init_model.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "OOOOOOOOOOO!\n",
      "OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO!\n",
      "“We are all a little crazy.”\n",
      "It’s been one year since I first heard about them. And every time I hear that name, I think about you.\n",
      "And when I do, I think about how lucky I am to have such a great friend like you.\n",
      "There is no doubt in my mind that we will be together forever.\n",
      "“I wish I could tell you that sometimes life just doesn’t make sense.”\n",
      "“I don’t know what I was thinking when I first met you, but it doesn’t matter now.”\n",
      "“It doesn’t matter how we met.”\n",
      "“I think about you all the time.”\n",
      "“When I hear your voice, I am transported back to a simpler time. A time before phones and Facebook.”\n",
      "“There is something about you that I can’t put my finger on.”\n",
      "“I don’t know what I was thinking when we first met, but it doesn’t matter now.”\n",
      "“I wish I could tell you that sometimes life just doesn’t make sense.”\n",
      "“When I hear your voice, I am taken back to a simpler time.”\n",
      "“It doesn’t matter how we first crossed paths.”\n",
      "“I don’t know what I was thinking when we first met, but it doesn’t matter now.”\n"
     ]
    }
   ],
   "source": [
    "# Example: reuse your existing OpenAI setup\n",
    "from openai import OpenAI\n",
    "\n",
    "# Point to the local server\n",
    "client = OpenAI(base_url=\"http://localhost:1234/v1\", api_key=\"lm-studio\")\n",
    "\n",
    "completion = client.chat.completions.create(\n",
    "  model=\"TheBloke/medalpaca-13B-GGUF/medalpaca-13b.Q3_K_S.gguf\",\n",
    "  messages=[\n",
    "    {\"role\": \"system\", \"content\": \"Always answer with yes, no, or maybe only\"},\n",
    "    {\"role\": \"user\", \"content\": \"Introduce yourself.\"}\n",
    "  ],\n",
    "  temperature=0.7,\n",
    ")\n",
    "\n",
    "print(completion.choices[0].message.content)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>pubid</th>\n",
       "      <th>question</th>\n",
       "      <th>context</th>\n",
       "      <th>long_answer</th>\n",
       "      <th>final_decision</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>21645374</td>\n",
       "      <td>Do mitochondria play a role in remodelling lac...</td>\n",
       "      <td>{'contexts': ['Programmed cell death (PCD) is ...</td>\n",
       "      <td>Results depicted mitochondrial dynamics in viv...</td>\n",
       "      <td>yes</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>16418930</td>\n",
       "      <td>Landolt C and snellen e acuity: differences in...</td>\n",
       "      <td>{'contexts': ['Assessment of visual acuity dep...</td>\n",
       "      <td>Using the charts described, there was only a s...</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9488747</td>\n",
       "      <td>Syncope during bathing in infants, a pediatric...</td>\n",
       "      <td>{'contexts': ['Apparent life-threatening event...</td>\n",
       "      <td>\"Aquagenic maladies\" could be a pediatric form...</td>\n",
       "      <td>yes</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>17208539</td>\n",
       "      <td>Are the long-term results of the transanal pul...</td>\n",
       "      <td>{'contexts': ['The transanal endorectal pull-t...</td>\n",
       "      <td>Our long-term study showed significantly bette...</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10808977</td>\n",
       "      <td>Can tailored interventions increase mammograph...</td>\n",
       "      <td>{'contexts': ['Telephone counseling and tailor...</td>\n",
       "      <td>The effects of the intervention were most pron...</td>\n",
       "      <td>yes</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>995</th>\n",
       "      <td>8921484</td>\n",
       "      <td>Does gestational age misclassification explain...</td>\n",
       "      <td>{'contexts': ['After 34 weeks gestation, summa...</td>\n",
       "      <td>Gestational age misclassification is an unlike...</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>996</th>\n",
       "      <td>16564683</td>\n",
       "      <td>Is there any interest to perform ultrasonograp...</td>\n",
       "      <td>{'contexts': ['To evaluate the accuracy of ult...</td>\n",
       "      <td>Sonography has no place in the diagnosis of un...</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>997</th>\n",
       "      <td>23147106</td>\n",
       "      <td>Is peak concentration needed in therapeutic dr...</td>\n",
       "      <td>{'contexts': ['We analyzed the pharmacokinetic...</td>\n",
       "      <td>These results suggest little need to use peak ...</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>998</th>\n",
       "      <td>21550158</td>\n",
