[GPT-J] add deprecation warning (#21869)

* add deprecation warning

* remove pos ids from args docstirng

* fix failing test

src/transformers/models/gptj/modeling_gptj.py

@@ -391,11 +391,6 @@ GPTJ_INPUTS_DOCSTRING = r"""
             - 1 corresponds to a *sentence B* token.
 
             [What are token type IDs?](../glossary#token-type-ids)
-        position_ids (`torch.LongTensor` of shape `({0})`, *optional*):
-            Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0,
-            config.n_positions - 1]`.
-
-            [What are position IDs?](../glossary#position-ids)
         head_mask (`torch.FloatTensor` of shape `(num_attention_heads,)` or `(n_layer, num_attention_heads)`, *optional*):
             Mask to nullify selected heads of the self-attention modules. Mask values selected in `[0, 1]`:
 
@@ -549,14 +544,24 @@ class GPTJModel(GPTJPreTrainedModel):
         past_key_values: Optional[Tuple[Tuple[torch.Tensor]]] = None,
         attention_mask: Optional[torch.FloatTensor] = None,
         token_type_ids: Optional[torch.LongTensor] = None,
-        position_ids: Optional[torch.LongTensor] = None,
         head_mask: Optional[torch.FloatTensor] = None,
         inputs_embeds: Optional[torch.FloatTensor] = None,
         use_cache: Optional[bool] = None,
         output_attentions: Optional[bool] = None,
         output_hidden_states: Optional[bool] = None,
         return_dict: Optional[bool] = None,
+        **deprecated_arguments,
     ) -> Union[Tuple, BaseModelOutputWithPast]:
+        if deprecated_arguments.pop("position_ids", False) is not False:
+            # `position_ids` could have been `torch.Tensor` or `None` so defaulting pop to `False` allows to detect if users were passing explicitly `None`
+            warnings.warn(
+                "`position_ids` have no functionality in GPT-J and will be removed in v5.0.0. You can safely ignore"
+                " passing `position_ids`.",
+                FutureWarning,
+            )
+        if len(deprecated_arguments) > 0:
+            raise ValueError(f"Got unexpected arguments: {deprecated_arguments}")
+
         output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
         output_hidden_states = (
             output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
@@ -576,23 +581,11 @@ class GPTJModel(GPTJPreTrainedModel):
         else:
             raise ValueError("You have to specify either input_ids or inputs_embeds")
 
-        device = input_ids.device if input_ids is not None else inputs_embeds.device
-
         if token_type_ids is not None:
             token_type_ids = token_type_ids.view(-1, input_shape[-1])
 
-        if position_ids is not None:
-            position_ids = position_ids.view(-1, input_shape[-1])
-
         if past_key_values is None:
-            past_length = 0
             past_key_values = tuple([None] * len(self.h))
-        else:
-            past_length = past_key_values[0][0].size(-2)
-
-        if position_ids is None:
-            position_ids = torch.arange(past_length, input_shape[-1] + past_length, dtype=torch.long, device=device)
-            position_ids = position_ids.unsqueeze(0).view(-1, input_shape[-1])
 
         # Attention mask.
         if attention_mask is not None:
@@ -773,7 +766,9 @@ class GPTJForCausalLM(GPTJPreTrainedModel):
     def set_output_embeddings(self, new_embeddings):
         self.lm_head = new_embeddings
 
