fix (#4976)

src/transformers/modeling_albert.py

@@ -423,7 +423,7 @@ ALBERT_INPUTS_DOCSTRING = r"""
             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.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_bart.py

@@ -90,7 +90,7 @@ BART_INPUTS_DOCSTRING = r"""
             Default behavior: generate a tensor that ignores pad tokens in decoder_input_ids. Causal mask will also be used by default.
             If you want to change padding behavior, you should read :func:`~transformers.modeling_bart._prepare_decoder_inputs` and modify.
             See diagram 1 in the paper for more info on the default strategy
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_bert.py

@@ -594,7 +594,7 @@ BERT_INPUTS_DOCSTRING = r"""
             is used in the cross-attention if the model is configured as a decoder.
             Mask values selected in ``[0, 1]``:
             ``1`` for tokens that are NOT MASKED, ``0`` for MASKED tokens.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_camembert.py

@@ -49,7 +49,7 @@ CAMEMBERT_START_DOCSTRING = r"""
             model. Initializing with a config file does not load the weights associated with the model, only the
             configuration.
             Check out the :meth:`~transformers.PreTrainedModel.from_pretrained` method to load the model weights.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_ctrl.py

@@ -284,7 +284,7 @@ CTRL_INPUTS_DOCSTRING = r"""
         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`.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_distilbert.py

@@ -377,7 +377,7 @@ DISTILBERT_INPUTS_DOCSTRING = r"""
             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.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_electra.py

@@ -220,7 +220,7 @@ ELECTRA_INPUTS_DOCSTRING = r"""
             is used in the cross-attention if the model is configured as a decoder.
             Mask values selected in ``[0, 1]``:
             ``1`` for tokens that are NOT MASKED, ``0`` for MASKED tokens.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_flaubert.py

@@ -100,7 +100,7 @@ FLAUBERT_INPUTS_DOCSTRING = r"""
             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.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_gpt2.py

@@ -335,7 +335,7 @@ GPT2_INPUTS_DOCSTRING = r"""
             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`.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_longformer.py

@@ -487,7 +487,7 @@ LONGFORMER_INPUTS_DOCSTRING = r"""
             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.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_mmbt.py

@@ -141,7 +141,7 @@ MMBT_INPUTS_DOCSTRING = r"""    Inputs:
             is used in the cross-attention if the model is configured as a decoder.
             Mask values selected in ``[0, 1]``:
             ``1`` for tokens that are NOT MASKED, ``0`` for MASKED tokens.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_openai.py

@@ -322,9 +322,7 @@ OPENAI_GPT_INPUTS_DOCSTRING = r"""
             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.
-        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`False`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
-            Should the model returns attentions weights.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 
@@ -607,7 +605,7 @@ class OpenAIGPTDoubleHeadsModel(OpenAIGPTPreTrainedModel):
         r"""
         mc_token_ids (:obj:`torch.LongTensor` of shape :obj:`(batch_size, num_choices)`, `optional`, default to index of the last token of the input)
             Index of the classification token in each input sequence.
-            Selected in the range ``[0, input_ids.size(-1) - 1[``.
+            Selected in the range ``[0, input_ids.size(-1) - 1]``.
         labels (:obj:`torch.LongTensor` of shape :obj:`(batch_size, sequence_length)`, `optional`, defaults to :obj:`None`)
             Labels for language modeling.
             Note that the labels **are shifted** inside the model, i.e. you can set ``labels = input_ids``

src/transformers/modeling_reformer.py

@@ -1505,7 +1505,7 @@ REFORMER_INPUTS_DOCSTRING = r"""
             bucketing. Setting `num_hashes` overwrites the default `num_hashes` defined
             in `config.num_hashes`.
             For more information, see `num_hashes` in :class:`transformers.ReformerConfig`.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_roberta.py

@@ -130,7 +130,7 @@ ROBERTA_INPUTS_DOCSTRING = r"""
             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.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_t5.py

@@ -841,7 +841,7 @@ T5_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**.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_tf_albert.py

@@ -688,7 +688,7 @@ ALBERT_INPUTS_DOCSTRING = r"""
         training (:obj:`boolean`, `optional`, defaults to :obj:`False`):
             Whether to activate dropout modules (if set to :obj:`True`) during training or to de-activate them
             (if set to :obj:`False`) for evaluation.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_tf_bert.py

@@ -682,7 +682,7 @@ BERT_INPUTS_DOCSTRING = r"""
         training (:obj:`boolean`, `optional`, defaults to :obj:`False`):
             Whether to activate dropout modules (if set to :obj:`True`) during training or to de-activate them
             (if set to :obj:`False`) for evaluation.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_tf_camembert.py

@@ -62,7 +62,7 @@ CAMEMBERT_START_DOCSTRING = r"""
         config (:class:`~transformers.CamembertConfig`): Model configuration class with all the parameters of the
             model. Initializing with a config file does not load the weights associated with the model, only the configuration.
             Check out the :meth:`~transformers.PreTrainedModel.from_pretrained` method to load the model weights.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_tf_ctrl.py

@@ -464,7 +464,7 @@ CTRL_INPUTS_DOCSTRING = r"""
         training (:obj:`boolean`, `optional`, defaults to :obj:`False`):
             Whether to activate dropout modules (if set to :obj:`True`) during training or to de-activate them
             (if set to :obj:`False`) for evaluation.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_tf_distilbert.py

@@ -549,7 +549,7 @@ DISTILBERT_INPUTS_DOCSTRING = r"""
             Whether to activate dropout modules (if set to :obj:`True`) during training or to de-activate them
             (if set to :obj:`False`) for evaluation.
 
