-                "For fp16: Apex AMP optimization level selected in ['O0', 'O1', 'O2', and 'O3']."
+                "For fp16: Apex AMP optimization level selected in ['O0', 'O1', 'O2', and 'O3']. "
                 "See details at https://nvidia.github.io/apex/amp.html"
             )
         },

src/transformers/benchmark/benchmark_tf.py

@@ -231,10 +231,10 @@ class TensorFlowBenchmark(Benchmark):
 
     def _measure_memory(self, func: Callable[[], None]) -> [Memory, MemorySummary]:
         logger.info(
-            "Note that TensorFlow allocates more memory than"
-            "it might need to speed up computation."
-            "The memory reported here corresponds to the memory"
-            "reported by `nvidia-smi`, which can vary depending"
+            "Note that TensorFlow allocates more memory than "
+            "it might need to speed up computation. "
+            "The memory reported here corresponds to the memory "
+            "reported by `nvidia-smi`, which can vary depending "
             "on total available memory on the GPU that is used."
         )
         with self.args.strategy.scope():

src/transformers/benchmark/benchmark_utils.py

@@ -801,7 +801,7 @@ class Benchmark(ABC):
                 info["cpu_ram_mb"] = bytes_to_mega_bytes(psutil.virtual_memory().total)
             else:
                 logger.warning(
-                    "Psutil not installed, we won't log available CPU memory."
+                    "Psutil not installed, we won't log available CPU memory. "
                     "Install psutil (pip install psutil) to log available CPU memory."
                 )
                 info["cpu_ram_mb"] = "N/A"

src/transformers/configuration_utils.py

@@ -314,7 +314,7 @@ class PretrainedConfig(PushToHubMixin):
         allowed_problem_types = ("regression", "single_label_classification", "multi_label_classification")
         if self.problem_type is not None and self.problem_type not in allowed_problem_types:
             raise ValueError(
-                f"The config parameter `problem_type` was not understood: received {self.problem_type}"
+                f"The config parameter `problem_type` was not understood: received {self.problem_type} "
                 "but only 'regression', 'single_label_classification' and 'multi_label_classification' are valid."
             )
 

src/transformers/convert_pytorch_checkpoint_to_tf2.py

@@ -444,7 +444,7 @@ if __name__ == "__main__":
         type=str,
         help="The config json file corresponding to the pre-trained model. \n"
         "This specifies the model architecture. If not given and "
-        "--pytorch_checkpoint_path is not given or is a shortcut name"
+        "--pytorch_checkpoint_path is not given or is a shortcut name "
         "use the configuration associated to the shortcut name on the AWS",
     )
     parser.add_argument(

src/transformers/data/data_collator.py

@@ -905,7 +905,7 @@ class DataCollatorForWholeWordMask(DataCollatorForLanguageModeling):
         """
         if not isinstance(self.tokenizer, (BertTokenizer, BertTokenizerFast)):
             warnings.warn(
-                "DataCollatorForWholeWordMask is only suitable for BertTokenizer-like tokenizers."
+                "DataCollatorForWholeWordMask is only suitable for BertTokenizer-like tokenizers. "
                 "Please refer to the documentation for more information."
             )
 

src/transformers/feature_extraction_sequence_utils.py

@@ -137,7 +137,7 @@ class SequenceFeatureExtractor(FeatureExtractionMixin):
         # The model's main input name, usually `input_values`, has be passed for padding
         if self.model_input_names[0] not in processed_features:
             raise ValueError(
-                "You should supply an instance of :class:`~transformers.BatchFeature` or list of :class:`~transformers.BatchFeature` to this method"
+                "You should supply an instance of :class:`~transformers.BatchFeature` or list of :class:`~transformers.BatchFeature` to this method "
                 f"that includes {self.model_input_names[0]}, but you provided {list(processed_features.keys())}"
             )
 

src/transformers/generation_beam_search.py

@@ -194,9 +194,9 @@ class BeamSearchScorer(BeamScorer):
 
         if "max_length" in kwargs:
             warnings.warn(
-                "Passing `max_length` to BeamSearchScorer is deprecated and has no effect."
+                "Passing `max_length` to BeamSearchScorer is deprecated and has no effect. "
                 "`max_length` should be passed directly to `beam_search(...)`, `beam_sample(...)`"
-                ",or `group_beam_search(...)`."
+                ", or `group_beam_search(...)`."
             )
 
