gpt2 typo (#4629)

* gpt2 typo

* Add files via upload

src/transformers/modeling_albert.py

@@ -425,7 +425,7 @@ ALBERT_INPUTS_DOCSTRING = r"""
             Mask to nullify selected heads of the self-attention modules.
             Mask values selected in ``[0, 1]``:
             :obj:`1` indicates the head is **not masked**, :obj:`0` indicates the head is **masked**.
-        input_embeds (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
+        inputs_embeds (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
             Optionally, instead of passing :obj:`input_ids` you can choose to directly pass an embedded representation.
             This is useful if you want more control over how to convert `input_ids` indices into associated vectors
             than the model's internal embedding lookup matrix.

src/transformers/modeling_ctrl.py

@@ -244,10 +244,10 @@ CTRL_INPUTS_DOCSTRING = r"""
             Mask to nullify selected heads of the self-attention modules.
             Mask values selected in ``[0, 1]``:
             :obj:`1` indicates the head is **not masked**, :obj:`0` indicates the head is **masked**.
-        input_embeds (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
+        inputs_embeds (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
             This is useful if you want more control over how to convert `input_ids` indices into associated vectors
             than the model's internal embedding lookup matrix.
-            If `past` is used, optionally only the last `input_embeds` have to be input (see `past`).
+            If `past` is used, optionally only the last `inputs_embeds` have to be input (see `past`).
         use_cache (:obj:`bool`):
             If `use_cache` is True, `past` key value states are returned and
             can be used to speed up decoding (see `past`). Defaults to `True`.

src/transformers/modeling_flaubert.py

@@ -95,7 +95,7 @@ FLAUBERT_INPUTS_DOCSTRING = r"""
             Mask to nullify selected heads of the self-attention modules.
             Mask values selected in ``[0, 1]``:
             :obj:`1` indicates the head is **not masked**, :obj:`0` indicates the head is **masked**.
-        input_embeds (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
+        inputs_embeds (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
             Optionally, instead of passing :obj:`input_ids` you can choose to directly pass an embedded representation.
             This is useful if you want more control over how to convert `input_ids` indices into associated vectors
             than the model's internal embedding lookup matrix.

src/transformers/modeling_gpt2.py

@@ -323,10 +323,10 @@ GPT2_INPUTS_DOCSTRING = r"""
             Mask to nullify selected heads of the self-attention modules.
             Mask values selected in ``[0, 1]``:
             :obj:`1` indicates the head is **not masked**, :obj:`0` indicates the head is **masked**.
-        input_embeds (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
+        inputs_embeds (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
             This is useful if you want more control over how to convert `input_ids` indices into associated vectors
             than the model's internal embedding lookup matrix.
-            If `past` is used, optionally only the last `input_embeds` have to be input (see `past`).
+            If `past` is used, optionally only the last `inputs_embeds` have to be input (see `past`).
         use_cache (:obj:`bool`):
             If `use_cache` is True, `past` key value states are returned and can be used to speed up decoding (see `past`). Defaults to `True`.
 """

src/transformers/modeling_openai.py

@@ -313,7 +313,7 @@ OPENAI_GPT_INPUTS_DOCSTRING = r"""
             Mask to nullify selected heads of the self-attention modules.
             Mask values selected in ``[0, 1]``:
             :obj:`1` indicates the head is **not masked**, :obj:`0` indicates the head is **masked**.
-        input_embeds (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
+        inputs_embeds (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
             Optionally, instead of passing :obj:`input_ids` you can choose to directly pass an embedded representation.
             This is useful if you want more control over how to convert `input_ids` indices into associated vectors
             than the model's internal embedding lookup matrix.

src/transformers/modeling_reformer.py

@@ -1662,7 +1662,7 @@ class ReformerModel(ReformerPreTrainedModel):
                 padded_position_ids = position_ids.unsqueeze(0).expand(input_shape[0], padding_length)
                 position_ids = torch.cat([position_ids, padded_position_ids], dim=-1)
 
-        # Extend `input_embeds` with padding to match least common multiple chunk_length
+        # Extend `inputs_embeds` with padding to match least common multiple chunk_length
         if inputs_embeds is not None:
             padded_inputs_embeds = self.embeddings(padded_input_ids, position_ids)
             inputs_embeds = torch.cat([inputs_embeds, padded_inputs_embeds], dim=-2)

src/transformers/modeling_tf_albert.py

@@ -657,7 +657,7 @@ ALBERT_INPUTS_DOCSTRING = r"""
             Mask to nullify selected heads of the self-attention modules.
             Mask values selected in ``[0, 1]``:
             ``1`` indicates the head is **not masked**, ``0`` indicates the head is **masked**.
-        input_embeds (:obj:`tf.Tensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
+        inputs_embeds (:obj:`tf.Tensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
             Optionally, instead of passing :obj:`input_ids` you can choose to directly pass an embedded representation.
             This is useful if you want more control over how to convert `input_ids` indices into associated vectors
             than the model's internal embedding lookup matrix.

src/transformers/modeling_tf_ctrl.py

@@ -449,7 +449,7 @@ CTRL_INPUTS_DOCSTRING = r"""
             Mask to nullify selected heads of the self-attention modules.
             Mask values selected in ``[0, 1]``:
             :obj:`1` indicates the head is **not masked**, :obj:`0` indicates the head is **masked**.
-        input_embeds (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
+        inputs_embeds (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
             Optionally, instead of passing :obj:`input_ids` you can choose to directly pass an embedded representation.
             This is useful if you want more control over how to convert `input_ids` indices into associated vectors
             than the model's internal embedding lookup matrix.

