Upload 3 files

app.py

@@ -0,0 +1,177 @@
+import os
+import json
+import requests
+import logging
+from pathlib import Path
+from dotenv import load_dotenv
+from typing_extensions import List, TypedDict
+
+import gradio as gr
+
+from langchain_core.documents import Document
+from langchain_core.tools import tool
+from langchain_text_splitters import RecursiveCharacterTextSplitter
+from langchain_community.vectorstores import FAISS
+from langchain_community.embeddings import HuggingFaceEmbeddings
+from langchain_community.tools import DuckDuckGoSearchResults
+
+# ---------------- ENV & LOGGING ----------------
+
+load_dotenv()
+GROQ_API_KEY = os.getenv("GROQ_API_KEY")
+os.environ["USER_AGENT"] = "my-custom-agent"
+
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s [%(levelname)s] %(message)s",
+    handlers=[
+        logging.FileHandler("assistant.log"),
+        logging.StreamHandler()
+    ]
+)
+logger = logging.getLogger(__name__)
+
+# ---------------- EMBEDDINGS & VECTOR STORE ----------------
+
+embedding_model_name = "sentence-transformers/all-MiniLM-L6-v2"
+embeddings = HuggingFaceEmbeddings(model_name=embedding_model_name)
+
+vector_store_path = "faiss_index"
+if Path(vector_store_path).exists():
+    logger.info("📂 Loading existing vector store...")
+    vector_store = FAISS.load_local(vector_store_path, embeddings, allow_dangerous_deserialization=True)
+else:
+    logger.info("📦 Initializing new vector store...")
+    dummy_doc = Document(page_content="Init doc")
+    vector_store = FAISS.from_documents([dummy_doc], embedding=embeddings)
+    vector_store.index.reset()
+    vector_store.docstore._dict.clear()
+    vector_store.index_to_docstore_id.clear()
+
+splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200, add_start_index=True)
+
+# ---------------- GROQ LLM ----------------
+
+def query_groq_llm(prompt: str) -> str:
+    url = "https://api.groq.com/openai/v1/chat/completions"
+    headers = {
+        "Authorization": f"Bearer {GROQ_API_KEY}",
+        "Content-Type": "application/json"
+    }
+    payload = {
+        "model": "llama3-8b-8192",  # You can change to "mixtral-8x7b-32768" or others
+        "messages": [
+            {"role": "system", "content": "You are a helpful assistant."},
+            {"role": "user", "content": prompt}
+        ],
+        "temperature": 0.7
+    }
+    response = requests.post(url, headers=headers, json=payload)
+    response.raise_for_status()
+    return response.json()['choices'][0]['message']['content']
+
+# ---------------- DUCKDUCKGO TOOL ----------------
+
+@tool
+def search_tool(query: str) -> str:
+    """Search the web using DuckDuckGo."""
+    search = DuckDuckGoSearchResults()
+    return search.run(query)
+
+# ---------------- UTILS ----------------
+
+def is_vague(text: str) -> bool:
+    if not text.strip():
+        return True
+    vague_phrases = [
+        "i don't know", "not sure", "cannot answer", "no context", "not enough info",
+        "uncertain", "please provide", "you haven't", "unknown", "not found"
+    ]
+    return any(phrase in text.lower() for phrase in vague_phrases)
+
+def is_realtime_query(text: str) -> bool:
+    keywords = ["current", "today", "latest", "now", "who is", "trending", "new", "recent"]
+    return any(k in text.lower() for k in keywords)
+
+# ---------------- FALLBACK WEB SEARCH ----------------
+
+def use_tool_only(question: str) -> str:
+    logger.info(f"Real-time query detected: '{question}'")
+    result = search_tool.invoke({"query": question})
+    prompt = f"Here is information from the web:\n\n{result}\n\nAnswer this question: {question}"
+    response = query_groq_llm(prompt)
+
+    # Store new info if not already in memory
+    new_doc = Document(page_content=result)
+    chunks = splitter.split_documents([new_doc])
+    existing = vector_store.similarity_search(result, k=5)
+    already_exists = any(c.page_content.strip() == result.strip() for c in existing)
+
+    if not already_exists:
+        logger.info("Adding new content to vector store.")
+        vector_store.add_documents(chunks)
+        vector_store.save_local(vector_store_path)
+    else:
+        logger.info("Content already exists. Skipping add.")
