Upload app.py with huggingface_hub

app.py

@@ -10,17 +10,15 @@ Original file is located at
 import gradio as gr
 import torch
 import torch.nn.functional as F
-from transformers import AutoTokenizer, AutoModelForSequenceClassification
-from safetensors.torch import load_file
+from transformers import BertTokenizer, BertForSequenceClassification
 import os
 
 def load_model():
-    # Load the model architecture
+    # Use BertForSequenceClassification instead of AutoModelForSequenceClassification
     model_name = "indobenchmark/indobert-lite-large-p2"
-    model = AutoModelForSequenceClassification.from_pretrained(
+    model = BertForSequenceClassification.from_pretrained(
         model_name, 
         num_labels=3,
-        # Add these parameters to avoid random initialization warnings
         local_files_only=False,
         ignore_mismatched_sizes=True
     )
@@ -30,7 +28,7 @@ def load_model():
         # Option 1: Load from local safetensors file
         local_model_path = "model.safetensors"
         if os.path.exists(local_model_path):
-            weights = load_file(local_model_path)
+            weights = torch.load(local_model_path)
             model.load_state_dict(weights)
         else:
             print("Warning: No local model weights found. Using pre-trained weights.")
@@ -45,7 +43,7 @@ def load_model():
 model = load_model()
 
 # Load the tokenizer
-tokenizer = AutoTokenizer.from_pretrained('indobenchmark/indobert-lite-large-p2')
+tokenizer = BertTokenizer.from_pretrained('indobenchmark/indobert-lite-large-p2')
 
 def predict_stress_with_accuracy(text_input):
     if not text_input.strip():