Raising exceptions instead of using assertions for few models (#14219)

* raising exceptions instead of using assertions for few models

* fixed formatting issues

* fixing copy inconsistencies

src/transformers/models/bart/modeling_bart.py

@@ -67,7 +67,8 @@ def shift_tokens_right(input_ids: torch.Tensor, pad_token_id: int, decoder_start
     shifted_input_ids[:, 1:] = input_ids[:, :-1].clone()
     shifted_input_ids[:, 0] = decoder_start_token_id
 
-    assert pad_token_id is not None, "self.model.config.pad_token_id has to be defined."
+    if pad_token_id is None:
+        raise ValueError("self.model.config.pad_token_id has to be defined.")
     # replace possible -100 values in labels by `pad_token_id`
     shifted_input_ids.masked_fill_(shifted_input_ids == -100, pad_token_id)
 
@@ -786,9 +787,11 @@ class BartEncoder(BartPretrainedModel):
 
         # check if head_mask has a correct number of layers specified if desired
         if head_mask is not None:
-            assert head_mask.size()[0] == (
-                len(self.layers)
-            ), f"The head_mask should be specified for {len(self.layers)} layers, but it is for {head_mask.size()[0]}."
+            if head_mask.size()[0] != (len(self.layers)):
+                raise ValueError(
+                    f"The head_mask should be specified for {len(self.layers)} layers, but it is for {head_mask.size()[0]}."
+                )
+
         for idx, encoder_layer in enumerate(self.layers):
             if output_hidden_states:
                 encoder_states = encoder_states + (hidden_states,)
@@ -1023,9 +1026,11 @@ class BartDecoder(BartPretrainedModel):
         # check if head_mask/cross_attn_head_mask has a correct number of layers specified if desired
         for attn_mask, mask_name in zip([head_mask, cross_attn_head_mask], ["head_mask", "cross_attn_head_mask"]):
             if attn_mask is not None:
-                assert attn_mask.size()[0] == (
-                    len(self.layers)
-                ), f"The `{mask_name}` should be specified for {len(self.layers)} layers, but it is for {head_mask.size()[0]}."
+                if attn_mask.size()[0] != (len(self.layers)):
+                    raise ValueError(
+                        "The `{mask_name}` should be specified for {len(self.layers)} layers, but it is for {head_mask.size()[0]}."
+                    )
+
         for idx, decoder_layer in enumerate(self.layers):
             # add LayerDrop (see https://arxiv.org/abs/1909.11556 for description)
             if output_hidden_states:

src/transformers/models/bert/modeling_bert.py

@@ -1369,7 +1369,9 @@ class BertForMaskedLM(BertPreTrainedModel):
         effective_batch_size = input_shape[0]
 
         #  add a dummy token
-        assert self.config.pad_token_id is not None, "The PAD token should be defined for generation"
+        if self.config.pad_token_id is None:
+            raise ValueError("The PAD token should be defined for generation")
+
         attention_mask = torch.cat([attention_mask, attention_mask.new_zeros((attention_mask.shape[0], 1))], dim=-1)
         dummy_token = torch.full(
             (effective_batch_size, 1), self.config.pad_token_id, dtype=torch.long, device=input_ids.device

src/transformers/models/blenderbot/modeling_blenderbot.py

@@ -69,7 +69,8 @@ def shift_tokens_right(input_ids: torch.Tensor, pad_token_id: int, decoder_start
     shifted_input_ids[:, 1:] = input_ids[:, :-1].clone()
     shifted_input_ids[:, 0] = decoder_start_token_id
 
-    assert pad_token_id is not None, "self.model.config.pad_token_id has to be defined."
+    if pad_token_id is None:
+        raise ValueError("self.model.config.pad_token_id has to be defined.")
     # replace possible -100 values in labels by `pad_token_id`
     shifted_input_ids.masked_fill_(shifted_input_ids == -100, pad_token_id)
 

src/transformers/models/blenderbot_small/modeling_blenderbot_small.py

@@ -66,7 +66,8 @@ def shift_tokens_right(input_ids: torch.Tensor, pad_token_id: int, decoder_start
     shifted_input_ids[:, 1:] = input_ids[:, :-1].clone()
     shifted_input_ids[:, 0] = decoder_start_token_id
 
-    assert pad_token_id is not None, "self.model.config.pad_token_id has to be defined."
+    if pad_token_id is None:
+        raise ValueError("self.model.config.pad_token_id has to be defined.")
     # replace possible -100 values in labels by `pad_token_id`
     shifted_input_ids.masked_fill_(shifted_input_ids == -100, pad_token_id)
 

src/transformers/models/m2m_100/modeling_m2m_100.py

@@ -64,7 +64,8 @@ def shift_tokens_right(input_ids: torch.Tensor, pad_token_id: int, decoder_start
     shifted_input_ids[:, 1:] = input_ids[:, :-1].clone()
     shifted_input_ids[:, 0] = decoder_start_token_id
 
-    assert pad_token_id is not None, "self.model.config.pad_token_id has to be defined."
+    if pad_token_id is None:
+        raise ValueError("self.model.config.pad_token_id has to be defined.")
     # replace possible -100 values in labels by `pad_token_id`
     shifted_input_ids.masked_fill_(shifted_input_ids == -100, pad_token_id)
 

src/transformers/models/marian/modeling_marian.py

@@ -67,7 +67,8 @@ def shift_tokens_right(input_ids: torch.Tensor, pad_token_id: int, decoder_start
     shifted_input_ids[:, 1:] = input_ids[:, :-1].clone()
     shifted_input_ids[:, 0] = decoder_start_token_id
 
-    assert pad_token_id is not None, "self.model.config.pad_token_id has to be defined."
+    if pad_token_id is None:
+        raise ValueError("self.model.config.pad_token_id has to be defined.")
     # replace possible -100 values in labels by `pad_token_id`
     shifted_input_ids.masked_fill_(shifted_input_ids == -100, pad_token_id)
 

src/transformers/models/pegasus/modeling_pegasus.py

@@ -66,7 +66,8 @@ def shift_tokens_right(input_ids: torch.Tensor, pad_token_id: int, decoder_start
     shifted_input_ids[:, 1:] = input_ids[:, :-1].clone()
     shifted_input_ids[:, 0] = decoder_start_token_id
 
-    assert pad_token_id is not None, "self.model.config.pad_token_id has to be defined."
+    if pad_token_id is None:
+        raise ValueError("self.model.config.pad_token_id has to be defined.")
     # replace possible -100 values in labels by `pad_token_id`
     shifted_input_ids.masked_fill_(shifted_input_ids == -100, pad_token_id)
 

src/transformers/models/speech_to_text/modeling_speech_to_text.py

@@ -63,7 +63,8 @@ def shift_tokens_right(input_ids: torch.Tensor, pad_token_id: int, decoder_start
     shifted_input_ids[:, 1:] = input_ids[:, :-1].clone()
     shifted_input_ids[:, 0] = decoder_start_token_id
 
-    assert pad_token_id is not None, "self.model.config.pad_token_id has to be defined."
+    if pad_token_id is None:
+        raise ValueError("self.model.config.pad_token_id has to be defined.")
     # replace possible -100 values in labels by `pad_token_id`
     shifted_input_ids.masked_fill_(shifted_input_ids == -100, pad_token_id)