Added both the model and its config in the same modeling.py file, deleted extra configuration file.

configuration_musheff.py

@@ -1,11 +0,0 @@
-# configuration_musheff.py
-from transformers import PretrainedConfig
-
-
-class MusheffConfig(PretrainedConfig):
-    model_type = "efficientnet_b3"
-
-    def __init__(self, num_classes=12, dropout_rate=0.3, **kwargs):
-        self.num_classes = num_classes
-        self.dropout_rate = dropout_rate
-        super().__init__(**kwargs)

modeling.py

@@ -3,7 +3,16 @@ from transformers import PreTrainedModel
 
 import torch.nn as nn
 
-from configuration_musheff import MusheffConfig
+from transformers import PretrainedConfig
+
+
+class MusheffConfig(PretrainedConfig):
+    model_type = "efficientnet_b3"
+
+    def __init__(self, num_classes=12, dropout_rate=0.3, **kwargs):
+        self.num_classes = num_classes
+        self.dropout_rate = dropout_rate
+        super().__init__(**kwargs)
 
 
 class Musheff(PreTrainedModel):