{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:58:21.040680Z",
     "start_time": "2024-05-16T13:58:19.777010Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:16.520137Z",
     "iopub.status.busy": "2025-02-20T15:27:16.519855Z",
     "iopub.status.idle": "2025-02-20T15:27:19.071188Z",
     "shell.execute_reply": "2025-02-20T15:27:19.070700Z",
     "shell.execute_reply.started": "2025-02-20T15:27:16.520121Z"
    }
   },
   "outputs": [],
   "source": [
    "import ast\n",
    "import os\n",
    "import shutil\n",
    "import sys\n",
    "from collections import defaultdict\n",
    "import json\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from preference_data_preparation_diff import *\n",
    "from sklearn.model_selection import train_test_split\n",
    "from suno_utils.utils.s3 import download_s3_files\n",
    "from suno_utils.utils.text import read_json, read_jsonl, write_json, write_jsonl\n",
    "from tqdm import tqdm\n",
    "import matplotlib.pyplot as plt\n",
    "from suno_utils.audio import Audio\n",
    "from suno_analytics.preference_helper import get_preference_counts\n",
    "\n",
    "pd.set_option(\"display.max_rows\", 500)\n",
    "pd.set_option(\"display.max_columns\", 500)\n",
    "pd.set_option(\"display.width\", 1000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:58:21.082172Z",
     "start_time": "2024-05-16T13:58:21.041926Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:19.071950Z",
     "iopub.status.busy": "2025-02-20T15:27:19.071691Z",
     "iopub.status.idle": "2025-02-20T15:27:19.076334Z",
     "shell.execute_reply": "2025-02-20T15:27:19.075989Z",
     "shell.execute_reply.started": "2025-02-20T15:27:19.071936Z"
    }
   },
   "outputs": [],
   "source": [
    "OUT_DATA_DIR = \"/app/suno/data/dpo/diff_v4_t7_comb/\"\n",
    "os.makedirs(OUT_DATA_DIR, exist_ok=True)\n",
    "NPZ_DIR = \"/app/suno/data/dpo/diff_v4\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:58:53.962528Z",
     "start_time": "2024-05-16T13:58:21.105919Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:19.076949Z",
     "iopub.status.busy": "2025-02-20T15:27:19.076749Z",
     "iopub.status.idle": "2025-02-20T15:27:20.727762Z",
     "shell.execute_reply": "2025-02-20T15:27:20.727256Z",
     "shell.execute_reply.started": "2025-02-20T15:27:19.076936Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preference data shape (116496, 96)\n",
      "unique users 25326\n"
     ]
    }
   ],
   "source": [
    "df = pd.read_pickle(\n",
    "    \"/home/tony/Data/Preference/up_v4/interesting_clips_up_u_4_20250219_full.pkl\"\n",
    ")  # , engine='python')\n",
    "print(\"Preference data shape\", df.shape)\n",
    "print(\"unique users\", df[\"user_id\"].nunique())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:20.728471Z",
     "iopub.status.busy": "2025-02-20T15:27:20.728281Z",
     "iopub.status.idle": "2025-02-20T15:27:20.731883Z",
     "shell.execute_reply": "2025-02-20T15:27:20.731513Z",
     "shell.execute_reply.started": "2025-02-20T15:27:20.728457Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "is_public\n",
      "False    108120\n",
      "True       8376\n",
      "Name: count, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print(df[\"is_public\"].value_counts())\n",
    "# # remove public for now cause fucking users\n",
    "# df = df[~df[\"is_public\"]]\n",
    "# print(df[\"is_public\"].value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:20.732511Z",
     "iopub.status.busy": "2025-02-20T15:27:20.732292Z",
     "iopub.status.idle": "2025-02-20T15:27:20.844820Z",
     "shell.execute_reply": "2025-02-20T15:27:20.844337Z",
     "shell.execute_reply.started": "2025-02-20T15:27:20.732498Z"
    }
   },
   "outputs": [],
   "source": [
    "df[\"upsample_clip_id\"] = df[\"metadata\"].apply(lambda x: x.get(\"upsample_clip_id\", \"\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:58:56.199480Z",
     "start_time": "2024-05-16T13:58:53.963687Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:20.846234Z",
