Update modeling_quasarv4.py

modeling_quasarv4.py

@@ -1143,8 +1143,8 @@ class QuasarV4ForCausalLM(QuasarV4PreTrainedModel):
     def get_decoder(self):
         return self.model
 
-    @add_start_docstrings_to_model_forward(QWEN2_INPUTS_DOCSTRING)
-    @replace_return_docstrings(output_type=CausalLMOutputWithPast, config_class=_CONFIG_FOR_DOC)
+    @add_start_docstrings_to_model_forward(QUASAR_V4_INPUTS_DOCSTRING)
+    @replace_return_docstrings(output_type=CausalLMOutputWithPast, config_class=QuasarV4Config)
     def forward(
         self,
         input_ids: torch.LongTensor = None,
@@ -1314,13 +1314,13 @@ class QuasarV4ForCausalLM(QuasarV4PreTrainedModel):
     padding tokens when `inputs_embeds` are passed instead of `input_ids`, it does the same (take the last value in
     each row of the batch).
     """,
-    QWEN2_START_DOCSTRING,
+    QUASAR_V4_START_DOCSTRING,
 )
 class QuasarV4ForSequenceClassification(QuasarV4PreTrainedModel):
     def __init__(self, config):
         super().__init__(config)
         self.num_labels = config.num_labels
-        self.model = Qwen2Model(config)
+        self.model = QuasarV4Model(config)
         self.score = nn.Linear(config.hidden_size, self.num_labels, bias=False)
 
         # Initialize weights and apply final processing
@@ -1332,7 +1332,7 @@ class QuasarV4ForSequenceClassification(QuasarV4PreTrainedModel):
     def set_input_embeddings(self, value):
         self.model.embed_tokens = value
 
-    @add_start_docstrings_to_model_forward(QWEN2_INPUTS_DOCSTRING)
+    @add_start_docstrings_to_model_forward(QUASAR_V4_INPUTS_DOCSTRING)
     def forward(
         self,
         input_ids: torch.LongTensor = None,