     @property

src/transformers/generation_logits_process.py

@@ -438,7 +438,7 @@ class NoBadWordsLogitsProcessor(LogitsProcessor):
                     banned_mask_list.append([idx, token])
                 else:
                     logger.error(
-                        f"An invalid bad word ID is defined: {token}. This ID is not contained in the"
+                        f"An invalid bad word ID is defined: {token}. This ID is not contained in the "
                         f"vocabulary, and is therefore ignored."
                     )
         if not banned_mask_list and self.static_bad_words_mask is None:

src/transformers/generation_tf_utils.py

@@ -533,7 +533,7 @@ class TFGenerationMixin:
         # We cannot generate if the model does not have a LM head
         if self.get_output_embeddings() is None:
             raise AttributeError(
-                "You tried to generate sequences with a model that does not have a LM Head."
+                "You tried to generate sequences with a model that does not have a LM Head. "
                 "Please use another model class (e.g. `TFOpenAIGPTLMHeadModel`, `TFXLNetLMHeadModel`, `TFGPT2LMHeadModel`, `TFCTRLLMHeadModel`, `TFT5ForConditionalGeneration`, `TFTransfoXLLMHeadModel`)"
             )
 

src/transformers/generation_utils.py

@@ -935,7 +935,7 @@ class GenerationMixin:
         if input_ids.shape[-1] >= max_length:
             input_ids_string = "decoder_input_ids" if self.config.is_encoder_decoder else "input_ids"
             logger.warning(
-                f"Input length of {input_ids_string} is {input_ids.shape[-1]}, but ``max_length`` is set to {max_length}."
+                f"Input length of {input_ids_string} is {input_ids.shape[-1]}, but ``max_length`` is set to {max_length}. "
                 "This can lead to unexpected behavior. You should consider increasing ``config.max_length`` or ``max_length``."
             )
 

src/transformers/hf_argparser.py

@@ -84,8 +84,8 @@ class HfArgumentParser(ArgumentParser):
             # it is provided as a third-party extension mechanism.
             if isinstance(field.type, str):
                 raise ImportError(
-                    "This implementation is not compatible with Postponed Evaluation of Annotations (PEP 563),"
-                    "which can be opted in from Python 3.7 with `from __future__ import annotations`."
+                    "This implementation is not compatible with Postponed Evaluation of Annotations (PEP 563), "
+                    "which can be opted in from Python 3.7 with `from __future__ import annotations`. "
                     "We will add compatibility when Python 3.9 is released."
                 )
             typestring = str(field.type)

src/transformers/modeling_flax_pytorch_utils.py

@@ -230,7 +230,7 @@ def load_flax_weights_in_pytorch_model(pt_model, flax_state):
         if flax_key in pt_model_dict:
             if flax_tensor.shape != pt_model_dict[flax_key].shape:
                 raise ValueError(
-                    f"Flax checkpoint seems to be incorrect. Weight {flax_key_tuple} was expected"
+                    f"Flax checkpoint seems to be incorrect. Weight {flax_key_tuple} was expected "
                     f"to be of shape {pt_model_dict[flax_key].shape}, but is {flax_tensor.shape}."
                 )
             else:

src/transformers/modeling_tf_utils.py

@@ -304,7 +304,7 @@ def booleans_processing(config, **kwargs):
             or ("use_cache" in kwargs and kwargs["use_cache"] not in (None, config.use_cache))
         ):
             tf_logger.warning(
-                "The parameters `output_attentions`, `output_hidden_states` and `use_cache` cannot be updated when calling a model."
+                "The parameters `output_attentions`, `output_hidden_states` and `use_cache` cannot be updated when calling a model. "
                 "They have to be set to True/False in the config object (i.e.: `config=XConfig.from_pretrained('name', output_attentions=True)`)."
             )
 

src/transformers/modeling_utils.py

@@ -777,7 +777,7 @@ class PreTrainedModel(nn.Module, ModuleUtilsMixin, GenerationMixin, PushToHubMix
 
         if not isinstance(old_embeddings, nn.Embedding):
             raise TypeError(
-                f"Old embeddings are of type {type(old_embeddings)}, which is not an instance of {nn.Embedding}."
+                f"Old embeddings are of type {type(old_embeddings)}, which is not an instance of {nn.Embedding}. "
                 f"You should either use a different resize function or make sure that `old_embeddings` are an instance of {nn.Embedding}."
             )
 
