Spaces:

bardicreels
/

rag

Running

App Files Files Community

user commited on Sep 11, 2024

Commit

b5553ae

1 Parent(s): e5a2f3a

modifications for remote development using huggingface resouces

Browse files

Files changed (7) hide show

app.py +31 -20
character_generator.py +10 -0
index.html +17 -44
model_loader.py +8 -2
pdf_processor.py +13 -9
requirements.txt +2 -5
run.py +4 -0

app.py CHANGED Viewed

@@ -1,16 +1,25 @@
 from flask import Flask, request, jsonify, send_from_directory
-from pdf_processor import extract_text_from_pdf, process_pdf
-from model_loader import load_model, load_generator
-import torch
-import numpy as np
 app = Flask(__name__)
-# Load data and models
-pdf_text = extract_text_from_pdf("data/example.pdf")
-chunks, index = process_pdf(pdf_text)
-tokenizer, model = load_model()
-generator = load_generator()
 @app.route('/')
 def home():
@@ -20,18 +29,20 @@ def home():
 def ask():
     prompt = request.json['question']
-    inputs = tokenizer(prompt, return_tensors="pt", padding=True, truncation=True, max_length=512)
-    with torch.no_grad():
-        query_vector = model(**inputs).last_hidden_state.mean(dim=1).squeeze().numpy()
-    k = 3
-    distances, indices = index.search(query_vector.astype('float32').reshape(1, -1), k)
-    context = " ".join([chunks[i] for i in indices[0]])
-    input_text = f"Context: {context}\n\nQuestion: {prompt}\n\nAnswer:"
-    response = generator(input_text, max_length=150, num_return_sequences=1)[0]['generated_text']
     return jsonify({'response': response})
 if __name__ == '__main__':
-    app.run(host='0.0.0.0', port=7860)

 from flask import Flask, request, jsonify, send_from_directory
+import requests
+import os
+from dotenv import load_dotenv
+load_dotenv()
 app = Flask(__name__)
+API_URL = "https://api-inference.huggingface.co/models/"
+headers = {"Authorization": f"Bearer {os.getenv('HUGGINGFACE_API_KEY')}"}
+# Sample text for testing
+sample_text = """
+This is a sample text for testing our RAG chatbot.
+It contains information about artificial intelligence and machine learning.
+AI and ML are revolutionizing various industries and improving efficiency.
+"""
+def query(payload, model):
+    response = requests.post(API_URL + model, headers=headers, json=payload)
+    return response.json()
 @app.route('/')
 def home():
 def ask():
     prompt = request.json['question']
+    # Use sentence-transformers model for embedding
+    embedding_model = "sentence-transformers/all-MiniLM-L6-v2"
+    context_embedding = query({"inputs": sample_text}, embedding_model)[0]
+    query_embedding = query({"inputs": prompt}, embedding_model)[0]
+    # Simple dot product similarity
+    similarity = sum(a*b for a, b in zip(context_embedding, query_embedding))
+    # Generate response using T5 model
+    generator_model = "google/flan-t5-small"
+    input_text = f"Context: {sample_text}\n\nQuestion: {prompt}\n\nAnswer:"
+    response = query({"inputs": input_text}, generator_model)[0]["generated_text"]
     return jsonify({'response': response})
 if __name__ == '__main__':
+    app.run(debug=True)

character_generator.py ADDED Viewed

	@@ -0,0 +1,10 @@

+def generate_character_response(context, prompt):
+    # Define character traits based on your PDF content
+    character_traits = "I am a knowledgeable and friendly AI assistant based on the content of the provided PDF."
+    input_text = f"Character: {character_traits}\nContext: {context}\nUser: {prompt}\nCharacter's response:"
+    # Use your text generation model here
+    response = generator(input_text, max_length=150, num_return_sequences=1)[0]['generated_text']
+    return response

index.html CHANGED Viewed

@@ -3,59 +3,32 @@
 <head>
     <meta charset="UTF-8">
     <meta name="viewport" content="width=device-width, initial-scale=1.0">
-    <title>PDF RAG Chatbot</title>
-    <style>
-        body { font-family: Arial, sans-serif; max-width: 800px; margin: 0 auto; padding: 20px; }
-        #chat-container { border: 1px solid #ddd; height: 400px; overflow-y: scroll; padding: 10px; margin-bottom: 10px; }
-        #user-input { width: 70%; padding: 5px; }
-        #send-button { padding: 5px 10px; }
-    </style>
 </head>
 <body>
-    <h1>PDF RAG Chatbot</h1>
     <div id="chat-container"></div>
     <input type="text" id="user-input" placeholder="Ask a question...">
-    <button id="send-button">Send</button>
     <script>
-        const chatContainer = document.getElementById('chat-container');
-        const userInput = document.getElementById('user-input');
-        const sendButton = document.getElementById('send-button');
-        function addMessage(role, content) {
-            const messageDiv = document.createElement('div');
-            messageDiv.innerHTML = `<strong>${role}:</strong> ${content}`;
-            chatContainer.appendChild(messageDiv);
-            chatContainer.scrollTop = chatContainer.scrollHeight;
-        }
         async function sendMessage() {
-            const question = userInput.value.trim();
-            if (question) {
-                addMessage('User', question);
-                userInput.value = '';
-                try {
-                    const response = await fetch('/ask', {
-                        method: 'POST',
-                        headers: {
-                            'Content-Type': 'application/json',
-                        },
-                        body: JSON.stringify({ question }),
-                    });
-                    const data = await response.json();
-                    addMessage('Bot', data.response);
-                } catch (error) {
-                    console.error('Error:', error);
-                    addMessage('Bot', 'Sorry, there was an error processing your request.');
-                }
-            }
-        }
-        sendButton.addEventListener('click', sendMessage);
-        userInput.addEventListener('keypress', (e) => {
-            if (e.key === 'Enter') sendMessage();
-        });
     </script>
 </body>
 </html>

