Fix annotations (#24582)

* fix annotations

* fix annotations

* fix annotations

* fix annotations

* fix annotations

* fix annotations

* fix annotations

* fix annotations

* fix annotations

* fix annotations

* fix annotations

* fix annotations

* fix annotations

* fix annotations

* fix annotations

* fix annotations

* fix annotations

* fix annotations

* fix annotations

* fix annotations

* fix annotations

* fix annotations

* fix annotations

src/transformers/models/table_transformer/modeling_table_transformer.py

@@ -584,7 +584,7 @@ class TableTransformerEncoderLayer(nn.Module):
     ):
         """
         Args:
-            hidden_states (`torch.FloatTensor`): input to the layer of shape `(seq_len, batch, embed_dim)`
+            hidden_states (`torch.FloatTensor`): input to the layer of shape `(batch, seq_len, embed_dim)`
             attention_mask (`torch.FloatTensor`): attention mask of size
                 `(batch, 1, target_len, source_len)` where padding elements are indicated by very large negative
                 values.
@@ -668,7 +668,7 @@ class TableTransformerDecoderLayer(nn.Module):
     ):
         """
         Args:
-            hidden_states (`torch.FloatTensor`): input to the layer of shape `(seq_len, batch, embed_dim)`
+            hidden_states (`torch.FloatTensor`): input to the layer of shape `(batch, seq_len, embed_dim)`
             attention_mask (`torch.FloatTensor`): attention mask of size
                 `(batch, 1, target_len, source_len)` where padding elements are indicated by very large negative
                 values.
@@ -679,7 +679,7 @@ class TableTransformerDecoderLayer(nn.Module):
                 position embeddings that are added to the queries and keys
             in the self-attention layer.
             encoder_hidden_states (`torch.FloatTensor`):
-                cross attention input to the layer of shape `(seq_len, batch, embed_dim)`
+                cross attention input to the layer of shape `(batch, seq_len, embed_dim)`
             encoder_attention_mask (`torch.FloatTensor`): encoder attention mask of size
                 `(batch, 1, target_len, source_len)` where padding elements are indicated by very large negative
                 values.

src/transformers/models/trocr/modeling_trocr.py

@@ -358,11 +358,11 @@ class TrOCRDecoderLayer(nn.Module):
     ):
         """
         Args:
-            hidden_states (`torch.FloatTensor`): input to the layer of shape `(seq_len, batch, embed_dim)`
+            hidden_states (`torch.FloatTensor`): input to the layer of shape `(batch, seq_len, embed_dim)`
             attention_mask (`torch.FloatTensor`): attention mask of size
                 `(batch, 1, tgt_len, src_len)` where padding elements are indicated by very large negative values.
             encoder_hidden_states (`torch.FloatTensor`):
-                cross attention input to the layer of shape `(seq_len, batch, embed_dim)`
+                cross attention input to the layer of shape `(batch, seq_len, embed_dim)`
             encoder_attention_mask (`torch.FloatTensor`): encoder attention mask of size
                 `(batch, 1, tgt_len, src_len)` where padding elements are indicated by very large negative values.
             layer_head_mask (`torch.FloatTensor`): mask for attention heads in a given layer of size

src/transformers/models/whisper/modeling_tf_whisper.py

@@ -313,7 +313,7 @@ class TFWhisperEncoderLayer(tf.keras.layers.Layer):
     ):
         """
         Args:
-            hidden_states (`tf.Tensor`): input to the layer of shape `(seq_len, batch, embed_dim)`
+            hidden_states (`tf.Tensor`): input to the layer of shape `(batch, seq_len, embed_dim)`
             attention_mask (`tf.Tensor`): attention mask of size
                 `(batch, 1, tgt_len, src_len)` where padding elements are indicated by very large negative values.
             layer_head_mask (`tf.Tensor`): mask for attention heads in a given layer of size
@@ -391,11 +391,11 @@ class TFWhisperDecoderLayer(tf.keras.layers.Layer):
     ) -> Tuple[tf.Tensor, tf.Tensor, Tuple[Tuple[tf.Tensor]]]:
         """
         Args:
-            hidden_states (`tf.Tensor`): input to the layer of shape `(seq_len, batch, embed_dim)`
+            hidden_states (`tf.Tensor`): input to the layer of shape `(batch, seq_len, embed_dim)`
             attention_mask (`tf.Tensor`): attention mask of size
                 `(batch, 1, tgt_len, src_len)` where padding elements are indicated by very large negative values.
             encoder_hidden_states (`tf.Tensor`):
-                cross attention input to the layer of shape `(seq_len, batch, embed_dim)`
+                cross attention input to the layer of shape `(batch, seq_len, embed_dim)`
             encoder_attention_mask (`tf.Tensor`): encoder attention mask of size
                 `(batch, 1, tgt_len, src_len)` where padding elements are indicated by very large negative values.
             layer_head_mask (`tf.Tensor`): mask for attention heads in a given layer of size

