Update app.py

app.py

@@ -1,14 +1,25 @@
 import gradio as gr
 from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
 
-from transformers import GPT2LMHeadModel, GPT2Config
-
-model_name = "facebook/bart-large-cnn"
-tokenizer = AutoTokenizer.from_pretrained(model_name)
-model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
+# Define the models and their corresponding names
+models = {
+    "Model 1 (facebook/bart-large-cnn)": {
+        "model_name": "facebook/bart-large-cnn",
+        "description": "Model 1",
+    },
+    "Model 2 ()": {
+        "model_name": "google/pegasus-multi_news",
+        "description": "Model 2",
+    },
+}
 
 # Define the Gradio interface
-def summarize_text(input_text):
+def summarize_text(input_text, selected_model):
+    # Get the selected model and its tokenizer
+    model_info = models[selected_model]
+    tokenizer = AutoTokenizer.from_pretrained(model_info["model_name"])
+    model = AutoModelForSeq2SeqLM.from_pretrained(model_info["model_name"])
+    
     # Tokenize and generate summary
     input_ids = tokenizer.encode(input_text, return_tensors="pt", max_length=1024, truncation=True)
     summary_ids = model.generate(input_ids, max_length=10, min_length=1, length_penalty=1.0, num_beams=4, early_stopping=True)
@@ -17,11 +28,11 @@ def summarize_text(input_text):
 
 iface = gr.Interface(
     fn=summarize_text,
-    inputs="text",
+    inputs=[gr.inputs.Textbox(label="Input Text"), gr.inputs.Radio(model_names, label="Select Model")],
     outputs="text",
-    title="Concept Tagger",
-    description="Concept tag your text using Theus model (1.3B Concept Tagging).",
+    title="Text Summarization App",
+    description="Choose a model for text summarization and enter the text to summarize.",
 )
 
 if __name__ == "__main__":
-    iface.launch()
+    iface.launch()