Create app.py

app.py

@@ -0,0 +1,80 @@
+import numpy as np
+import faiss
+from transformers import AutoTokenizer, AutoModel
+from groq import Groq
+from datasets import load_dataset
+import gradio as gr
+
+# Initialize the Groq client
+groq_api_key = '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'
+client = Groq(api_key=groq_api_key)
+
+# Load dataset
+dataset = load_dataset("midrees2806/7K_Dataset")
+
+# Preprocessing function
+def preprocess_data(text):
+    tokenizer = AutoTokenizer.from_pretrained('bert-base-uncased')
+    model = AutoModel.from_pretrained('bert-base-uncased')
+    inputs = tokenizer(text, return_tensors='pt', max_length=512, truncation=True, padding=True)
+    return model(**inputs).last_hidden_state.mean(dim=1).detach().numpy()
+
+# Prepare embeddings
+print("Preparing embeddings...")
+train_dataset = dataset['train']
+chunked_embeddings = []
+for data in train_dataset:
+    text = data['text']
+    chunked_embeddings.append(preprocess_data(text))
+chunked_embeddings = np.vstack(chunked_embeddings)
+
+# Initialize FAISS index
+dimension = chunked_embeddings.shape[1]
+index = faiss.IndexFlatL2(dimension)
+index.add(chunked_embeddings)
+
+# Groq response function
+def get_groq_response(query):
+    chat_completion = client.chat.completions.create(
+        messages=[{"role": "user", "content": query}],
+        model="llama3-70b-8192",
+    )
+    return chat_completion.choices[0].message.content
+
+# FAISS search function
+def search_in_faiss(query):
+    query_embedding = preprocess_data(query)
+    distances, indices = index.search(query_embedding, k=5)
+    return [dataset['train'][int(idx)]['text'] for idx in indices[0]]
+
+# Gradio Chat Interface
+def respond(message, chat_history):
+    try:
+        faiss_results = search_in_faiss(message)
+        model_response = get_groq_response(message)
+
+        bot_response = "**Relevant Information from Dataset:**\n\n"
+        for result in faiss_results:
+            bot_response += f"- {result}\n\n"
+        bot_response = "**Model Response:**\n\n" + model_response
+                   #+
+        return "", chat_history + [(message, bot_response)]
+    except Exception as e:
+        print(f"Error: {str(e)}")
+        return "", chat_history + [(message, f"Error processing request: {str(e)}")]
+
+# Create interface
+with gr.Blocks(theme=gr.themes.Soft()) as demo:
+    gr.Markdown("# <center>UOE Chatbot</center>")
+    gr.Markdown("<center>Ask any question and get answers powered by Groq and FAISS</center>")
+
+    chatbot = gr.Chatbot(height=500)
+    msg = gr.Textbox(label="Type your message here...", placeholder="Ask me anything...")
+    clear = gr.Button("Clear")
+
+    msg.submit(respond, [msg, chatbot], [msg, chatbot])
+    clear.click(lambda: None, None, chatbot, queue=False)
+
+# Launch application
+if __name__ == "__main__":
+    demo.launch()