Fix doc formatting in forward passes & modular

examples/modular-transformers/modeling_new_task_model.py

@@ -349,7 +349,6 @@ def forward(
         num_logits_to_keep: int = 0,
     ) -> Union[Tuple, NewTaskModelCausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.text_config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/aria/modeling_aria.py

@@ -1193,7 +1193,6 @@ def forward(
         **kwargs: Unpack[KwargsForCausalLM],
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored
@@ -1458,7 +1457,6 @@ def forward(
         **loss_kwargs,
     ) -> Union[Tuple, AriaCausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or `model.image_token_id` (where `model` is your instance of `Idefics3ForConditionalGeneration`).

src/transformers/models/aria/modular_aria.py

@@ -1437,7 +1437,6 @@ def forward(
         **loss_kwargs,
     ) -> Union[Tuple, AriaCausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or `model.image_token_id` (where `model` is your instance of `Idefics3ForConditionalGeneration`).

src/transformers/models/bamba/modeling_bamba.py

@@ -1495,7 +1495,6 @@ def forward(
         **kwargs,
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/bamba/modular_bamba.py

@@ -1205,7 +1205,6 @@ def forward(
         **kwargs,
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/chameleon/modeling_chameleon.py

@@ -1554,7 +1554,6 @@ def forward(
         cache_position: Optional[torch.LongTensor] = None,
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/cohere/modeling_cohere.py

@@ -833,7 +833,6 @@ def forward(
         **kwargs: Unpack[KwargsForCausalLM],
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/cohere/modular_cohere.py

@@ -321,7 +321,6 @@ def forward(
         **kwargs: Unpack[KwargsForCausalLM],
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/cohere2/modeling_cohere2.py

@@ -834,7 +834,6 @@ def forward(
         **kwargs: Unpack[KwargsForCausalLM],
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/dbrx/modeling_dbrx.py

@@ -1283,9 +1283,7 @@ def forward(
         logits_to_keep: Union[int, torch.Tensor] = 0,
         **kwargs,
     ) -> Union[Tuple, MoeCausalLMOutputWithPast]:
-        r"""Forward function for causal language modeling.
-
-        Args:
+        r"""
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/deprecated/open_llama/modeling_open_llama.py

@@ -716,7 +716,6 @@ def forward(
         return_dict: Optional[bool] = None,
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/diffllama/modeling_diffllama.py

@@ -1070,7 +1070,6 @@ def forward(
         **kwargs: Unpack[KwargsForCausalLM],
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/emu3/modeling_emu3.py

@@ -1650,7 +1650,6 @@ def forward(
         **kwargs: Unpack[KwargsForCausalLM],
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored
@@ -1878,7 +1877,6 @@ def forward(
         logits_to_keep: Union[int, torch.Tensor] = 0,
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/emu3/modular_emu3.py

@@ -1077,7 +1077,6 @@ def __init__(self, config):
     @replace_return_docstrings(output_type=CausalLMOutputWithPast, config_class="Emu3TextConfig")
     def forward(**super_kwargs):
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored
@@ -1186,7 +1185,6 @@ def forward(
         logits_to_keep: Union[int, torch.Tensor] = 0,
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/gemma/modeling_gemma.py

@@ -803,7 +803,6 @@ def forward(
         **kwargs: Unpack[KwargsForCausalLM],
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/gemma/modular_gemma.py

@@ -483,7 +483,6 @@ def forward(
 class GemmaForCausalLM(LlamaForCausalLM):
     def forward(**super_kwargs):
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/gemma2/modeling_gemma2.py

@@ -841,7 +841,6 @@ def forward(
         **loss_kwargs,
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored
@@ -859,9 +858,9 @@ def forward(
         Example:
 
         ```python
-        >>> from transformers import AutoTokenizer, GemmaForCausalLM
+        >>> from transformers import AutoTokenizer, Gemma2ForCausalLM
 
-        >>> model = GemmaForCausalLM.from_pretrained("google/gemma-2-9b")
+        >>> model = Gemma2ForCausalLM.from_pretrained("google/gemma-2-9b")
         >>> tokenizer = AutoTokenizer.from_pretrained("google/gemma-2-9b")
 
         >>> prompt = "What is your favorite condiment?"

