make code check pass

huggingface · Oct 19, 2024 · 8b48846 · 8b48846
1 parent 4fc5d23
commit 8b48846
Show file tree

Hide file tree

Showing 7 changed files with 37 additions and 26 deletions.
diff --git a/src/transformers/generation/candidate_generator.py b/src/transformers/generation/candidate_generator.py
@@ -459,9 +459,11 @@ def get_candidates(self, input_ids: torch.LongTensor) -> Tuple[torch.LongTensor,
             prompt_use_length = new_assistant_ids.shape[1]
             prompt_use = self.prev_assistant_ids[:, -prompt_use_length:]
 
-            discrepancy_length, new_tokens_only, discrepancy_only = (
-                AssistedCandidateGeneratorDifferentTokenizers._get_tokens_diag(prompt_use, new_assistant_ids)
-            )
+            (
+                discrepancy_length,
+                new_tokens_only,
+                discrepancy_only,
+            ) = AssistedCandidateGeneratorDifferentTokenizers._get_tokens_diag(prompt_use, new_assistant_ids)
             assistant_input_ids = self.prev_assistant_ids
 
             if new_tokens_only is not None:

diff --git a/src/transformers/models/big_bird/modeling_big_bird.py b/src/transformers/models/big_bird/modeling_big_bird.py
@@ -919,9 +919,11 @@ def bigbird_block_sparse_attention(
             attention_probs[:, :, -2 * from_block_size : -from_block_size, :to_block_size] = second_last_attn_weights[
                 :, :, :, :to_block_size
             ]  # 1st key block (global)
-            attention_probs[:, :, -2 * from_block_size : -from_block_size, -3 * to_block_size :] = (
-                second_last_attn_weights[:, :, :, to_block_size : 4 * to_block_size]
-            )  # last three blocks (global + sliding)
+            attention_probs[
+                :, :, -2 * from_block_size : -from_block_size, -3 * to_block_size :
+            ] = second_last_attn_weights[
+                :, :, :, to_block_size : 4 * to_block_size
+            ]  # last three blocks (global + sliding)
             # random keys
             for p1, i1, w1 in zip(range(bsz), rand_attn, second_last_attn_weights):
                 # p1, i1, w1 corresponds to batch_dim i.e. following operation is done for each sequence in batch

diff --git a/src/transformers/models/bigbird_pegasus/modeling_bigbird_pegasus.py b/src/transformers/models/bigbird_pegasus/modeling_bigbird_pegasus.py
@@ -718,9 +718,11 @@ def bigbird_block_sparse_attention(
             attention_probs[:, :, -2 * from_block_size : -from_block_size, :to_block_size] = second_last_attn_weights[
                 :, :, :, :to_block_size
             ]  # 1st key block (global)
-            attention_probs[:, :, -2 * from_block_size : -from_block_size, -3 * to_block_size :] = (
-                second_last_attn_weights[:, :, :, to_block_size : 4 * to_block_size]
-            )  # last three blocks (global + sliding)
+            attention_probs[
+                :, :, -2 * from_block_size : -from_block_size, -3 * to_block_size :
+            ] = second_last_attn_weights[
+                :, :, :, to_block_size : 4 * to_block_size
+            ]  # last three blocks (global + sliding)
             # random keys
             for p1, i1, w1 in zip(range(bsz), rand_attn, second_last_attn_weights):
                 # p1, i1, w1 corresponds to batch_dim i.e. following operation is done for each sequence in batch

diff --git a/src/transformers/models/moshi/modeling_moshi.py b/src/transformers/models/moshi/modeling_moshi.py
@@ -2216,16 +2216,19 @@ def generate(
         # needs to prepare generation config, even though it'll be done again in `generate`
         generation_config, kwargs = self._prepare_generation_config(kwargs.pop("generation_config", None), **kwargs)
 
-        input_ids, user_audio_codes, moshi_audio_codes, concat_unconditional_inputs = (
-            self._check_and_maybe_initalize_inputs(
-                input_ids=input_ids,
-                user_input_values=user_input_values,
-                user_audio_codes=user_audio_codes,
-                moshi_input_values=moshi_input_values,
-                moshi_audio_codes=moshi_audio_codes,
-                inputs_embeds=inputs_embeds,
-                concat_unconditional_inputs=concat_unconditional_inputs,
-            )
+        (
+            input_ids,
+            user_audio_codes,
+            moshi_audio_codes,
+            concat_unconditional_inputs,
+        ) = self._check_and_maybe_initalize_inputs(
+            input_ids=input_ids,
+            user_input_values=user_input_values,
+            user_audio_codes=user_audio_codes,
+            moshi_input_values=moshi_input_values,
+            moshi_audio_codes=moshi_audio_codes,
+            inputs_embeds=inputs_embeds,
+            concat_unconditional_inputs=concat_unconditional_inputs,
         )
 
         inputs = inputs_embeds if input_ids is None else input_ids

diff --git a/src/transformers/tokenization_utils.py b/src/transformers/tokenization_utils.py
@@ -1035,10 +1035,12 @@ def get_special_tokens_mask(
         return [0] * ((len(token_ids_1) if token_ids_1 else 0) + len(token_ids_0))
 
     @overload
-    def convert_ids_to_tokens(self, ids: int, skip_special_tokens: bool = False) -> str: ...
+    def convert_ids_to_tokens(self, ids: int, skip_special_tokens: bool = False) -> str:
+        ...
 
     @overload
-    def convert_ids_to_tokens(self, ids: List[int], skip_special_tokens: bool = False) -> List[str]: ...
+    def convert_ids_to_tokens(self, ids: List[int], skip_special_tokens: bool = False) -> List[str]:
+        ...
 
     def convert_ids_to_tokens(
         self, ids: Union[int, List[int]], skip_special_tokens: bool = False

diff --git a/tests/models/video_llava/test_modeling_video_llava.py b/tests/models/video_llava/test_modeling_video_llava.py
@@ -164,9 +164,9 @@ def prepare_config_and_inputs_for_common(self):
         input_ids = ids_tensor([self.batch_size, self.seq_length], config.text_config.vocab_size - 1) + 1
         attention_mask = input_ids.ne(1).to(torch_device)
 
-        input_ids[(input_ids == config.image_token_index) | (input_ids == config.video_token_index)] = (
-            self.pad_token_id
-        )
+        input_ids[
+            (input_ids == config.image_token_index) | (input_ids == config.video_token_index)
+        ] = self.pad_token_id
         input_ids[:, : self.num_image_tokens] = config.image_token_index
         input_ids[:, self.num_image_tokens : self.num_video_tokens + self.num_image_tokens] = config.video_token_index
         inputs_dict = {

diff --git a/utils/check_copies.py b/utils/check_copies.py
@@ -757,9 +757,9 @@ def is_copy_consistent(filename: str, overwrite: bool = False, buffer: dict = No
                 else:
                     # not in the target --> add it
                     theoretical_code_blocks[f"_ignored_new_block_{ignored_new_block_index}"] = code
-                    name_mappings_1[f"_ignored_new_block_{ignored_new_block_index}"] = (
+                    name_mappings_1[
                         f"_ignored_new_block_{ignored_new_block_index}"
-                    )
+                    ] = f"_ignored_new_block_{ignored_new_block_index}"
 
                     del observed_code_blocks[name]
                     observed_code_blocks[f"_ignored_new_block_{ignored_new_block_index}"] = code