+
+    return f"[From Web Search]\n\n{response}"
+
+# ---------------- ASK FUNCTION ----------------
+
+def ask(question: str) -> str:
+    logger.info(f"New Question: {question}")
+
+    # Use FAISS with score
+    retrieved_docs = vector_store.similarity_search_with_score(question, k=5)
+    threshold = 0.25  # adjust as needed
+    filtered_docs = [doc for doc, score in retrieved_docs if score < threshold]
+
+    if not filtered_docs:
+        logger.info("No relevant memory found. Using web search.")
+        return use_tool_only(question)
+
+    context = "\n\n".join(doc.page_content for doc in filtered_docs)
+    prompt_text = f"""You are a helpful assistant.
+
+Context:
+{context}
+
+Question: {question}"""
+
+    response = query_groq_llm(prompt_text)
+
+    if not is_vague(response):
+        logger.info("Answered using memory.")
+        return f"[From Memory]\n\n{response}"
+
+    logger.info("Memory response vague. Using web search.")
+    return use_tool_only(question)
+
+
+# ---------------- GRADIO UI ----------------
+
+with gr.Blocks(title="RAG + Groq LLM Assistant") as iface:
+    gr.Markdown(
+        """
+        # 🤖 RAG + Real-Time AI Assistant (Groq + HuggingFace Embeddings)  
+        Ask anything. The assistant uses memory or searches the web if needed.
+        """
+    )
+    with gr.Row():
+        with gr.Column(scale=1):
+            chatbot = gr.Chatbot(label="Assistant", height=400)
+        with gr.Column(scale=1):
+            msg = gr.Textbox(label="Your message", placeholder="Ask a question...", show_label=False)
+            submit_btn = gr.Button("Submit")
+
+    def user_message_handler(message, history):
+        logger.info(f"User: {message}")
+        response = ask(message)
+        history.append((message, response))
+        return "", history
+
+    submit_btn.click(user_message_handler, [msg, chatbot], [msg, chatbot])
+    msg.submit(user_message_handler, [msg, chatbot], [msg, chatbot])
+
+if __name__ == "__main__":
+    iface.launch(share=True)

assistant.log

@@ -0,0 +1,236 @@
+2025-05-04 23:17:54,437 [INFO] Loading faiss.
+2025-05-04 23:17:54,461 [INFO] Successfully loaded faiss.
+2025-05-04 23:17:54,466 [INFO] Failed to load GPU Faiss: name 'GpuIndexIVFFlat' is not defined. Will not load constructor refs for GPU indexes. This is only an error if you're trying to use GPU Faiss.
+2025-05-04 23:17:54,726 [INFO] HTTP Request: GET http://127.0.0.1:7860/gradio_api/startup-events "HTTP/1.1 200 OK"
+2025-05-04 23:17:54,747 [INFO] HTTP Request: HEAD http://127.0.0.1:7860/ "HTTP/1.1 200 OK"
+2025-05-04 23:17:54,893 [INFO] HTTP Request: GET https://api.gradio.app/pkg-version "HTTP/1.1 200 OK"
+2025-05-04 23:17:55,058 [INFO] HTTP Request: GET https://api.gradio.app/v3/tunnel-request "HTTP/1.1 200 OK"
+2025-05-04 23:18:03,030 [INFO] New Question: who does a pen work
+2025-05-04 23:18:03,812 [INFO] HTTP Request: POST https://api.openai.com/v1/embeddings "HTTP/1.1 200 OK"
+2025-05-04 23:18:24,777 [INFO] HTTP Request: POST https://api.openai.com/v1/chat/completions "HTTP/1.1 200 OK"
+2025-05-04 23:18:24,791 [INFO] Answered using memory.
+2025-05-04 23:20:32,691 [INFO] New Question: Who is Prashanth Reddy Varanganti
+2025-05-04 23:20:34,828 [INFO] HTTP Request: POST https://api.openai.com/v1/chat/completions "HTTP/1.1 200 OK"
+2025-05-04 23:20:35,852 [INFO] response: https://html.duckduckgo.com/html 200
+2025-05-04 23:20:43,926 [INFO] HTTP Request: POST https://api.openai.com/v1/chat/completions "HTTP/1.1 200 OK"
+2025-05-04 23:20:44,863 [INFO] HTTP Request: POST https://api.openai.com/v1/embeddings "HTTP/1.1 200 OK"
+2025-05-04 23:20:44,867 [INFO] Adding new content to vector store.
