Create app.py

app.py

@@ -0,0 +1,56 @@
+import os
+import gradio as gr
+import asyncio
+from langchain_core.prompts import PromptTemplate
+from langchain_community.document_loaders import PyPDFLoader
+from langchain.chains.question_answering import load_qa_chain
+import torch
+from transformers import AutoTokenizer, AutoModelForCausalLM
+
+# Load Mistral model
+model_path = "nvidia/Mistral-NeMo-Minitron-8B-Base"
+tokenizer = AutoTokenizer.from_pretrained(model_path)
+device = 'cuda' if torch.cuda.is_available() else 'cpu'
+dtype = torch.bfloat16
+model = AutoModelForCausalLM.from_pretrained(model_path, torch_dtype=dtype, device_map=device)
+
+async def initialize(file_path, question):
+    prompt_template = """Answer the question as precise as possible using the provided context. If the answer is not contained in the context, say "answer not available in context" \n\n Context: \n {context}?\n Question: \n {question} \n Answer: """
+    prompt = PromptTemplate(template=prompt_template, input_variables=["context", "question"])
+
+    if os.path.exists(file_path):
+        pdf_loader = PyPDFLoader(file_path)
+        pages = pdf_loader.load_and_split()
+        context = "\n".join(str(page.page_content) for page in pages[:30])
+        
+        # Prepare input for Mistral model
+        input_text = prompt.format(context=context, question=question)
+        inputs = tokenizer.encode(input_text, return_tensors='pt').to(device)
+        
+        # Generate the output
+        with torch.no_grad():
+            outputs = model.generate(inputs, max_length=500)  # Adjust max_length as needed
+        
+        # Decode and return the output
+        answer = tokenizer.decode(outputs[0], skip_special_tokens=True)
+        return answer
+    else:
+        return "Error: Unable to process the document. Please ensure the PDF file is valid."
+
+# Define Gradio Interface
+input_file = gr.File(label="Upload PDF File")
+input_question = gr.Textbox(label="Ask about the document")
+output_text = gr.Textbox(label="Answer - Mistral Model")
+
+async def pdf_qa(file, question):
+    answer = await initialize(file.name, question)
+    return answer
+
+# Create Gradio Interface
+gr.Interface(
+    fn=pdf_qa,
+    inputs=[input_file, input_question],
+    outputs=output_text,
+    title="RAG Knowledge Retrieval using Mistral Model",
+    description="Upload a PDF file and ask questions about the content."
+).launch()