     "iopub.status.busy": "2025-02-20T15:27:20.846053Z",
     "iopub.status.idle": "2025-02-20T15:27:21.316709Z",
     "shell.execute_reply": "2025-02-20T15:27:21.316183Z",
     "shell.execute_reply.started": "2025-02-20T15:27:20.846220Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "182545\n",
      "59663\n",
      "122882\n",
      "pre-downloaded df (116496, 97)\n",
      "downloaded df (116460, 97)\n",
      "vae downloaded df (116460, 97)\n"
     ]
    }
   ],
   "source": [
    "all_converted_paths = os.listdir(NPZ_DIR)\n",
    "print(len(all_converted_paths))\n",
    "\n",
    "converted_paths = set(\n",
    "    [f.replace(\".npz\", \"\") for f in all_converted_paths if \"vae\" not in f]\n",
    ")\n",
    "print(len(converted_paths))\n",
    "vae_converted_paths = set(\n",
    "    [f.replace(\"_vae.npz\", \"\") for f in all_converted_paths if \"vae\" in f]\n",
    ")\n",
    "print(len(vae_converted_paths))\n",
    "\n",
    "print(\"pre-downloaded df\", df.shape)\n",
    "df[df[\"upsample_clip_id\"].isin(converted_paths)].shape\n",
    "df = df[df[\"upsample_clip_id\"].isin(converted_paths)].copy()\n",
    "print(\"downloaded df\", df.shape)\n",
    "df[df[\"s3_id\"].isin(vae_converted_paths)].shape\n",
    "df = df[df[\"s3_id\"].isin(vae_converted_paths)].copy()\n",
    "print(\"vae downloaded df\", df.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:58:56.467253Z",
     "start_time": "2024-05-16T13:58:56.207647Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:21.317394Z",
     "iopub.status.busy": "2025-02-20T15:27:21.317210Z",
     "iopub.status.idle": "2025-02-20T15:27:21.340480Z",
     "shell.execute_reply": "2025-02-20T15:27:21.340115Z",
     "shell.execute_reply.started": "2025-02-20T15:27:21.317380Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "is_up\n",
       "True    116460\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[\"is_up\"] = df[\"model_name\"].str.contains(\"up\")\n",
    "df[\"is_up\"].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# LET's do the data prep"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:58:56.592883Z",
     "start_time": "2024-05-16T13:58:56.470781Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:21.341121Z",
     "iopub.status.busy": "2025-02-20T15:27:21.340908Z",
     "iopub.status.idle": "2025-02-20T15:27:21.417572Z",
     "shell.execute_reply": "2025-02-20T15:27:21.417079Z",
     "shell.execute_reply.started": "2025-02-20T15:27:21.341107Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "preference  model_name     \n",
      "False       chirp-v4-up-u-4    58230\n",
      "True        chirp-v4-up-u-4    58230\n",
      "Name: count, dtype: int64\n",
      "(116460, 98)\n",
      "(116460, 98)\n"
     ]
    }
   ],
   "source": [
    "## for 13b this is easy for now\n",
    "print(df.groupby([\"preference\"])[\"model_name\"].value_counts())\n",
    "print(df.shape)\n",
    "df = df[df[\"model_name\"].isin([\"chirp-v4-up-u-4\"])]\n",
    "print(df.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:58:56.909539Z",
     "start_time": "2024-05-16T13:58:56.595736Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:21.418247Z",
     "iopub.status.busy": "2025-02-20T15:27:21.418072Z",
     "iopub.status.idle": "2025-02-20T15:27:21.515223Z",
     "shell.execute_reply": "2025-02-20T15:27:21.514722Z",
     "shell.execute_reply.started": "2025-02-20T15:27:21.418233Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(116460, 98)\n",
      "(116460, 98)\n",
      "preference  model_name     \n",
      "False       chirp-v4-up-u-4    58230\n",
      "True        chirp-v4-up-u-4    58230\n",
      "Name: count, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print(df.shape)\n",
    "df = df[\n",
    "    df[\"request_id\"].isin(\n",
    "        df[\"request_id\"].value_counts().index[df[\"request_id\"].value_counts() == 2]\n",
    "    )\n",
    "]\n",
    "print(df.shape)\n",
    "print(df.groupby([\"preference\"])[\"model_name\"].value_counts())\n",