+2025-05-04 23:20:45,271 [INFO] HTTP Request: POST https://api.openai.com/v1/embeddings "HTTP/1.1 200 OK"
+2025-05-04 23:20:58,838 [INFO] New Question: Who is Prashanth Reddy Varanganti
+2025-05-04 23:21:01,146 [INFO] HTTP Request: POST https://api.openai.com/v1/chat/completions "HTTP/1.1 200 OK"
+2025-05-04 23:21:02,168 [INFO] response: https://lite.duckduckgo.com/lite/ 200
+2025-05-04 23:21:13,843 [INFO] HTTP Request: POST https://api.openai.com/v1/chat/completions "HTTP/1.1 200 OK"
+2025-05-04 23:21:14,153 [INFO] HTTP Request: POST https://api.openai.com/v1/embeddings "HTTP/1.1 200 OK"
+2025-05-04 23:21:14,157 [INFO] Adding new content to vector store.
+2025-05-04 23:21:14,766 [INFO] HTTP Request: POST https://api.openai.com/v1/embeddings "HTTP/1.1 200 OK"
+2025-05-04 23:23:04,507 [INFO] Loading faiss.
+2025-05-04 23:23:04,529 [INFO] Successfully loaded faiss.
+2025-05-04 23:23:04,533 [INFO] Failed to load GPU Faiss: name 'GpuIndexIVFFlat' is not defined. Will not load constructor refs for GPU indexes. This is only an error if you're trying to use GPU Faiss.
+2025-05-04 23:23:04,810 [INFO] HTTP Request: GET http://127.0.0.1:7860/gradio_api/startup-events "HTTP/1.1 200 OK"
+2025-05-04 23:23:04,832 [INFO] HTTP Request: HEAD http://127.0.0.1:7860/ "HTTP/1.1 200 OK"
+2025-05-04 23:23:04,996 [INFO] HTTP Request: GET https://api.gradio.app/pkg-version "HTTP/1.1 200 OK"
+2025-05-04 23:23:05,134 [INFO] HTTP Request: GET https://api.gradio.app/v3/tunnel-request "HTTP/1.1 200 OK"
+2025-05-04 23:23:14,324 [INFO] New Question: Who is prashanth reddy varanganti
+2025-05-04 23:23:15,089 [INFO] HTTP Request: POST https://api.openai.com/v1/embeddings "HTTP/1.1 200 OK"
+2025-05-04 23:23:19,977 [INFO] HTTP Request: POST https://api.openai.com/v1/chat/completions "HTTP/1.1 200 OK"
+2025-05-04 23:23:19,992 [INFO] Answered using memory.
+2025-05-04 23:24:25,078 [INFO] Loading faiss.
+2025-05-04 23:24:25,097 [INFO] Successfully loaded faiss.
+2025-05-04 23:24:25,101 [INFO] Failed to load GPU Faiss: name 'GpuIndexIVFFlat' is not defined. Will not load constructor refs for GPU indexes. This is only an error if you're trying to use GPU Faiss.
+2025-05-04 23:24:25,362 [INFO] HTTP Request: GET http://127.0.0.1:7860/gradio_api/startup-events "HTTP/1.1 200 OK"
+2025-05-04 23:24:25,375 [INFO] HTTP Request: HEAD http://127.0.0.1:7860/ "HTTP/1.1 200 OK"
+2025-05-04 23:24:25,604 [INFO] HTTP Request: GET https://api.gradio.app/pkg-version "HTTP/1.1 200 OK"
+2025-05-04 23:24:25,681 [INFO] HTTP Request: GET https://api.gradio.app/v3/tunnel-request "HTTP/1.1 200 OK"
+2025-05-04 23:24:30,150 [INFO] New message: HI 
+2025-05-04 23:24:30,151 [INFO] New Question: HI 
+2025-05-04 23:24:30,998 [INFO] HTTP Request: POST https://api.openai.com/v1/embeddings "HTTP/1.1 200 OK"
+2025-05-04 23:24:32,099 [INFO] HTTP Request: POST https://api.openai.com/v1/chat/completions "HTTP/1.1 200 OK"
+2025-05-04 23:24:32,110 [INFO] Answered using memory.
+2025-05-04 23:24:39,429 [INFO] New message: What is todays date
+2025-05-04 23:24:39,430 [INFO] New Question: What is todays date
+2025-05-04 23:24:40,901 [INFO] HTTP Request: POST https://api.openai.com/v1/chat/completions "HTTP/1.1 200 OK"
+2025-05-04 23:25:04,899 [INFO] New message: what is today..?
