Fix capitalization issues

onnx/modeling_phi4mm.py

@@ -1412,7 +1412,7 @@ class Phi4MMAttention(nn.Module):
 
 class Phi4MMFlashAttention2(Phi4MMAttention):
     """
-    Phi-O flash attention module. This module inherits from `Phi4MMAttention` as the weights of the module stays
+    Phi-4-MM flash attention module. This module inherits from `Phi4MMAttention` as the weights of the module stays
     untouched. The only required change would be on the forward pass where it needs to correctly call the public API of
     flash attention and deal with padding tokens in case the input contains any of them.
     """
@@ -1736,7 +1736,7 @@ PHI4MM_START_DOCSTRING = r"""
 
 
 @add_start_docstrings(
-    "The bare Phi-O model outputting raw hidden-states without any specific head on top.",
+    "The bare Phi-4-MM model outputting raw hidden-states without any specific head on top.",
     PHI4MM_START_DOCSTRING,
 )
 class Phi4MMPreTrainedModel(PreTrainedModel):
@@ -1839,7 +1839,7 @@ PHI4MM_INPUTS_DOCSTRING = r"""
 
 
 @add_start_docstrings(
-    "The bare Phi-O model outputting raw hidden-states without any specific head on top.",
+    "The bare Phi-4-MM model outputting raw hidden-states without any specific head on top.",
     PHI4MM_START_DOCSTRING,
 )
 class Phi4MMModel(Phi4MMPreTrainedModel):
@@ -1883,7 +1883,7 @@ class Phi4MMModel(Phi4MMPreTrainedModel):
     def set_input_embeddings(self, value):
         self.embed_tokens = value
 
-    @add_start_docstrings_to_model_forward(Phi4MM_INPUTS_DOCSTRING)
+    @add_start_docstrings_to_model_forward(PHI4MM_INPUTS_DOCSTRING)
     def forward(
         self,
         input_ids: torch.LongTensor = None,
@@ -2260,7 +2260,7 @@ class Phi4MMForCausalLM(Phi4MMPreTrainedModel, GenerationMixin):
         return self.model
 
     # Ignore copy
-    @add_start_docstrings_to_model_forward(Phi4MM_INPUTS_DOCSTRING)
+    @add_start_docstrings_to_model_forward(PHI4MM_INPUTS_DOCSTRING)
     @replace_return_docstrings(output_type=CausalLMOutputWithPast, config_class=_CONFIG_FOR_DOC)
     def forward(
         self,
@@ -2473,7 +2473,7 @@ class Phi4MMForSequenceClassification(Phi4MMPreTrainedModel):
     def set_input_embeddings(self, value):
         self.model.embed_tokens = value
 
-    @add_start_docstrings_to_model_forward(Phi4MM_INPUTS_DOCSTRING)
+    @add_start_docstrings_to_model_forward(PHI4MM_INPUTS_DOCSTRING)
     def forward(
         self,
         input_ids: torch.LongTensor = None,
@@ -2572,7 +2572,7 @@ class Phi4MMForTokenClassification(Phi4MMPreTrainedModel):
         # Initialize weights and apply final processing
         self.post_init()
 
-    @add_start_docstrings_to_model_forward(Phi4MM_INPUTS_DOCSTRING)
+    @add_start_docstrings_to_model_forward(PHI4MM_INPUTS_DOCSTRING)
     @add_code_sample_docstrings(
         checkpoint=_CHECKPOINT_FOR_DOC,
         output_type=TokenClassifierOutput,