src/transformers/modeling_tf_flaubert.py

@@ -91,7 +91,7 @@ FLAUBERT_INPUTS_DOCSTRING = r"""
             Mask to nullify selected heads of the self-attention modules.
             Mask values selected in ``[0, 1]``:
             :obj:`1` indicates the head is **not masked**, :obj:`0` indicates the head is **masked**.
-        input_embeds (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
+        inputs_embeds (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
             Optionally, instead of passing :obj:`input_ids` you can choose to directly pass an embedded representation.
             This is useful if you want more control over how to convert `input_ids` indices into associated vectors
             than the model's internal embedding lookup matrix.

src/transformers/modeling_tf_gpt2.py

@@ -458,7 +458,7 @@ GPT2_INPUTS_DOCSTRING = r"""
             Mask to nullify selected heads of the self-attention modules.
             Mask values selected in ``[0, 1]``:
             :obj:`1` indicates the head is **not masked**, :obj:`0` indicates the head is **masked**.
-        input_embeds (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
+        inputs_embeds (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
             Optionally, instead of passing :obj:`input_ids` you can choose to directly pass an embedded representation.
             This is useful if you want more control over how to convert `input_ids` indices into associated vectors
             than the model's internal embedding lookup matrix.

src/transformers/modeling_tf_openai.py

@@ -411,7 +411,7 @@ OPENAI_GPT_INPUTS_DOCSTRING = r"""
             Mask to nullify selected heads of the self-attention modules.
             Mask values selected in ``[0, 1]``:
             :obj:`1` indicates the head is **not masked**, :obj:`0` indicates the head is **masked**.
-        input_embeds (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
+        inputs_embeds (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
             Optionally, instead of passing :obj:`input_ids` you can choose to directly pass an embedded representation.
             This is useful if you want more control over how to convert `input_ids` indices into associated vectors
             than the model's internal embedding lookup matrix.

src/transformers/modeling_tf_transfo_xl.py

@@ -679,7 +679,7 @@ TRANSFO_XL_INPUTS_DOCSTRING = r"""
             Mask to nullify selected heads of the self-attention modules.
             Mask values selected in ``[0, 1]``:
             :obj:`1` indicates the head is **not masked**, :obj:`0` indicates the head is **masked**.
-        input_embeds (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
+        inputs_embeds (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
             Optionally, instead of passing :obj:`input_ids` you can choose to directly pass an embedded representation.
             This is useful if you want more control over how to convert `input_ids` indices into associated vectors
             than the model's internal embedding lookup matrix.

src/transformers/modeling_tf_xlm.py

@@ -560,7 +560,7 @@ XLM_INPUTS_DOCSTRING = r"""
             Mask to nullify selected heads of the self-attention modules.
             Mask values selected in ``[0, 1]``:
             :obj:`1` indicates the head is **not masked**, :obj:`0` indicates the head is **masked**.
-        input_embeds (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
+        inputs_embeds (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
             Optionally, instead of passing :obj:`input_ids` you can choose to directly pass an embedded representation.
             This is useful if you want more control over how to convert `input_ids` indices into associated vectors
             than the model's internal embedding lookup matrix.

src/transformers/modeling_tf_xlnet.py

@@ -779,7 +779,7 @@ XLNET_INPUTS_DOCSTRING = r"""
             Mask to nullify selected heads of the self-attention modules.
             Mask values selected in ``[0, 1]``:
             :obj:`1` indicates the head is **not masked**, :obj:`0` indicates the head is **masked**.
-        input_embeds (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
+        inputs_embeds (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
             Optionally, instead of passing :obj:`input_ids` you can choose to directly pass an embedded representation.
             This is useful if you want more control over how to convert `input_ids` indices into associated vectors
             than the model's internal embedding lookup matrix.

src/transformers/modeling_transfo_xl.py

@@ -538,7 +538,7 @@ TRANSFO_XL_INPUTS_DOCSTRING = r"""
             Mask to nullify selected heads of the self-attention modules.
             Mask values selected in ``[0, 1]``:
             :obj:`1` indicates the head is **not masked**, :obj:`0` indicates the head is **masked**.
-        input_embeds (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
+        inputs_embeds (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
             Optionally, instead of passing :obj:`input_ids` you can choose to directly pass an embedded representation.
             This is useful if you want more control over how to convert `input_ids` indices into associated vectors
             than the model's internal embedding lookup matrix.

src/transformers/modeling_xlm.py

@@ -299,7 +299,7 @@ XLM_INPUTS_DOCSTRING = r"""
             Mask to nullify selected heads of the self-attention modules.
             Mask values selected in ``[0, 1]``:
             :obj:`1` indicates the head is **not masked**, :obj:`0` indicates the head is **masked**.
-        input_embeds (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
+        inputs_embeds (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
             Optionally, instead of passing :obj:`input_ids` you can choose to directly pass an embedded representation.
             This is useful if you want more control over how to convert `input_ids` indices into associated vectors
             than the model's internal embedding lookup matrix.

src/transformers/modeling_xlnet.py

@@ -552,7 +552,7 @@ XLNET_INPUTS_DOCSTRING = r"""
             Mask to nullify selected heads of the self-attention modules.
             Mask values selected in ``[0, 1]``:
             :obj:`1` indicates the head is **not masked**, :obj:`0` indicates the head is **masked**.
-        input_embeds (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
+        inputs_embeds (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
             Optionally, instead of passing :obj:`input_ids` you can choose to directly pass an embedded representation.
             This is useful if you want more control over how to convert `input_ids` indices into associated vectors
             than the model's internal embedding lookup matrix.