 <head>
     <meta charset="UTF-8">
     <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <title>RAG Chatbot</title>
 </head>
 <body>
+    <h1>RAG Chatbot</h1>
     <div id="chat-container"></div>
     <input type="text" id="user-input" placeholder="Ask a question...">
+    <button onclick="sendMessage()">Send</button>
     <script>
         async function sendMessage() {
+            const input = document.getElementById('user-input');
+            const message = input.value;
+            input.value = '';
+            const response = await fetch('/ask', {
+                method: 'POST',
+                headers: {
+                    'Content-Type': 'application/json',
+                },
+                body: JSON.stringify({ question: message }),
+            });
+            const data = await response.json();
+            const chatContainer = document.getElementById('chat-container');
+            chatContainer.innerHTML += `<p><strong>You:</strong> ${message}</p><p><strong>Bot:</strong> ${data.response}</p>`;
+        }
     </script>
 </body>
 </html>

model_loader.py CHANGED Viewed

@@ -1,9 +1,15 @@
 from transformers import AutoTokenizer, AutoModel, pipeline
 def load_model():
     tokenizer = AutoTokenizer.from_pretrained("sentence-transformers/all-MiniLM-L6-v2")
-    model = AutoModel.from_pretrained("sentence-transformers/all-MiniLM-L6-v2")
     return tokenizer, model
 def load_generator():
-    return pipeline('text2text-generation', model='google/flan-t5-base')

 from transformers import AutoTokenizer, AutoModel, pipeline
+from huggingface_hub import hf_hub_download
 def load_model():
     tokenizer = AutoTokenizer.from_pretrained("sentence-transformers/all-MiniLM-L6-v2")
+    model = AutoModel.from_pretrained("sentence-transformers/all-MiniLM-L6-v2", device_map="cpu")
     return tokenizer, model
 def load_generator():
+    return pipeline('text2text-generation', model='google/flan-t5-base', device_map="cpu")
+def download_pdf():
+    # Replace 'your_pdf_file.pdf' with the actual name of your PDF file on Hugging Face
+    file_path = hf_hub_download(repo_id="your_username/your_repo_name", filename="your_pdf_file.pdf")
+    return file_path

pdf_processor.py CHANGED Viewed

@@ -1,22 +1,26 @@
-import fitz
-import faiss
 import numpy as np
-import torch
-from model_loader import load_model
 def extract_text_from_pdf(file_path):
     with fitz.open(file_path) as doc:
-        return " ".join(page.get_text() for page in doc)
 def process_pdf(pdf_text):
     chunks = [pdf_text[i:i+512] for i in range(0, len(pdf_text), 512)]
-    tokenizer, model = load_model()
     embeddings = []
     for chunk in chunks:
-        inputs = tokenizer(chunk, return_tensors="pt", padding=True, truncation=True, max_length=512)
-        with torch.no_grad():
-            outputs = model(**inputs)
         embeddings.append(outputs.last_hidden_state.mean(dim=1).squeeze().numpy())
     embeddings = np.array(embeddings)
     index = faiss.IndexFlatL2(embeddings.shape[1])
     index.add(embeddings.astype('float32'))

+import fitz  # PyMuPDF
 import numpy as np
+from transformers import AutoTokenizer, AutoModel
+import faiss
 def extract_text_from_pdf(file_path):
+    text = ""
     with fitz.open(file_path) as doc:
+        for page in doc:
+            text += page.get_text()
+    return text
 def process_pdf(pdf_text):
     chunks = [pdf_text[i:i+512] for i in range(0, len(pdf_text), 512)]
+    tokenizer = AutoTokenizer.from_pretrained("sentence-transformers/all-MiniLM-L6-v2")
+    model = AutoModel.from_pretrained("sentence-transformers/all-MiniLM-L6-v2")
     embeddings = []
     for chunk in chunks:
+        inputs = tokenizer(chunk, padding=True, truncation=True, max_length=512, return_tensors="pt")
+        outputs = model(**inputs)
         embeddings.append(outputs.last_hidden_state.mean(dim=1).squeeze().numpy())
     embeddings = np.array(embeddings)
     index = faiss.IndexFlatL2(embeddings.shape[1])
     index.add(embeddings.astype('float32'))

requirements.txt CHANGED Viewed

@@ -1,6 +1,3 @@
 flask
-transformers
-torch
-PyMuPDF
-faiss-cpu
-numpy

 flask
+requests
+python-dotenv

run.py ADDED Viewed

	@@ -0,0 +1,4 @@

+from app import app
+if __name__ == '__main__':
+    app.run(debug=True)