Update app.py

app.py

@@ -1,64 +1,99 @@
 import gradio as gr
 from huggingface_hub import InferenceClient
+from sentence_transformers import SentenceTransformer
+import faiss
+import numpy as np
+import pdfplumber
+import re
 
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
+# Initialize the InferenceClient
 client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
 
+# Function to extract text from PDFs
+def extract_text_from_pdf(pdf_path):
+    text = ""
+    with pdfplumber.open(pdf_path) as pdf:
+        for page in pdf.pages:
+            page_text = page.extract_text()
+            if page_text:
+                text += page_text
+    return text
 
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
+# Clean the extracted text
+def clean_extracted_text(text):
+    # Removing any unnecessary characters, such as file paths and non-text data
+    cleaned_text = re.sub(r'file://[^\n]*', '', text)  # Remove file paths
+    cleaned_text = re.sub(r'\d{1,2}/\d{1,2}/\d{4}', '', cleaned_text)  # Remove dates
+    cleaned_text = re.sub(r'[^a-zA-Z0-9\u0600-\u06FF\s\u00C0-\u00FF]+', '', cleaned_text)  # Keep Arabic and basic text
+    return cleaned_text.strip()
 
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
+# Path to the uploaded PDF file
+pdf_path = "/content/Noor-Book.com  القاموس عربي فرنسي بالمصطلحات العلمية و الصور 3  (1).pdf"
 
-    messages.append({"role": "user", "content": message})
+# Extract and clean text from the provided PDF
+pdf_text = extract_text_from_pdf(pdf_path)
+cleaned_text = clean_extracted_text(pdf_text)
 
+# Split the cleaned text into chunks for processing
+def chunk_text(text, chunk_size=300):
+    sentences = text.split('. ')
+    chunks, current_chunk = [], ""
+    for sentence in sentences:
+        if len(current_chunk) + len(sentence) <= chunk_size:
+            current_chunk += sentence + ". "
+        else:
+            chunks.append(current_chunk.strip())
+            current_chunk = sentence + ". "
+    if current_chunk:
+        chunks.append(current_chunk.strip())
+    return chunks
+
+# Chunk the cleaned text
+chunked_text = chunk_text(cleaned_text)
+
+# Load pre-trained Sentence Transformer model for embeddings
+model = SentenceTransformer("all-MiniLM-L6-v2")
+index = faiss.IndexFlatL2(model.get_sentence_embedding_dimension())
+
+# Generate embeddings for the chunks
+embeddings = model.encode(chunked_text, convert_to_tensor=True).detach().cpu().numpy()
+index.add(embeddings)
+
+# Function to generate response from the model
+def respond(message, history, system_message, max_tokens, temperature, top_p):
+    # Step 1: Retrieve relevant chunks based on the user query
+    query_embedding = model.encode([message], convert_to_tensor=True).detach().cpu().numpy()
+    k = 5  # Number of relevant chunks to retrieve
+    _, indices = index.search(query_embedding, k)
+    relevant_chunks = " ".join([chunked_text[idx] for idx in indices[0]])
+
+    # Step 2: Create prompt for the language model
+    prompt = f"{system_message}\n\nUser Query: {message}\n\nRelevant Information: {relevant_chunks}"
     response = ""
 
+    # Step 3: Generate response using the HuggingFace model
     for message in client.chat_completion(
-        messages,
+        [{"role": "system", "content": system_message}, {"role": "user", "content": message}],
         max_tokens=max_tokens,
         stream=True,
         temperature=temperature,
         top_p=top_p,
     ):
         token = message.choices[0].delta.content
-
         response += token
         yield response
 
-
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
+# Create the Gradio interface with additional inputs
 demo = gr.ChatInterface(
     respond,
     additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
+        gr.Textbox(value="You are a helpful and empathetic mental health assistant.", label="System message"),
         gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
         gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
+        gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)"),
     ],
 )
 
-
+# Launch the Gradio interface
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()