+2025-05-04 23:25:04,900 [INFO] New Question: what is today..?
+2025-05-04 23:25:08,551 [INFO] HTTP Request: POST https://api.openai.com/v1/chat/completions "HTTP/1.1 200 OK"
+2025-05-04 23:25:31,300 [INFO] New message: what is todya
+2025-05-04 23:25:31,300 [INFO] New Question: what is todya
+2025-05-04 23:25:31,562 [INFO] HTTP Request: POST https://api.openai.com/v1/embeddings "HTTP/1.1 200 OK"
+2025-05-04 23:25:34,050 [INFO] HTTP Request: POST https://api.openai.com/v1/chat/completions "HTTP/1.1 200 OK"
+2025-05-04 23:25:34,053 [INFO] Answered using memory.
+2025-05-04 23:25:39,609 [INFO] New message: stoday 
+2025-05-04 23:25:39,610 [INFO] New Question: stoday 
+2025-05-04 23:25:41,626 [INFO] HTTP Request: POST https://api.openai.com/v1/chat/completions "HTTP/1.1 200 OK"
+2025-05-04 23:25:49,224 [INFO] New message: what is today
+2025-05-04 23:25:49,224 [INFO] New Question: what is today
+2025-05-04 23:25:51,902 [INFO] HTTP Request: POST https://api.openai.com/v1/chat/completions "HTTP/1.1 200 OK"
+2025-05-04 23:26:13,965 [INFO] New message: ok ,what do you mean by ai engineering..?
+2025-05-04 23:26:13,965 [INFO] New Question: ok ,what do you mean by ai engineering..?
+2025-05-04 23:26:14,805 [INFO] HTTP Request: POST https://api.openai.com/v1/embeddings "HTTP/1.1 200 OK"
+2025-05-04 23:26:29,245 [INFO] HTTP Request: POST https://api.openai.com/v1/chat/completions "HTTP/1.1 200 OK"
+2025-05-04 23:26:29,250 [INFO] Answered using memory.
+2025-05-04 23:27:17,887 [INFO] New message: Long note book
+2025-05-04 23:27:17,887 [INFO] New Question: Long note book
+2025-05-04 23:27:18,398 [INFO] HTTP Request: POST https://api.openai.com/v1/embeddings "HTTP/1.1 200 OK"
+2025-05-04 23:27:41,336 [INFO] HTTP Request: POST https://api.openai.com/v1/chat/completions "HTTP/1.1 200 OK"
+2025-05-04 23:27:41,341 [INFO] Answered using memory.
+2025-05-04 23:28:53,480 [INFO] Loading faiss.
+2025-05-04 23:28:53,499 [INFO] Successfully loaded faiss.
+2025-05-04 23:28:53,503 [INFO] Failed to load GPU Faiss: name 'GpuIndexIVFFlat' is not defined. Will not load constructor refs for GPU indexes. This is only an error if you're trying to use GPU Faiss.
+2025-05-04 23:28:53,707 [INFO] HTTP Request: GET http://127.0.0.1:7860/gradio_api/startup-events "HTTP/1.1 200 OK"
+2025-05-04 23:28:53,720 [INFO] HTTP Request: HEAD http://127.0.0.1:7860/ "HTTP/1.1 200 OK"
+2025-05-04 23:28:53,833 [INFO] HTTP Request: GET https://api.gradio.app/pkg-version "HTTP/1.1 200 OK"
+2025-05-04 23:28:54,010 [INFO] HTTP Request: GET https://api.gradio.app/v3/tunnel-request "HTTP/1.1 200 OK"
+2025-05-04 23:28:58,676 [INFO] New message: HI
+2025-05-04 23:28:58,677 [INFO] New Question: HI
+2025-05-04 23:28:59,676 [INFO] HTTP Request: POST https://api.openai.com/v1/embeddings "HTTP/1.1 429 Too Many Requests"
+2025-05-04 23:28:59,677 [INFO] Retrying request to /embeddings in 0.423430 seconds
+2025-05-04 23:29:00,509 [INFO] HTTP Request: POST https://api.openai.com/v1/embeddings "HTTP/1.1 429 Too Many Requests"
+2025-05-04 23:29:00,510 [INFO] Retrying request to /embeddings in 0.900636 seconds
+2025-05-04 23:29:01,926 [INFO] HTTP Request: POST https://api.openai.com/v1/embeddings "HTTP/1.1 429 Too Many Requests"
+2025-05-04 23:36:05,589 [INFO] Use pytorch device_name: mps
+2025-05-04 23:36:05,590 [INFO] Load pretrained SentenceTransformer: sentence-transformers/all-MiniLM-L6-v2
+2025-05-04 23:36:07,140 [WARNING] Xet Storage is enabled for this repo, but the 'hf_xet' package is not installed. Falling back to regular HTTP download. For better performance, install the package with: `pip install huggingface_hub[hf_xet]` or `pip install hf_xet`
+2025-05-04 23:36:12,372 [INFO] 📂 Loading existing vector store...
