Update app.py

app.py

@@ -1,4 +1,8 @@
-#import gradio as gr  
+import gradio as gr  
+from transformers import pipeline
+from peft import AutoPeftModelForCausalLM
+from transformers import AutoTokenizer
+
 #from llama_cpp import Llama  
   
 # Load the Llama model  
@@ -21,17 +25,26 @@
     # Extract the model's reply  
   #  model_reply = response['choices'][0]['message']['content']  
    # return model_reply  
+def generate_answer(user_input):
+    model = AutoPeftModelForCausalLM.from_pretrained(
+        "GSridhar1982/AIML_QA_Llama31_FineTuned_UsingLora", # YOUR MODEL YOU USED FOR TRAINING
+        load_in_4bit = load_in_4bit,
+    )
+    tokenizer = AutoTokenizer.from_pretrained("GSridhar1982/AIML_QA_Llama31_FineTuned_UsingLora")
+    # Create a text generation pipeline
+    generator = pipeline(task="text-generation", model=model, tokenizer=tokenizer)
 
-from transformers import pipeline
-
-# Replace with your Hugging Face model space URL
-model_id = "GSridhar1982/AIML_QA_Llama31_FineTuned_UsingLora" 
-
-pipe = pipeline("text2text-generation", model=model_id)
+    # Generate predictions on the test dataset
+    # Access the input column of the dataset using the column name
+    predictions = generator(
+        user_input,
+        max_new_tokens=100,
+        num_beams=1,
+    )
 
-def generate_text(prompt):
-  """Generates text using the loaded Hugging Face model."""
-  return pipe(prompt)[0]['generated_text']
+    # Extract the generated text from the pipeline output
+    predictions = [pred[0]['generated_text'] for pred in predictions]
+    return predictions
 
 # Create a Gradio interface  
 iface = gr.Interface(