@@ -848,7 +848,7 @@ class PreTrainedModel(nn.Module, ModuleUtilsMixin, GenerationMixin, PushToHubMix
 
         if not isinstance(old_lm_head, nn.Linear):
             raise TypeError(
-                f"Old language model head is of type {type(old_lm_head)}, which is not an instance of {nn.Linear}."
+                f"Old language model head is of type {type(old_lm_head)}, which is not an instance of {nn.Linear}. "
                 f"You should either use a different resize function or make sure that `old_lm_head` are an instance of {nn.Linear}."
             )
 
@@ -1344,8 +1344,8 @@ class PreTrainedModel(nn.Module, ModuleUtilsMixin, GenerationMixin, PushToHubMix
                     except (UnicodeDecodeError, ValueError):
                         raise OSError(
                             f"Unable to load weights from pytorch checkpoint file for '{pretrained_model_name_or_path}' "
-                            f"at '{resolved_archive_file}'"
-                            "If you tried to load a PyTorch model from a TF 2.0 checkpoint, please set from_tf=True. "
+                            f"at '{resolved_archive_file}'. "
+                            "If you tried to load a PyTorch model from a TF 2.0 checkpoint, please set from_tf=True."
                         )
 
             # set dtype to instantiate the model under:

src/transformers/models/bart/configuration_bart.py

@@ -175,7 +175,7 @@ class BartConfig(PretrainedConfig):
         if self.forced_bos_token_id is None and kwargs.get("force_bos_token_to_be_generated", False):
             self.forced_bos_token_id = self.bos_token_id
             warnings.warn(
-                f"Please make sure the config includes `forced_bos_token_id={self.bos_token_id}` in future versions."
+                f"Please make sure the config includes `forced_bos_token_id={self.bos_token_id}` in future versions. "
                 "The config can simply be saved and uploaded again to be fixed."
             )
 

src/transformers/models/beit/feature_extraction_beit.py

@@ -132,7 +132,7 @@ class BeitFeatureExtractor(FeatureExtractionMixin, ImageFeatureExtractionMixin):
 
         if not valid_images:
             raise ValueError(
-                "Images must of type `PIL.Image.Image`, `np.ndarray` or `torch.Tensor` (single example),"
+                "Images must of type `PIL.Image.Image`, `np.ndarray` or `torch.Tensor` (single example), "
                 "`List[PIL.Image.Image]`, `List[np.ndarray]` or `List[torch.Tensor]` (batch of examples)."
             )
 

src/transformers/models/bert_japanese/tokenization_bert_japanese.py

@@ -272,7 +272,7 @@ class MecabTokenizer:
                 dic_dir = unidic.DICDIR
                 if not os.path.isdir(dic_dir):
                     raise RuntimeError(
-                        "The unidic dictionary itself is not found."
+                        "The unidic dictionary itself is not found. "
                         "See https://github.com/polm/unidic-py for installation."
                     )
 

src/transformers/models/big_bird/modeling_big_bird.py

@@ -2066,7 +2066,7 @@ class BigBirdModel(BigBirdPreTrainedModel):
                 "+ additional buffer: config.num_random_blocks * config.block_size "
                 f"= {max_tokens_to_attend} with config.block_size "
                 f"= {self.config.block_size}, config.num_random_blocks "
-                f"= {self.config.num_random_blocks}."
+                f"= {self.config.num_random_blocks}. "
                 "Changing attention type to 'original_full'..."
             )
             self.set_attention_type("original_full")

src/transformers/models/bigbird_pegasus/modeling_bigbird_pegasus.py

@@ -1858,7 +1858,7 @@ class BigBirdPegasusEncoder(BigBirdPegasusPreTrainedModel):
                 "+ additional buffer: config.num_random_blocks * config.block_size "
                 f"= {max_tokens_to_attend} with config.block_size "
                 f"= {self.config.block_size}, config.num_random_blocks "
-                f"= {self.config.num_random_blocks}."
+                f"= {self.config.num_random_blocks}. "
                 "Changing attention type to 'original_full'..."
             )
             self.set_attention_type("original_full")