-    def prepare_inputs_for_generation(self, input_ids, past_key_values=None, inputs_embeds=None, **kwargs):
+    def prepare_inputs_for_generation(
+        self, input_ids, past_key_values=None, inputs_embeds=None, use_cache=None, **kwargs
+    ):
         token_type_ids = kwargs.get("token_type_ids", None)
         # only last token for inputs_ids if past is defined in kwargs
         if past_key_values:
@@ -782,16 +777,6 @@ class GPTJForCausalLM(GPTJPreTrainedModel):
                 token_type_ids = token_type_ids[:, -1].unsqueeze(-1)
 
         attention_mask = kwargs.get("attention_mask", None)
-        position_ids = kwargs.get("position_ids", None)
-
-        if attention_mask is not None and position_ids is None:
-            # create position_ids on the fly for batch generation
-            position_ids = attention_mask.long().cumsum(-1) - 1
-            position_ids.masked_fill_(attention_mask == 0, 1)
-            if past_key_values:
-                position_ids = position_ids[:, -1].unsqueeze(-1)
-        else:
-            position_ids = None
 
         # if `inputs_embeds` are passed, we only want to use them in the 1st generation step
         if inputs_embeds is not None and past_key_values is None:
@@ -802,8 +787,7 @@ class GPTJForCausalLM(GPTJPreTrainedModel):
         model_inputs.update(
             {
                 "past_key_values": past_key_values,
-                "use_cache": kwargs.get("use_cache"),
-                "position_ids": position_ids,
+                "use_cache": use_cache,
                 "attention_mask": attention_mask,
                 "token_type_ids": token_type_ids,
             }
@@ -824,7 +808,6 @@ class GPTJForCausalLM(GPTJPreTrainedModel):
         past_key_values: Optional[Tuple[Tuple[torch.Tensor]]] = None,
         attention_mask: Optional[torch.FloatTensor] = None,
         token_type_ids: Optional[torch.LongTensor] = None,
-        position_ids: Optional[torch.LongTensor] = None,
         head_mask: Optional[torch.FloatTensor] = None,
         inputs_embeds: Optional[torch.FloatTensor] = None,
         labels: Optional[torch.LongTensor] = None,
@@ -832,6 +815,7 @@ class GPTJForCausalLM(GPTJPreTrainedModel):
         output_attentions: Optional[bool] = None,
         output_hidden_states: Optional[bool] = None,
         return_dict: Optional[bool] = None,
+        **deprecated_arguments,
     ) -> Union[Tuple, CausalLMOutputWithPast]:
         r"""
         labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
@@ -839,6 +823,16 @@ class GPTJForCausalLM(GPTJPreTrainedModel):
             `labels = input_ids` Indices are selected in `[-100, 0, ..., config.vocab_size]` All labels set to `-100`
             are ignored (masked), the loss is only computed for labels in `[0, ..., config.vocab_size]`
         """
+        if deprecated_arguments.pop("position_ids", False) is not False:
+            # `position_ids` could have been `torch.Tensor` or `None` so defaulting pop to `False` allows to detect if users were passing explicitly `None`
+            warnings.warn(
+                "`position_ids` have no functionality in GPT-J and will be removed in v5.0.0. You can safely ignore"
+                " passing `position_ids`.",
+                FutureWarning,
+            )
+        if len(deprecated_arguments) > 0:
+            raise ValueError(f"Got unexpected arguments: {deprecated_arguments}")
+
         return_dict = return_dict if return_dict is not None else self.config.use_return_dict
 