       "      <td>Can autologous platelet-rich plasma gel enhanc...</td>\n",
       "      <td>{'contexts': ['This investigation assesses the...</td>\n",
       "      <td>The PRP group recorded reduced pain, swelling,...</td>\n",
       "      <td>yes</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999</th>\n",
       "      <td>17559449</td>\n",
       "      <td>Are sugars-free medicines more erosive than su...</td>\n",
       "      <td>{'contexts': ['The reduced use of sugars-conta...</td>\n",
       "      <td>Paediatric SF medicines were not more erosive ...</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1000 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        pubid                                           question  \\\n",
       "0    21645374  Do mitochondria play a role in remodelling lac...   \n",
       "1    16418930  Landolt C and snellen e acuity: differences in...   \n",
       "2     9488747  Syncope during bathing in infants, a pediatric...   \n",
       "3    17208539  Are the long-term results of the transanal pul...   \n",
       "4    10808977  Can tailored interventions increase mammograph...   \n",
       "..        ...                                                ...   \n",
       "995   8921484  Does gestational age misclassification explain...   \n",
       "996  16564683  Is there any interest to perform ultrasonograp...   \n",
       "997  23147106  Is peak concentration needed in therapeutic dr...   \n",
       "998  21550158  Can autologous platelet-rich plasma gel enhanc...   \n",
       "999  17559449  Are sugars-free medicines more erosive than su...   \n",
       "\n",
       "                                               context  \\\n",
       "0    {'contexts': ['Programmed cell death (PCD) is ...   \n",
       "1    {'contexts': ['Assessment of visual acuity dep...   \n",
       "2    {'contexts': ['Apparent life-threatening event...   \n",
       "3    {'contexts': ['The transanal endorectal pull-t...   \n",
       "4    {'contexts': ['Telephone counseling and tailor...   \n",
       "..                                                 ...   \n",
       "995  {'contexts': ['After 34 weeks gestation, summa...   \n",
       "996  {'contexts': ['To evaluate the accuracy of ult...   \n",
       "997  {'contexts': ['We analyzed the pharmacokinetic...   \n",
       "998  {'contexts': ['This investigation assesses the...   \n",
       "999  {'contexts': ['The reduced use of sugars-conta...   \n",
       "\n",
       "                                           long_answer final_decision  \n",
       "0    Results depicted mitochondrial dynamics in viv...            yes  \n",
       "1    Using the charts described, there was only a s...             no  \n",
       "2    \"Aquagenic maladies\" could be a pediatric form...            yes  \n",
       "3    Our long-term study showed significantly bette...             no  \n",
       "4    The effects of the intervention were most pron...            yes  \n",
       "..                                                 ...            ...  \n",
       "995  Gestational age misclassification is an unlike...             no  \n",
       "996  Sonography has no place in the diagnosis of un...             no  \n",
       "997  These results suggest little need to use peak ...             no  \n",
       "998  The PRP group recorded reduced pain, swelling,...            yes  \n",
       "999  Paediatric SF medicines were not more erosive ...             no  \n",
       "\n",
       "[1000 rows x 5 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "completion = client.chat.completions.create(\n",
    "  model=\"TheBloke/medalpaca-13B-GGUF/medalpaca-13b.Q3_K_S.gguf\",\n",
    "  messages=[\n",
    "    {\"role\": \"system\", \"content\": \"Always answer with yes, no, or maybe only\"},\n",
    "    {\"role\": \"user\", \"content\": \"Introduce yourself.\"}\n",
    "  ],\n",
    "  temperature=0.7,\n",
    ")\n",
    "\n",
    "print(completion.choices[0].message.content)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "# Load your DataFrame here\n",
    "# test_df = pd.read_parquet('path_to_your_file.pubmedqa-labeled.parquet')\n",
    "\n",
    "client = OpenAI(base_url=\"http://localhost:1234/v1\", api_key=\"lm-studio\")\n",
    "\n",
    "# Simulating the LLM response, replace this function with actual API calls\n",
    "def simulate_llm_response(question, context):\n",
    "    completion = client.chat.completions.create(\n",
    "    model=\"TheBloke/medalpaca-13B-GGUF/medalpaca-13b.Q4_K_M.gguf\",\n",
    "    messages=[\n",
    "    {\"role\": \"system\", \"content\": f\"Always answer with yes, no, or maybe only. Only have a one word response.\"},\n",
    "    {\"role\": \"user\", \"content\": f\"{context}. {question}. Only have a one word response with yes/no/maybe.\"}\n",
    "    ],\n",
