Fix E722 flake8 warnings (x26).

examples/contrib/run_swag.py

@@ -44,7 +44,7 @@ from transformers import (
 
 try:
     from torch.utils.tensorboard import SummaryWriter
-except:
+except ImportError:
     from tensorboardX import SummaryWriter
 
 

examples/distillation/distiller.py

@@ -37,7 +37,7 @@ from utils import logger
 
 try:
     from torch.utils.tensorboard import SummaryWriter
-except:
+except ImportError:
     from tensorboardX import SummaryWriter
 
 

examples/distillation/run_squad_w_distillation.py

@@ -67,7 +67,7 @@ from ..utils_squad_evaluate import main as evaluate_on_squad
 
 try:
     from torch.utils.tensorboard import SummaryWriter
-except:
+except ImportError:
     from tensorboardX import SummaryWriter
 
 

examples/mm-imdb/run_mmimdb.py

@@ -62,7 +62,7 @@ from utils_mmimdb import ImageEncoder, JsonlDataset, collate_fn, get_image_trans
 
 try:
     from torch.utils.tensorboard import SummaryWriter
-except:
+except ImportError:
     from tensorboardX import SummaryWriter
 
 

examples/pplm/run_pplm.py

@@ -697,8 +697,8 @@ def run_pplm_example(
             print("= Perturbed generated text {} =".format(i + 1))
             print(pert_gen_text)
             print()
-        except:
+        except Exception as exc:
-            pass
+            print("Ignoring error while generating perturbed text:", exc)
 
         # keep the prefix, perturbed seq, original seq for each index
         generated_texts.append((tokenized_cond_text, pert_gen_tok_text, unpert_gen_tok_text))

examples/pplm/run_pplm_discrim_train.py

@@ -285,7 +285,7 @@ def train_discriminator(
             for i, line in enumerate(f):
                 try:
                     data.append(eval(line))
-                except:
+                except Exception:
                     print("Error evaluating line {}: {}".format(i, line))
                     continue
         x = []
@@ -303,7 +303,7 @@ def train_discriminator(
                         continue
                     x.append(seq)
                     y.append(d["label"])
-                except:
+                except Exception:
                     print("Error evaluating / tokenizing" " line {}, skipping it".format(i))
                     pass
 
@@ -343,7 +343,7 @@ def train_discriminator(
                         continue
                     x.append(seq)
                     y.append(int(np.sum(d["label"]) > 0))
-                except:
+                except Exception:
                     print("Error evaluating / tokenizing" " line {}, skipping it".format(i))
                     pass
 
@@ -402,7 +402,7 @@ def train_discriminator(
                         x.append(seq)
                         y.append(class2idx[label])
 
-                    except:
+                    except Exception:
                         print("Error tokenizing line {}, skipping it".format(i))
                         pass
 

examples/run_glue.py

@@ -64,7 +64,7 @@ from transformers import glue_processors as processors
 
 try:
     from torch.utils.tensorboard import SummaryWriter
-except:
+except ImportError:
     from tensorboardX import SummaryWriter
 
 

examples/run_lm_finetuning.py

@@ -63,7 +63,7 @@ from transformers import (
 
 try:
     from torch.utils.tensorboard import SummaryWriter
-except:
+except ImportError:
     from tensorboardX import SummaryWriter
 
 

examples/run_multiple_choice.py

@@ -48,7 +48,7 @@ from utils_multiple_choice import convert_examples_to_features, processors
 
 try:
     from torch.utils.tensorboard import SummaryWriter
-except:
+except ImportError:
     from tensorboardX import SummaryWriter
 
 

examples/run_squad.py

@@ -64,7 +64,7 @@ from transformers.data.processors.squad import SquadResult, SquadV1Processor, Sq
 
 try:
     from torch.utils.tensorboard import SummaryWriter
-except:
+except ImportError:
     from tensorboardX import SummaryWriter
 
 

examples/run_xnli.py

@@ -52,7 +52,7 @@ from transformers import xnli_processors as processors
 
 try:
     from torch.utils.tensorboard import SummaryWriter
-except:
+except ImportError:
     from tensorboardX import SummaryWriter
 
 

templates/adding_a_new_example_script/run_xxx.py

@@ -63,7 +63,7 @@ from utils_squad_evaluate import main as evaluate_on_squad
 
 try:
     from torch.utils.tensorboard import SummaryWriter
-except:
+except ImportError:
     from tensorboardX import SummaryWriter
 