+2025-05-04 23:36:12,374 [INFO] Loading faiss.
+2025-05-04 23:36:12,556 [INFO] Successfully loaded faiss.
+2025-05-04 23:36:12,561 [INFO] Failed to load GPU Faiss: name 'GpuIndexIVFFlat' is not defined. Will not load constructor refs for GPU indexes. This is only an error if you're trying to use GPU Faiss.
+2025-05-04 23:36:12,891 [INFO] HTTP Request: GET http://127.0.0.1:7860/gradio_api/startup-events "HTTP/1.1 200 OK"
+2025-05-04 23:36:12,906 [INFO] HTTP Request: HEAD http://127.0.0.1:7860/ "HTTP/1.1 200 OK"
+2025-05-04 23:36:13,154 [INFO] HTTP Request: GET https://api.gradio.app/pkg-version "HTTP/1.1 200 OK"
+2025-05-04 23:36:13,189 [INFO] HTTP Request: GET https://api.gradio.app/v3/tunnel-request "HTTP/1.1 200 OK"
+2025-05-04 23:36:23,475 [INFO] User: Hey
+2025-05-04 23:36:23,475 [INFO] New Question: Hey
+2025-05-04 23:36:49,116 [INFO] Use pytorch device_name: mps
+2025-05-04 23:36:49,117 [INFO] Load pretrained SentenceTransformer: sentence-transformers/all-MiniLM-L6-v2
+2025-05-04 23:36:50,429 [INFO] 📂 Loading existing vector store...
+2025-05-04 23:36:50,437 [INFO] Loading faiss.
+2025-05-04 23:36:50,771 [INFO] Successfully loaded faiss.
+2025-05-04 23:36:50,776 [INFO] Failed to load GPU Faiss: name 'GpuIndexIVFFlat' is not defined. Will not load constructor refs for GPU indexes. This is only an error if you're trying to use GPU Faiss.
+2025-05-04 23:36:51,009 [INFO] HTTP Request: GET http://127.0.0.1:7860/gradio_api/startup-events "HTTP/1.1 200 OK"
+2025-05-04 23:36:51,024 [INFO] HTTP Request: HEAD http://127.0.0.1:7860/ "HTTP/1.1 200 OK"
+2025-05-04 23:36:51,121 [INFO] HTTP Request: GET https://api.gradio.app/pkg-version "HTTP/1.1 200 OK"
+2025-05-04 23:36:51,336 [INFO] HTTP Request: GET https://api.gradio.app/v3/tunnel-request "HTTP/1.1 200 OK"
+2025-05-04 23:36:55,939 [INFO] User: Hey
+2025-05-04 23:36:55,940 [INFO] New Question: Hey
+2025-05-04 23:37:44,053 [INFO] Use pytorch device_name: mps
+2025-05-04 23:37:44,054 [INFO] Load pretrained SentenceTransformer: sentence-transformers/all-MiniLM-L6-v2
+2025-05-04 23:37:45,359 [INFO] 📦 Initializing new vector store...
+2025-05-04 23:37:46,731 [INFO] Loading faiss.
+2025-05-04 23:37:46,997 [INFO] Successfully loaded faiss.
+2025-05-04 23:37:47,003 [INFO] Failed to load GPU Faiss: name 'GpuIndexIVFFlat' is not defined. Will not load constructor refs for GPU indexes. This is only an error if you're trying to use GPU Faiss.
