updated model weights path

app.py

@@ -6,31 +6,13 @@ import gradio as gr
 from transformers import BertTokenizer, BertForSequenceClassification
 import os
 
-model_zip_path = "BioBERT_Model.zip"
-tokenizer_zip_path = "BioBERT_Tokenizer.zip"
+tokenizer = BertTokenizer.from_pretrained("dmis-lab/biobert-base-cased-v1.1")
+model = BertForSequenceClassification.from_pretrained("dmis-lab/biobert-base-cased-v1.1", num_labels=2)
 
-# Assume model and tokenizer are already loaded
-def load_model_and_tokenizer():
-    if not os.path.exists('BioBERT_Model'):
-        with zipfile.ZipFile(model_zip_path, 'r') as zip_ref:
-            zip_ref.extractall('BioBERT_Model')
+# loading the pretrained weights into the model
+model.load_state_dict(torch.load('Bio_BERT_model.pth'))
 
-    if not os.path.exists('BioBERT_Tokenizer'):
-        with zipfile.ZipFile(tokenizer_zip_path, 'r') as zip_ref:
-            zip_ref.extractall('BioBERT_Tokenizer')
-
-    model_path = 'BioBERT_Model/content/BioBERT_Model'
-    tokenizer_path = 'BioBERT_Tokenizer/content/BioBERT_Tokenizer'
-
-    model = BertForSequenceClassification.from_pretrained(model_path)
-    tokenizer = BertTokenizer.from_pretrained(tokenizer_path)
-
-    return model, tokenizer
-
-
-model, tokenizer = load_model_and_tokenizer()
 device = "cpu"
-model = model.to(device)
 
 def predict_drug_target_interaction(sentence):
     # Tokenize the input sentence