         transformer_outputs = self.transformer(
@@ -846,7 +840,6 @@ class GPTJForCausalLM(GPTJPreTrainedModel):
             past_key_values=past_key_values,
             attention_mask=attention_mask,
             token_type_ids=token_type_ids,
-            position_ids=position_ids,
             head_mask=head_mask,
             inputs_embeds=inputs_embeds,
             use_cache=use_cache,
@@ -948,7 +941,6 @@ class GPTJForSequenceClassification(GPTJPreTrainedModel):
         past_key_values: Optional[Tuple[Tuple[torch.Tensor]]] = None,
         attention_mask: Optional[torch.FloatTensor] = None,
         token_type_ids: Optional[torch.LongTensor] = None,
-        position_ids: Optional[torch.LongTensor] = None,
         head_mask: Optional[torch.FloatTensor] = None,
         inputs_embeds: Optional[torch.FloatTensor] = None,
         labels: Optional[torch.LongTensor] = None,
@@ -956,6 +948,7 @@ class GPTJForSequenceClassification(GPTJPreTrainedModel):
         output_attentions: Optional[bool] = None,
         output_hidden_states: Optional[bool] = None,
         return_dict: Optional[bool] = None,
+        **deprecated_arguments,
     ) -> Union[Tuple, SequenceClassifierOutputWithPast]:
         r"""
         labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
@@ -963,6 +956,16 @@ class GPTJForSequenceClassification(GPTJPreTrainedModel):
             config.num_labels - 1]`. If `config.num_labels == 1` a regression loss is computed (Mean-Square loss), If
             `config.num_labels > 1` a classification loss is computed (Cross-Entropy).
         """
+        if deprecated_arguments.pop("position_ids", False) is not False:
+            # `position_ids` could have been `torch.Tensor` or `None` so defaulting pop to `False` allows to detect if users were passing explicitly `None`
+            warnings.warn(
+                "`position_ids` have no functionality in GPT-J and will be removed in v5.0.0. You can safely ignore"
+                " passing `position_ids`.",
+                FutureWarning,
+            )
+        if len(deprecated_arguments) > 0:
+            raise ValueError(f"Got unexpected arguments: {deprecated_arguments}")
+
         return_dict = return_dict if return_dict is not None else self.config.use_return_dict
 
         transformer_outputs = self.transformer(
@@ -970,7 +973,6 @@ class GPTJForSequenceClassification(GPTJPreTrainedModel):
             past_key_values=past_key_values,
             attention_mask=attention_mask,
             token_type_ids=token_type_ids,
-            position_ids=position_ids,
             head_mask=head_mask,
             inputs_embeds=inputs_embeds,
             use_cache=use_cache,
@@ -1072,7 +1074,6 @@ class GPTJForQuestionAnswering(GPTJPreTrainedModel):
         input_ids: Optional[torch.LongTensor] = None,
         attention_mask: Optional[torch.FloatTensor] = None,
         token_type_ids: Optional[torch.LongTensor] = None,
-        position_ids: Optional[torch.LongTensor] = None,
         head_mask: Optional[torch.FloatTensor] = None,
         inputs_embeds: Optional[torch.FloatTensor] = None,
         start_positions: Optional[torch.LongTensor] = None,
@@ -1080,6 +1081,7 @@ class GPTJForQuestionAnswering(GPTJPreTrainedModel):
         output_attentions: Optional[bool] = None,
         output_hidden_states: Optional[bool] = None,
         return_dict: Optional[bool] = None,
+        **deprecated_arguments,
     ) -> Union[Tuple, QuestionAnsweringModelOutput]:
         r"""
         start_positions (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
@@ -1091,13 +1093,22 @@ class GPTJForQuestionAnswering(GPTJPreTrainedModel):
             Positions are clamped to the length of the sequence (`sequence_length`). Position outside of the sequence
             are not taken into account for computing the loss.
         """
+        if deprecated_arguments.pop("position_ids", False) is not False:
+            # `position_ids` could have been `torch.Tensor` or `None` so defaulting pop to `False` allows to detect if users were passing explicitly `None`
+            warnings.warn(
+                "`position_ids` have no functionality in GPT-J and will be removed in v5.0.0. You can safely ignore"
+                " passing `position_ids`.",
+                FutureWarning,
+            )
+        if len(deprecated_arguments) > 0:
+            raise ValueError(f"Got unexpected arguments: {deprecated_arguments}")
+
         return_dict = return_dict if return_dict is not None else self.config.use_return_dict
 
         outputs = self.transformer(
             input_ids,
             attention_mask=attention_mask,
             token_type_ids=token_type_ids,
-            position_ids=position_ids,
             head_mask=head_mask,
             inputs_embeds=inputs_embeds,
             output_attentions=output_attentions,