+2025-05-04 23:37:47,293 [INFO] HTTP Request: GET http://127.0.0.1:7860/gradio_api/startup-events "HTTP/1.1 200 OK"
+2025-05-04 23:37:47,310 [INFO] HTTP Request: HEAD http://127.0.0.1:7860/ "HTTP/1.1 200 OK"
+2025-05-04 23:37:47,461 [INFO] HTTP Request: GET https://api.gradio.app/pkg-version "HTTP/1.1 200 OK"
+2025-05-04 23:37:47,614 [INFO] HTTP Request: GET https://api.gradio.app/v3/tunnel-request "HTTP/1.1 200 OK"
+2025-05-04 23:37:51,102 [INFO] User: HEy
+2025-05-04 23:37:51,103 [INFO] New Question: HEy
+2025-05-04 23:37:51,869 [INFO] Answered using memory.
+2025-05-04 23:38:10,664 [INFO] User: What US did to India..?
+2025-05-04 23:38:10,664 [INFO] New Question: What US did to India..?
+2025-05-04 23:38:12,346 [INFO] Memory insufficient. Falling back to real-time web search.
+2025-05-04 23:38:12,347 [INFO] Real-time query detected: 'What US did to India..?'
+2025-05-04 23:38:13,584 [INFO] response: https://html.duckduckgo.com/html 200
+2025-05-04 23:38:14,931 [INFO] Adding new content to vector store.
+2025-05-04 23:38:37,643 [INFO] User: What US did to india..?
+2025-05-04 23:38:37,644 [INFO] New Question: What US did to india..?
+2025-05-04 23:38:38,397 [INFO] Answered using memory.
+2025-05-04 23:47:16,450 [INFO] Use pytorch device_name: mps
+2025-05-04 23:47:16,451 [INFO] Load pretrained SentenceTransformer: sentence-transformers/all-MiniLM-L6-v2
+2025-05-04 23:47:17,944 [INFO] 📂 Loading existing vector store...
+2025-05-04 23:47:17,955 [INFO] Loading faiss.
+2025-05-04 23:47:18,165 [INFO] Successfully loaded faiss.
+2025-05-04 23:47:18,172 [INFO] Failed to load GPU Faiss: name 'GpuIndexIVFFlat' is not defined. Will not load constructor refs for GPU indexes. This is only an error if you're trying to use GPU Faiss.
+2025-05-04 23:47:18,692 [INFO] HTTP Request: GET https://api.gradio.app/pkg-version "HTTP/1.1 200 OK"
+2025-05-04 23:48:14,916 [INFO] Use pytorch device_name: mps
+2025-05-04 23:48:14,917 [INFO] Load pretrained SentenceTransformer: sentence-transformers/all-MiniLM-L6-v2
+2025-05-04 23:48:16,180 [INFO] 📂 Loading existing vector store...
+2025-05-04 23:48:16,186 [INFO] Loading faiss.
+2025-05-04 23:48:16,358 [INFO] Successfully loaded faiss.
+2025-05-04 23:48:16,362 [INFO] Failed to load GPU Faiss: name 'GpuIndexIVFFlat' is not defined. Will not load constructor refs for GPU indexes. This is only an error if you're trying to use GPU Faiss.
+2025-05-04 23:48:16,629 [INFO] HTTP Request: GET http://127.0.0.1:7860/gradio_api/startup-events "HTTP/1.1 200 OK"
+2025-05-04 23:48:16,649 [INFO] HTTP Request: HEAD http://127.0.0.1:7860/ "HTTP/1.1 200 OK"
+2025-05-04 23:48:16,708 [INFO] HTTP Request: GET https://api.gradio.app/pkg-version "HTTP/1.1 200 OK"
+2025-05-04 23:48:16,947 [INFO] HTTP Request: GET https://api.gradio.app/v3/tunnel-request "HTTP/1.1 200 OK"
+2025-05-04 23:49:06,536 [INFO] Use pytorch device_name: mps
+2025-05-04 23:49:06,536 [INFO] Load pretrained SentenceTransformer: sentence-transformers/all-MiniLM-L6-v2
+2025-05-04 23:49:08,344 [INFO] 📂 Loading existing vector store...
+2025-05-04 23:49:08,347 [INFO] Loading faiss.
+2025-05-04 23:49:08,532 [INFO] Successfully loaded faiss.
+2025-05-04 23:49:08,536 [INFO] Failed to load GPU Faiss: name 'GpuIndexIVFFlat' is not defined. Will not load constructor refs for GPU indexes. This is only an error if you're trying to use GPU Faiss.