    "assert df.shape[0] == df[\"request_id\"].nunique() * 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:21.515894Z",
     "iopub.status.busy": "2025-02-20T15:27:21.515720Z",
     "iopub.status.idle": "2025-02-20T15:27:23.430033Z",
     "shell.execute_reply": "2025-02-20T15:27:23.429513Z",
     "shell.execute_reply.started": "2025-02-20T15:27:21.515881Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total pair quality scores: 61011\n",
      "Total unpacked pair quality scores: 712498\n"
     ]
    }
   ],
   "source": [
    "with open(\"/home/tony/Data/Preference/up_v4/full_pair_quality.json\", \"r\") as file:\n",
    "    full_pair_quality = json.load(file)\n",
    "print(\"Total pair quality scores:\", len(full_pair_quality))\n",
    "\n",
    "unpacked_pair_quality = {}\n",
    "for request_id, pairs_of_qualities in full_pair_quality.items():\n",
    "    for clip_id, pair_quality in pairs_of_qualities.items():\n",
    "        unpacked_pair_quality[clip_id] = pair_quality\n",
    "print(\"Total unpacked pair quality scores:\", len(unpacked_pair_quality))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:23.430715Z",
     "iopub.status.busy": "2025-02-20T15:27:23.430538Z",
     "iopub.status.idle": "2025-02-20T15:27:24.143994Z",
     "shell.execute_reply": "2025-02-20T15:27:24.143486Z",
     "shell.execute_reply.started": "2025-02-20T15:27:23.430701Z"
    }
   },
   "outputs": [],
   "source": [
    "def get_audio_quality_measures(s3_id):\n",
    "    audio_quality = unpacked_pair_quality.get(s3_id, [])\n",
    "    if not audio_quality:\n",
    "        return [None for _ in range(11)]\n",
    "    return [\n",
    "        float(audio_quality[\"pref\"]),\n",
    "        float(audio_quality[\"shimmer_score\"]),\n",
    "        float(audio_quality[\"loudness_factor\"]),\n",
    "        audio_quality[\"spectral_character\"],\n",
    "        float(audio_quality[\"spectral_centroid\"]),\n",
    "        float(audio_quality[\"bass_ratio\"]),\n",
    "        float(audio_quality[\"mid_ratio\"]),\n",
    "        float(audio_quality[\"high_ratio\"]),\n",
    "        float(audio_quality[\"stereo_width\"]),\n",
    "        int(audio_quality[\"total_clips\"]),\n",
    "        float(audio_quality[\"clips_per_second\"]),\n",
    "    ]\n",
    "\n",
    "\n",
    "df[\n",
    "    [\n",
    "        \"pair_quality\",\n",
    "        \"total_shimmer_score\",\n",
    "        \"loudness_factor\",\n",
    "        \"spectral_character\",\n",
    "        \"spectral_centroid\",\n",
    "        \"bass_ratio\",\n",
    "        \"mid_ratio\",\n",
    "        \"high_ratio\",\n",
    "        \"stereo_width\",\n",
    "        \"total_clips\",\n",
    "        \"clips_per_second\",\n",
    "    ]\n",
    "] = pd.DataFrame(df[\"s3_id\"].apply(get_audio_quality_measures).tolist(), index=df.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:24.144692Z",
     "iopub.status.busy": "2025-02-20T15:27:24.144517Z",
     "iopub.status.idle": "2025-02-20T15:27:24.351615Z",
     "shell.execute_reply": "2025-02-20T15:27:24.351098Z",
     "shell.execute_reply.started": "2025-02-20T15:27:24.144678Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(116460, 109)\n",
      "(115538, 109)\n",
      "(115538, 109)\n"
     ]
    }
   ],
   "source": [
    "print(df.shape)\n",
    "df = df.dropna(subset=[\n",
    "        \"pair_quality\",\n",
    "        \"total_shimmer_score\",\n",
    "        \"loudness_factor\",\n",
    "        \"spectral_character\",\n",
    "        \"spectral_centroid\",\n",
    "        \"bass_ratio\",\n",
    "        \"mid_ratio\",\n",
    "        \"high_ratio\",\n",
    "        \"stereo_width\",\n",
    "        \"total_clips\",\n",
    "        \"clips_per_second\",\n",
    "    ])\n",
    "print(df.shape)\n",
    "df = df[\n",
    "    df[\"request_id\"].isin(\n",
    "        df[\"request_id\"].value_counts().index[df[\"request_id\"].value_counts() == 2]\n",
    "    )\n",
    "]\n",