 

transformers/__init__.py

@@ -6,12 +6,12 @@ __version__ = "2.3.0"
 # and: https://github.com/tensorflow/tensorflow/issues/26691#issuecomment-500369493
 try:
     import absl.logging
-
+except ImportError:
+    pass
+else:
     absl.logging.set_verbosity("info")
     absl.logging.set_stderrthreshold("info")
     absl.logging._warn_preinit_stderr = False
-except:
-    pass
 
 import logging
 

transformers/hf_api.py

@@ -205,10 +205,8 @@ class HfFolder:
         try:
             with open(cls.path_token, "r") as f:
                 return f.read()
-        except:
+        except FileNotFoundError:
-            # this is too wide. When Py2 is dead use:
+            pass
-            # `except FileNotFoundError:` instead
-            return None
 
     @classmethod
     def delete_token(cls):
@@ -218,5 +216,5 @@ class HfFolder:
         """
         try:
             os.remove(cls.path_token)
-        except:
+        except FileNotFoundError:
-            return
+            pass

transformers/modeling_utils.py

@@ -439,7 +439,7 @@ class PreTrainedModel(nn.Module):
         if state_dict is None and not from_tf:
             try:
                 state_dict = torch.load(resolved_archive_file, map_location="cpu")
-            except:
+            except Exception:
                 raise OSError(
                     "Unable to load weights from pytorch checkpoint file. "
                     "If you tried to load a PyTorch model from a TF 2.0 checkpoint, please set from_tf=True. "

transformers/tests/modeling_tf_common_test.py

@@ -333,13 +333,13 @@ class TFCommonTestCases:
             # We used to fall back to just synthetically creating a dummy tensor of ones:
             try:
                 x = wte(input_ids, mode="embedding")
-            except:
+            except Exception:
                 try:
                     x = wte([input_ids], mode="embedding")
-                except:
+                except Exception:
                     try:
                         x = wte([input_ids, None, None, None], mode="embedding")
-                    except:
+                    except Exception:
                         if hasattr(self.model_tester, "embedding_size"):
                             x = tf.ones(input_ids.shape + [self.model_tester.embedding_size], dtype=tf.dtypes.float32)
                         else:

transformers/tokenization_ctrl.py

@@ -168,11 +168,12 @@ class CTRLTokenizer(PreTrainedTokenizer):
             while i < len(word):
                 try:
                     j = word.index(first, i)
-                    new_word.extend(word[i:j])
+                except ValueError:
-                    i = j
-                except:
                     new_word.extend(word[i:])
                     break
+                else:
+                    new_word.extend(word[i:j])
+                    i = j
 
                 if word[i] == first and i < len(word) - 1 and word[i + 1] == second:
                     new_word.append(first + second)

transformers/tokenization_gpt2.py

@@ -178,11 +178,12 @@ class GPT2Tokenizer(PreTrainedTokenizer):
             while i < len(word):
                 try:
                     j = word.index(first, i)
-                    new_word.extend(word[i:j])
+                except ValueError:
-                    i = j
-                except:
                     new_word.extend(word[i:])
                     break
+                else:
+                    new_word.extend(word[i:j])
+                    i = j
 
                 if word[i] == first and i < len(word) - 1 and word[i + 1] == second:
                     new_word.append(first + second)

transformers/tokenization_openai.py

@@ -136,11 +136,12 @@ class OpenAIGPTTokenizer(PreTrainedTokenizer):
             while i < len(word):
                 try:
                     j = word.index(first, i)
-                    new_word.extend(word[i:j])
+                except ValueError:
-                    i = j
-                except:
                     new_word.extend(word[i:])
                     break
+                else:
+                    new_word.extend(word[i:j])
+                    i = j
 
                 if word[i] == first and i < len(word) - 1 and word[i + 1] == second:
                     new_word.append(first + second)

transformers/tokenization_xlm.py

@@ -683,11 +683,12 @@ class XLMTokenizer(PreTrainedTokenizer):
             while i < len(word):
                 try:
                     j = word.index(first, i)
-                    new_word.extend(word[i:j])
+                except ValueError:
-                    i = j
-                except:
                     new_word.extend(word[i:])
                     break
+                else:
+                    new_word.extend(word[i:j])
+                    i = j
 
                 if word[i] == first and i < len(word) - 1 and word[i + 1] == second:
                     new_word.append(first + second)