+2025-05-04 23:49:08,755 [INFO] HTTP Request: GET http://127.0.0.1:7860/gradio_api/startup-events "HTTP/1.1 200 OK"
+2025-05-04 23:49:08,770 [INFO] HTTP Request: HEAD http://127.0.0.1:7860/ "HTTP/1.1 200 OK"
+2025-05-04 23:49:08,960 [INFO] HTTP Request: GET https://api.gradio.app/pkg-version "HTTP/1.1 200 OK"
+2025-05-04 23:49:09,077 [INFO] HTTP Request: GET https://api.gradio.app/v3/tunnel-request "HTTP/1.1 200 OK"
+2025-05-05 00:50:13,159 [INFO] Use pytorch device_name: mps
+2025-05-05 00:50:13,159 [INFO] Load pretrained SentenceTransformer: sentence-transformers/all-MiniLM-L6-v2
+2025-05-05 00:50:14,497 [INFO] 📂 Loading existing vector store...
+2025-05-05 00:50:14,503 [INFO] Loading faiss.
+2025-05-05 00:50:14,738 [INFO] Successfully loaded faiss.
+2025-05-05 00:50:14,744 [INFO] Failed to load GPU Faiss: name 'GpuIndexIVFFlat' is not defined. Will not load constructor refs for GPU indexes. This is only an error if you're trying to use GPU Faiss.
+2025-05-05 00:50:15,026 [INFO] HTTP Request: GET http://127.0.0.1:7860/gradio_api/startup-events "HTTP/1.1 200 OK"
+2025-05-05 00:50:15,044 [INFO] HTTP Request: HEAD http://127.0.0.1:7860/ "HTTP/1.1 200 OK"
+2025-05-05 00:50:15,174 [INFO] HTTP Request: GET https://api.gradio.app/pkg-version "HTTP/1.1 200 OK"
+2025-05-05 00:50:15,343 [INFO] HTTP Request: GET https://api.gradio.app/v3/tunnel-request "HTTP/1.1 200 OK"
+2025-05-05 00:50:39,936 [INFO] User: Hey , HI
+2025-05-05 00:50:39,936 [INFO] New Question: Hey , HI
+2025-05-05 00:50:41,481 [INFO] Answered using memory.
+2025-05-05 00:51:02,253 [INFO] User: WHt is th current staus of the indian army..?
+2025-05-05 00:51:02,254 [INFO] New Question: WHt is th current staus of the indian army..?
+2025-05-05 00:51:03,216 [INFO] Answered using memory.
+2025-05-05 00:51:43,085 [INFO] User: I cant understand why we use RAG
+2025-05-05 00:51:43,086 [INFO] New Question: I cant understand why we use RAG
+2025-05-05 00:51:44,973 [INFO] Answered using memory.
+2025-05-05 00:52:23,974 [INFO] Use pytorch device_name: mps
+2025-05-05 00:52:23,974 [INFO] Load pretrained SentenceTransformer: sentence-transformers/all-MiniLM-L6-v2
+2025-05-05 00:52:25,419 [INFO] 📂 Loading existing vector store...
+2025-05-05 00:52:25,423 [INFO] Loading faiss.
+2025-05-05 00:52:25,657 [INFO] Successfully loaded faiss.
+2025-05-05 00:52:25,661 [INFO] Failed to load GPU Faiss: name 'GpuIndexIVFFlat' is not defined. Will not load constructor refs for GPU indexes. This is only an error if you're trying to use GPU Faiss.
+2025-05-05 00:52:25,898 [INFO] HTTP Request: GET http://127.0.0.1:7860/gradio_api/startup-events "HTTP/1.1 200 OK"
+2025-05-05 00:52:25,911 [INFO] HTTP Request: HEAD http://127.0.0.1:7860/ "HTTP/1.1 200 OK"
+2025-05-05 00:52:26,117 [INFO] HTTP Request: GET https://api.gradio.app/pkg-version "HTTP/1.1 200 OK"
+2025-05-05 00:52:26,189 [INFO] HTTP Request: GET https://api.gradio.app/v3/tunnel-request "HTTP/1.1 200 OK"
+2025-05-05 00:52:30,627 [INFO] User: HEY
+2025-05-05 00:52:30,627 [INFO] New Question: HEY
+2025-05-05 00:52:31,308 [INFO] Answered using memory.
+2025-05-05 00:52:43,194 [INFO] User: I DONT WNAT THAT
+2025-05-05 00:52:43,195 [INFO] New Question: I DONT WNAT THAT
+2025-05-05 00:52:44,776 [INFO] Answered using memory.