src/transformers/models/xglm/modeling_tf_xglm.py

@@ -348,11 +348,11 @@ class TFXGLMDecoderLayer(tf.keras.layers.Layer):
     ) -> Tuple[tf.Tensor, tf.Tensor, Tuple[Tuple[tf.Tensor]]]:
         """
         Args:
-            hidden_states (`tf.Tensor`): input to the layer of shape *(seq_len, batch, embed_dim)*
+            hidden_states (`tf.Tensor`): input to the layer of shape *(batch, seq_len, embed_dim)*
             attention_mask (`tf.Tensor`): attention mask of size
                 *(batch, 1, tgt_len, src_len)* where padding elements are indicated by very large negative values.
             encoder_hidden_states (`tf.Tensor`):
-                cross attention input to the layer of shape *(seq_len, batch, embed_dim)*
+                cross attention input to the layer of shape *(batch, seq_len, embed_dim)*
             encoder_attention_mask (`tf.Tensor`): encoder attention mask of size
                 *(batch, 1, tgt_len, src_len)* where padding elements are indicated by very large negative values.
             layer_head_mask (`tf.Tensor`): mask for attention heads in a given layer of size

templates/adding_a_new_model/cookiecutter-template-{{cookiecutter.modelname}}/modeling_tf_{{cookiecutter.lowercase_modelname}}.py

@@ -1794,7 +1794,7 @@ class TF{{cookiecutter.camelcase_modelname}}EncoderLayer(tf.keras.layers.Layer):
     def call(self, hidden_states: tf.Tensor, attention_mask: tf.Tensor, layer_head_mask: tf.Tensor, training=False):
         """
         Args:
-            hidden_states (`tf.Tensor`): input to the layer of shape *(seq_len, batch, embed_dim)*
+            hidden_states (`tf.Tensor`): input to the layer of shape *(batch, seq_len, embed_dim)*
             attention_mask (`tf.Tensor`): attention mask of size
                 *(batch, 1, tgt_len, src_len)* where padding elements are indicated by very large negative values.
             layer_head_mask (`tf.Tensor`): mask for attention heads in a given layer of size
@@ -1867,10 +1867,10 @@ class TF{{cookiecutter.camelcase_modelname}}DecoderLayer(tf.keras.layers.Layer):
     ) -> Tuple[tf.Tensor, tf.Tensor, Tuple[Tuple[tf.Tensor]]]:
         """
         Args:
-            hidden_states (`tf.Tensor`): input to the layer of shape *(seq_len, batch, embed_dim)*
+            hidden_states (`tf.Tensor`): input to the layer of shape *(batch, seq_len, embed_dim)*
             attention_mask (`tf.Tensor`): attention mask of size
                 *(batch, 1, tgt_len, src_len)* where padding elements are indicated by very large negative values.
-            encoder_hidden_states (`tf.Tensor`): cross attention input to the layer of shape *(seq_len, batch, embed_dim)*
+            encoder_hidden_states (`tf.Tensor`): cross attention input to the layer of shape *(batch, seq_len, embed_dim)*
             encoder_attention_mask (`tf.Tensor`): encoder attention mask of size
                 *(batch, 1, tgt_len, src_len)* where padding elements are indicated by very large negative values.
             layer_head_mask (`tf.Tensor`): mask for attention heads in a given layer of size

templates/adding_a_new_model/cookiecutter-template-{{cookiecutter.modelname}}/modeling_{{cookiecutter.lowercase_modelname}}.py

@@ -1826,7 +1826,7 @@ class {{cookiecutter.camelcase_modelname}}EncoderLayer(nn.Module):
     ):
         """
         Args:
-            hidden_states (`torch.FloatTensor`): input to the layer of shape *(seq_len, batch, embed_dim)*
+            hidden_states (`torch.FloatTensor`): input to the layer of shape *(batch, seq_len, embed_dim)*
             attention_mask (`torch.FloatTensor`): attention mask of size
                 *(batch, 1, tgt_len, src_len)* where padding elements are indicated by very large negative values.
             layer_head_mask (`torch.FloatTensor`): mask for attention heads in a given layer of size
@@ -1907,10 +1907,10 @@ class {{cookiecutter.camelcase_modelname}}DecoderLayer(nn.Module):
     ):
         """
         Args:
-            hidden_states (`torch.FloatTensor`): input to the layer of shape *(seq_len, batch, embed_dim)*
+            hidden_states (`torch.FloatTensor`): input to the layer of shape *(batch, seq_len, embed_dim)*
             attention_mask (`torch.FloatTensor`): attention mask of size
                 *(batch, 1, tgt_len, src_len)* where padding elements are indicated by very large negative values.
-            encoder_hidden_states (`torch.FloatTensor`): cross attention input to the layer of shape *(seq_len, batch, embed_dim)*
+            encoder_hidden_states (`torch.FloatTensor`): cross attention input to the layer of shape *(batch, seq_len, embed_dim)*
             encoder_attention_mask (`torch.FloatTensor`): encoder attention mask of size
                 *(batch, 1, tgt_len, src_len)* where padding elements are indicated by very large negative values.
             layer_head_mask (`torch.FloatTensor`): mask for attention heads in a given layer of size