Update audio_text_multimodal.py

audio_text_multimodal.py

@@ -21,7 +21,7 @@ class MultiModalConfig(PretrainedConfig):
         super().__init__(**kwargs)
 
 
-class Wav2Vec2BertConfig(MultiModalConfig):
+class WavLMBertConfig(MultiModalConfig):
     ...
 
 
@@ -187,27 +187,27 @@ class AudioTextModelForSequenceBaseClassification(BaseClassificationModel):
         )
 
 
-class Wav2Vec2BertForSequenceClassification(AudioTextModelForSequenceBaseClassification):
+class WavLMBertForSequenceClassification(AudioTextModelForSequenceBaseClassification):
     """
-    Wav2Vec2BertForSequenceClassification is a model for sequence classification task
+    WavLMBertForSequenceClassification is a model for sequence classification task
      (e.g. sentiment analysis, text classification, etc.)
 
     Args:
-        config (Wav2Vec2BertConfig): config
+        config (WavLMBertConfig): config
 
     Attributes:
-        config (Wav2Vec2BertConfig): config
-        audio_config (Wav2Vec2Config): wav2vec2 config
+        config (WavLMBertConfig): config
+        audio_config (WavLMConfig): wav2vec2 config
         text_config (BertConfig): bert config
-        audio_model (Wav2Vec2Model): wav2vec2 model
+        audio_model (WavLMModel): wav2vec2 model
         text_model (BertModel): bert model
         classifier (torch.nn.Linear): classifier
     """
     def __init__(self, config):
         super().__init__(config)
-        self.audio_config = Wav2Vec2Config.from_dict(self.config.Wav2Vec2Model)
+        self.audio_config = WavLMConfig.from_dict(self.config.WavLMModel)
         self.text_config = BertConfig.from_dict(self.config.BertModel)
-        self.audio_model = Wav2Vec2Model(self.audio_config)
+        self.audio_model = WavLMModel(self.audio_config)
         self.text_model = BertModel(self.text_config)
         self.classifier = torch.nn.Linear(
             self.audio_config.hidden_size + self.text_config.hidden_size, self.num_labels