src/transformers/models/gemma2/modular_gemma2.py

@@ -591,10 +591,26 @@ def forward(
         **loss_kwargs,
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         r"""
+            labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
+                Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
+                config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored
+                (masked), the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`.
+
+            logits_to_keep (`int` or `torch.Tensor`, *optional*):
+                If an `int`, compute logits for the last `logits_to_keep` tokens. If `0`, calculate logits for all
+                `input_ids` (special case). Only last token logits are needed for generation, and calculating them only for that
+                token can save memory, which becomes pretty significant for long sequences or large vocabulary size.
+                If a `torch.Tensor`, must be 1D corresponding to the indices to keep in the sequence length dimension.
+                This is useful when using packed tensor format (single dimension for batch and sequence length).
+
+        Returns:
+
+        Example:
+
         ```python
-        >>> from transformers import AutoTokenizer, GemmaForCausalLM
+        >>> from transformers import AutoTokenizer, Gemma2ForCausalLM
 
-        >>> model = GemmaForCausalLM.from_pretrained("google/gemma-2-9b")
+        >>> model = Gemma2ForCausalLM.from_pretrained("google/gemma-2-9b")
         >>> tokenizer = AutoTokenizer.from_pretrained("google/gemma-2-9b")
 
         >>> prompt = "What is your favorite condiment?"

src/transformers/models/glm/modeling_glm.py

@@ -812,7 +812,6 @@ def forward(
         **kwargs: Unpack[KwargsForCausalLM],
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/got_ocr2/modeling_got_ocr2.py

@@ -769,7 +769,6 @@ def forward(
         logits_to_keep: Union[int, torch.Tensor] = 0,
     ) -> Union[Tuple, GotOcr2CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/got_ocr2/modular_got_ocr2.py

@@ -848,7 +848,6 @@ def forward(
         logits_to_keep: Union[int, torch.Tensor] = 0,
     ) -> Union[Tuple, LlavaCausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/granite/modeling_granite.py

@@ -815,7 +815,6 @@ def forward(
         **kwargs: Unpack[KwargsForCausalLM],
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/granitemoe/modeling_granitemoe.py

@@ -1287,7 +1287,6 @@ def forward(
         **kwargs,
     ) -> Union[Tuple, MoeCausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/granitemoeshared/modeling_granitemoeshared.py

@@ -1313,7 +1313,6 @@ def forward(
         **kwargs,
     ) -> Union[Tuple, MoeCausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/helium/modeling_helium.py

@@ -799,7 +799,6 @@ def forward(
         **kwargs: Unpack[KwargsForCausalLM],
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/idefics/modeling_idefics.py

@@ -1559,7 +1559,6 @@ def forward(
         cache_position: Optional[torch.LongTensor] = None,
     ) -> Union[Tuple, IdeficsCausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/idefics/modeling_tf_idefics.py

@@ -1687,7 +1687,6 @@ def call(
         training=False,
     ) -> Union[TFIdeficsCausalLMOutputWithPast, Tuple[tf.Tensor]]:
         r"""
-        Args:
             labels (`tf.Tensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/idefics2/modeling_idefics2.py

@@ -1537,7 +1537,6 @@ def forward(
         logits_to_keep: Union[int, torch.Tensor] = 0,
     ) -> Union[Tuple, Idefics2CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or `model.image_token_id` (where `model` is your instance of `Idefics2ForConditionalGeneration`).

src/transformers/models/idefics3/modeling_idefics3.py

@@ -1121,7 +1121,6 @@ def forward(
         logits_to_keep: Union[int, torch.Tensor] = 0,
     ) -> Union[Tuple, Idefics3CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or `model.image_token_id` (where `model` is your instance of `Idefics3ForConditionalGeneration`).

src/transformers/models/jamba/modeling_jamba.py

@@ -1456,7 +1456,6 @@ def forward(
         **loss_kwargs,
     ) -> Union[Tuple, MoeCausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/jetmoe/modeling_jetmoe.py

@@ -1299,7 +1299,6 @@ def forward(
         **kwargs,
     ) -> Union[Tuple, MoeCausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/llama/modeling_llama.py

@@ -801,7 +801,6 @@ def forward(
         **kwargs: Unpack[KwargsForCausalLM],
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/llava/modeling_llava.py

@@ -348,7 +348,6 @@ def forward(
         **lm_kwargs,
     ) -> Union[Tuple, LlavaCausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/llava_next/modeling_llava_next.py

@@ -561,7 +561,6 @@ def forward(
         **lm_kwargs,
     ) -> Union[Tuple, LlavaNextCausalLMOutputWithPast]:
         r"""
-        Args:
             labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                 Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                 config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored

src/transformers/models/llava_next_video/modeling_llava_next_video.py

@@ -601,7 +601,6 @@ def forward(
         **lm_kwargs,
     ) -> Union[Tuple, LlavaNextVideoCausalLMOutputWithPast]:
         r"""
-        Args:
             pixel_values_videos (`torch.FloatTensor` of shape `(batch_size, num_frames, num_channels, image_size, image_size)):
                 The tensors corresponding to the input videos. Pixel values can be obtained using
                 [`AutoImageProcessor`]. See [`LlavaNextVideoVideoProcessor.__call__`] for details. [`LlavaProcessor`] uses