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_tf_electra.py

@@ -348,7 +348,7 @@ ELECTRA_INPUTS_DOCSTRING = r"""
             Whether to activate dropout modules (if set to :obj:`True`) during training or to de-activate them
             (if set to :obj:`False`) for evaluation.
 
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_tf_flaubert.py

@@ -100,7 +100,7 @@ FLAUBERT_INPUTS_DOCSTRING = r"""
             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.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_tf_gpt2.py

@@ -467,7 +467,7 @@ GPT2_INPUTS_DOCSTRING = r"""
         training (:obj:`boolean`, `optional`, defaults to :obj:`False`):
             Whether to activate dropout modules (if set to :obj:`True`) during training or to de-activate them
             (if set to :obj:`False`) for evaluation.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_tf_openai.py

@@ -429,7 +429,7 @@ OPENAI_GPT_INPUTS_DOCSTRING = r"""
         training (:obj:`boolean`, `optional`, defaults to :obj:`False`):
             Whether to activate dropout modules (if set to :obj:`True`) during training or to de-activate them
             (if set to :obj:`False`) for evaluation.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 
@@ -568,7 +568,7 @@ class TFOpenAIGPTDoubleHeadsModel(TFOpenAIGPTPreTrainedModel):
         r"""
         mc_token_ids (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(batch_size, num_choices)`, `optional`, default to index of the last token of the input)
             Index of the classification token in each input sequence.
-            Selected in the range ``[0, input_ids.size(-1) - 1[``.
+            Selected in the range ``[0, input_ids.size(-1) - 1]``.
 
     Return:
         :obj:`tuple(tf.Tensor)` comprising various elements depending on the configuration (:class:`~transformers.OpenAIGPTConfig`) and inputs:

src/transformers/modeling_tf_roberta.py

@@ -182,7 +182,7 @@ ROBERTA_INPUTS_DOCSTRING = r"""
         training (:obj:`boolean`, `optional`, defaults to :obj:`False`):
             Whether to activate dropout modules (if set to :obj:`True`) during training or to de-activate them
             (if set to :obj:`False`) for evaluation.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_tf_t5.py

@@ -855,7 +855,7 @@ T5_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**.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_tf_transfo_xl.py

@@ -692,7 +692,7 @@ TRANSFO_XL_INPUTS_DOCSTRING = r"""
             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.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_tf_xlm.py

@@ -585,7 +585,7 @@ XLM_INPUTS_DOCSTRING = r"""
             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.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_tf_xlm_roberta.py

@@ -62,7 +62,7 @@ XLM_ROBERTA_START_DOCSTRING = r"""
         config (:class:`~transformers.XLMRobertaConfig`): Model configuration class with all the parameters of the
             model. Initializing with a config file does not load the weights associated with the model, only the configuration.
             Check out the :meth:`~transformers.PreTrainedModel.from_pretrained` method to load the model weights.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_tf_xlnet.py

@@ -809,7 +809,7 @@ XLNET_INPUTS_DOCSTRING = r"""
             than the model's internal embedding lookup matrix.
         use_cache (:obj:`bool`):
             If `use_cache` is True, `mems` are returned and can be used to speed up decoding (see `mems`). Defaults to `True`.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_transfo_xl.py

@@ -622,7 +622,7 @@ TRANSFO_XL_INPUTS_DOCSTRING = r"""
             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.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_xlm.py

@@ -302,7 +302,7 @@ XLM_INPUTS_DOCSTRING = r"""
             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.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_xlm_roberta.py

@@ -53,7 +53,7 @@ XLM_ROBERTA_START_DOCSTRING = r"""
         config (:class:`~transformers.XLMRobertaConfig`): Model configuration class with all the parameters of the
             model. Initializing with a config file does not load the weights associated with the model, only the configuration.
             Check out the :meth:`~transformers.PreTrainedModel.from_pretrained` method to load the model weights.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """
 

src/transformers/modeling_xlnet.py

@@ -618,7 +618,7 @@ XLNET_INPUTS_DOCSTRING = r"""
             than the model's internal embedding lookup matrix.
         use_cache (:obj:`bool`):
             If `use_cache` is True, `mems` are returned and can be used to speed up decoding (see `mems`). Defaults to `True`.
-        output_attentions (:obj:`bool`, `optional`, defaults to `:obj:`None`):
+        output_attentions (:obj:`bool`, `optional`, defaults to :obj:`None`):
             If set to ``True``, the attentions tensors of all attention layers are returned. See ``attentions`` under returned tensors for more detail.
 """