{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "51aa96d8-9958-465c-8b0d-cd4f35584307",
   "metadata": {},
   "outputs": [],
   "source": [
    "from ollama import chat\n",
    "from ollama import ChatResponse\n",
    "\n",
    "# Define file path \n",
    "file_path = r'C:\\Users\\ASUS\\Downloads\\responses.txt'\n",
    "\n",
    "# First query and save\n",
    "response = chat(model='partai/dorna-llama3', messages=[\n",
    "    {'role': 'user', 'content': 'چرا آسمان آبیست؟?'},\n",
    "])\n",
    "with open(file_path, 'w', encoding='utf-8') as f:\n",
    "    f.write(response['message']['content'] + '\\n\\n')\n",
    "\n",
    "# Second query and append\n",
    "response = chat(model='partai/dorna-llama3', messages=[\n",
    "    {'role': 'user', 'content': 'چرا اینترنت قطع میشه؟'},\n",
    "])\n",
    "with open(file_path, 'a', encoding='utf-8') as f:\n",
    "    f.write(response['message']['content'] + '\\n\\n')\n",
    "\n",
    "print(f\"Responses saved to {file_path}\")\n",
    "\n",
    "Masih Moafi, [1/24/2025 11:57 PM]\n",
    "from ollama import chat, embeddings\n",
    "import numpy as np\n",
    "from sklearn.metrics.pairwise import cosine_similarity\n",
    "\n",
    "# 1. Load and chunk document\n",
    "with open(r'C:\\Users\\ASUS\\Downloads\\data.txt', 'r', encoding='utf-8') as f:\n",
    "    text = f.read()\n",
    "\n",
    "# Split with overlap\n",
    "chunk_size = 1000\n",
    "overlap = 200\n",
    "chunks = [text[i:i+chunk_size] for i in range(0, len(text), chunk_size - overlap)]\n",
    "\n",
    "# 2. Create embeddings using dorna-llama3\n",
    "chunk_embeddings = []\n",
    "for chunk in chunks:\n",
    "    response = embeddings(model='partai/dorna-llama3', prompt=chunk)\n",
    "    chunk_embeddings.append(response['embedding'])\n",
    "\n",
    "# 3. Context retrieval system\n",
    "def find_relevant_chunks(query, top_k=3):\n",
    "    # Generate query embedding\n",
    "    query_embed = embeddings(model='partai/dorna-llama3', prompt=query)['embedding']\n",
    "    \n",
    "    # Calculate similarities\n",
    "    scores = cosine_similarity([query_embed], chunk_embeddings)[0]\n",
    "    \n",
    "    # Return top chunks\n",
    "    best_indices = np.argsort(scores)[-top_k:][::-1]\n",
    "    return \"\\n---\\n\".join([chunks[i] for i in best_indices])\n",
    "\n",
    "# 4. RAG-enhanced chat function\n",
    "def rag_chat(query):\n",
    "    # Retrieve context\n",
    "    context = find_relevant_chunks(query)\n",
    "    \n",
    "    # Create augmented prompt\n",
    "    prompt = f\"\"\"Answer the question using this context:\n",
    "{context}\n",
    "\n",
    "Question: {query}\n",
    "Answer clearly and concisely in Persian:\"\"\"\n",
    "    \n",
    "    # Get response\n",
    "    response = chat(model='partai/dorna-llama3', messages=[\n",
    "        {'role': 'user', 'content': prompt}\n",
    "    ])\n",
    "    \n",
    "    return response['message']['content']\n",
    "\n",
    "# Example usage\n",
    "response = rag_chat(\"چرا اینترنت قطع میشود؟\")\n",
    "print(\"پاسخ:\", response)\n",
    "\n",
    "Masih Moafi, [1/24/2025 11:57 PM]\n",
    "from ollama import chat\n",
    "from ollama import ChatResponse\n",
    "\n",
    "# Define file path \n",
    "file_path = r'C:\\Users\\ASUS\\Downloads\\responses.txt'\n",
    "\n",
    "\n",
    "response = chat(model='partai/dorna-llama3', messages=[\n",
    "    {'role': 'user', 'content': 'چرا اینترنت قطع میشه؟'},\n",
    "])\n",
    "with open(file_path, 'a', encoding='utf-8') as f:\n",
    "    f.write(response['message']['content'] + '\\n\\n')\n",
    "\n",
    "print(f\"Responses saved to {file_path}\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}