    "    temperature=0.1,\n",
    "    max_tokens=5,\n",
    "    n=1\n",
    "    )\n",
    "    # This function should be replaced with the actual LLM request\n",
    "    # For simulation, we'll randomly return 'yes', 'no', or 'maybe'\n",
    "    return completion.choices[0].message.content"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "ename": "TypeError",
     "evalue": "object of type 'StringMethods' has no len()",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[22], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m \u001b[38;5;28;43mlen\u001b[39;49m\u001b[43m(\u001b[49m\u001b[43mtest_df\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mhead\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m1\u001b[39;49m\u001b[43m)\u001b[49m\u001b[43m[\u001b[49m\u001b[38;5;124;43m'\u001b[39;49m\u001b[38;5;124;43mquestion\u001b[39;49m\u001b[38;5;124;43m'\u001b[39;49m\u001b[43m]\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstr\u001b[49m\u001b[43m)\u001b[49m\n",
      "\u001b[0;31mTypeError\u001b[0m: object of type 'StringMethods' has no len()"
     ]
    }
   ],
   "source": [
    "test_df.head(1)['question']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'contexts': array(['Programmed cell death (PCD) is the regulated death of cells within an organism. The lace plant (Aponogeton madagascariensis) produces perforations in its leaves through PCD. The leaves of the plant consist of a latticework of longitudinal and transverse veins enclosing areoles. PCD occurs in the cells at the center of these areoles and progresses outwards, stopping approximately five cells from the vasculature. The role of mitochondria during PCD has been recognized in animals; however, it has been less studied during PCD in plants.',\n",
      "       'The following paper elucidates the role of mitochondrial dynamics during developmentally regulated PCD in vivo in A. madagascariensis. A single areole within a window stage leaf (PCD is occurring) was divided into three areas based on the progression of PCD; cells that will not undergo PCD (NPCD), cells in early stages of PCD (EPCD), and cells in late stages of PCD (LPCD). Window stage leaves were stained with the mitochondrial dye MitoTracker Red CMXRos and examined. Mitochondrial dynamics were delineated into four categories (M1-M4) based on characteristics including distribution, motility, and membrane potential (ΔΨm). A TUNEL assay showed fragmented nDNA in a gradient over these mitochondrial stages. Chloroplasts and transvacuolar strands were also examined using live cell imaging. The possible importance of mitochondrial permeability transition pore (PTP) formation during PCD was indirectly examined via in vivo cyclosporine A (CsA) treatment. This treatment resulted in lace plant leaves with a significantly lower number of perforations compared to controls, and that displayed mitochondrial dynamics similar to that of non-PCD cells.'],\n",
      "      dtype=object), 'labels': array(['BACKGROUND', 'RESULTS'], dtype=object), 'meshes': array(['Alismataceae', 'Apoptosis', 'Cell Differentiation',\n",
      "       'Mitochondria', 'Plant Leaves'], dtype=object), 'reasoning_required_pred': array(['y', 'e', 's'], dtype=object), 'reasoning_free_pred': array(['y', 'e', 's'], dtype=object)}\n",
      "Do mitochondria play a role in remodelling lace plant leaves during programmed cell death?\n",
      "The role of mit\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Initialize variables\n",
    "accuracies = []\n",
    "correct_count = 0\n",
    "\n",
    "# Iterate through the first 100 rows of the DataFrame\n",
    "for index, row in test_df.head(1).iterrows():\n",
    "    # Simulate sending request to the LLM and getting a response\n",
    "    print(row['context'])\n",
    "    print(row['question'])\n",
    "    response = simulate_llm_response(row['question'], row['context'])\n",
    "    print(response)\n",
    "    # Check if the response matches the final decision\n",
    "    if row['final_decision'] in response:\n",
    "        correct_count += 1\n",
    "    \n",
    "    # Every 10 values, calculate the accuracy and reset the count\n",
    "    if (index + 1) % 10 == 0:\n",
    "        accuracy = correct_count / 10\n",
    "        accuracies.append(accuracy)\n",
    "        correct_count = 0  # Reset the count for the next batch\n",
    "\n",
    "# Plotting the accuracies\n",
    "plt.plot(accuracies, marker='o', linestyle='-')\n",
    "plt.title('Accuracy of LLM Responses per 10 Questions')\n",
    "plt.xlabel('Batch (each batch represents 10 questions)')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.grid(True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "GROQ_KEY = \"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\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.9.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}