+2025-05-05 00:56:20,576 [INFO] Use pytorch device_name: mps
+2025-05-05 00:56:20,576 [INFO] Load pretrained SentenceTransformer: sentence-transformers/all-MiniLM-L6-v2
+2025-05-05 00:56:21,926 [INFO] 📂 Loading existing vector store...
+2025-05-05 00:56:21,932 [INFO] Loading faiss.
+2025-05-05 00:56:22,197 [INFO] Successfully loaded faiss.
+2025-05-05 00:56:22,202 [INFO] Failed to load GPU Faiss: name 'GpuIndexIVFFlat' is not defined. Will not load constructor refs for GPU indexes. This is only an error if you're trying to use GPU Faiss.
+2025-05-05 00:56:22,435 [INFO] HTTP Request: GET http://127.0.0.1:7860/gradio_api/startup-events "HTTP/1.1 200 OK"
+2025-05-05 00:56:22,449 [INFO] HTTP Request: HEAD http://127.0.0.1:7860/ "HTTP/1.1 200 OK"
+2025-05-05 00:56:22,768 [INFO] HTTP Request: GET https://api.gradio.app/pkg-version "HTTP/1.1 200 OK"
+2025-05-05 00:56:22,768 [INFO] HTTP Request: GET https://api.gradio.app/v3/tunnel-request "HTTP/1.1 200 OK"
+2025-05-05 00:56:30,876 [INFO] User: Hey
+2025-05-05 00:56:30,877 [INFO] New Question: Hey
+2025-05-05 00:56:31,308 [INFO] No relevant memory found. Using web search.
+2025-05-05 00:56:31,309 [INFO] Real-time query detected: 'Hey'
+2025-05-05 00:56:32,679 [INFO] response: https://lite.duckduckgo.com/lite/ 200
+2025-05-05 00:56:33,208 [INFO] Adding new content to vector store.
+2025-05-05 00:56:54,640 [INFO] User: Hey, hi
+2025-05-05 00:56:54,641 [INFO] New Question: Hey, hi
+2025-05-05 00:56:54,796 [INFO] No relevant memory found. Using web search.
+2025-05-05 00:56:54,796 [INFO] Real-time query detected: 'Hey, hi'
+2025-05-05 00:56:55,660 [INFO] response: https://html.duckduckgo.com/html 200
+2025-05-05 00:56:56,238 [INFO] Adding new content to vector store.
+2025-05-05 01:02:28,907 [INFO] User: HEY , HI ASSISTANT
+2025-05-05 01:02:28,908 [INFO] New Question: HEY , HI ASSISTANT
+2025-05-05 01:02:30,049 [INFO] No relevant memory found. Using web search.
+2025-05-05 01:02:30,050 [INFO] Real-time query detected: 'HEY , HI ASSISTANT'
+2025-05-05 01:02:31,033 [INFO] response: https://html.duckduckgo.com/html 200
+2025-05-05 01:02:31,521 [INFO] Adding new content to vector store.
+2025-05-05 01:10:07,943 [INFO] Use pytorch device_name: mps
+2025-05-05 01:10:07,944 [INFO] Load pretrained SentenceTransformer: sentence-transformers/all-MiniLM-L6-v2
+2025-05-05 01:10:09,436 [INFO] 📂 Loading existing vector store...
+2025-05-05 01:10:09,439 [INFO] Loading faiss.
+2025-05-05 01:10:09,622 [INFO] Successfully loaded faiss.
+2025-05-05 01:10:09,627 [INFO] Failed to load GPU Faiss: name 'GpuIndexIVFFlat' is not defined. Will not load constructor refs for GPU indexes. This is only an error if you're trying to use GPU Faiss.
+2025-05-05 01:10:09,889 [INFO] HTTP Request: GET http://127.0.0.1:7860/gradio_api/startup-events "HTTP/1.1 200 OK"
+2025-05-05 01:10:09,904 [INFO] HTTP Request: HEAD http://127.0.0.1:7860/ "HTTP/1.1 200 OK"
+2025-05-05 01:10:10,025 [INFO] HTTP Request: GET https://api.gradio.app/pkg-version "HTTP/1.1 200 OK"
+2025-05-05 01:10:10,190 [INFO] HTTP Request: GET https://api.gradio.app/v3/tunnel-request "HTTP/1.1 200 OK"

requirements.txt

@@ -0,0 +1,4 @@
+python-dotenv
+langchain-core
+langchain-community
+faiss