    "print(df.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:59:36.043975Z",
     "start_time": "2024-05-16T13:58:56.910958Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:24.352323Z",
     "iopub.status.busy": "2025-02-20T15:27:24.352137Z",
     "iopub.status.idle": "2025-02-20T15:27:24.367277Z",
     "shell.execute_reply": "2025-02-20T15:27:24.366861Z",
     "shell.execute_reply.started": "2025-02-20T15:27:24.352310Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "unique_requests 57769\n"
     ]
    }
   ],
   "source": [
    "# Let's use the old selection for now -- for quality assurance\n",
    "# expand the metadata columns -- this takes forever...~ 6 mins\n",
    "# test_slice = df[\"metadata\"].apply(lambda x: ast.literal_eval(str(x)))\n",
    "# test_slice = df[\"metadata\"]  # .apply(lambda x: custom_parse(x))\n",
    "# test_slice_series = test_slice.apply(pd.Series)\n",
    "# df = pd.concat([df, test_slice_series], axis=1, join=\"inner\")\n",
    "print(\"unique_requests\", df[\"request_id\"].nunique())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:59:40.799375Z",
     "start_time": "2024-05-16T13:59:36.394236Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:24.367991Z",
     "iopub.status.busy": "2025-02-20T15:27:24.367749Z",
     "iopub.status.idle": "2025-02-20T15:27:24.642129Z",
     "shell.execute_reply": "2025-02-20T15:27:24.641631Z",
     "shell.execute_reply.started": "2025-02-20T15:27:24.367977Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "good_continue_at\n",
      "True    115538\n",
      "Name: count, dtype: int64\n",
      "\n",
      " Check some basics... \n",
      " preference\n",
      "False    57769\n",
      "True     57769\n",
      "Name: count, dtype: int64 model_name\n",
      "chirp-v4-up-u-4    115538\n",
      "Name: count, dtype: int64 preference  model_name     \n",
      "False       chirp-v4-up-u-4    57769\n",
      "True        chirp-v4-up-u-4    57769\n",
      "Name: count, dtype: int64\n",
      "task\n",
      "upsample    115538\n",
      "Name: count, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "df = df.loc[:, ~df.columns.duplicated()].copy()\n",
    "# get the original duration of the clips, if they are concacted\n",
    "df[\"original_duration_s\"] = df[\"total_start_s\"] + df[\"duration\"]\n",
    "# classify the continue at behavoirs by the duration choice\n",
    "audio_prompt_id_to_continue_at = {}\n",
    "for _, row in df[~df[\"continued_parent\"].isna()].iterrows():\n",
    "    audio_prompt_id = row[\"continued_parent\"]\n",
    "    if audio_prompt_id not in audio_prompt_id_to_continue_at:\n",
    "        audio_prompt_id_to_continue_at[audio_prompt_id] = row[\"continue_at\"]\n",
    "    else:\n",
    "        # pick the max\n",
    "        audio_prompt_id = max(\n",
    "            audio_prompt_id_to_continue_at[audio_prompt_id], row[\"continue_at\"]\n",
    "        )\n",
    "print(len(audio_prompt_id_to_continue_at))\n",
    "df[\"has_continue_and_start_continue_at\"] = df[\"s3_id\"].apply(\n",
    "    lambda x: audio_prompt_id_to_continue_at.get(x)\n",
    ")\n",
    "# we want continue at to be at most of the clip...\n",
    "df[\"good_continue_at\"] = (\n",
    "    (df[\"has_continue_and_start_continue_at\"] / df[\"duration\"]) > 0.9\n",
    ") | df[\"has_continue_and_start_continue_at\"].isna()\n",
    "print(df[\"good_continue_at\"].value_counts())\n",
    "\n",
    "\n",
    "print(\n",
    "    \"\\n Check some basics... \\n\",\n",
    "    df[\"preference\"].value_counts(),\n",
    "    df[\"model_name\"].value_counts(),\n",
    "    df.groupby([\"preference\"])[\"model_name\"].value_counts(),\n",
    ")\n",
    "\n",
    "df = df.sort_values(by=[\"request_id\", \"preference\"])\n",
    "df[\"duration_rel_diff\"] = df[\"duration\"].diff()\n",
    "df[\"play_rel_diff\"] = df[\"reaction_play_count\"].diff()\n",
    "print(df[\"task\"].value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:24.642814Z",
     "iopub.status.busy": "2025-02-20T15:27:24.642638Z",
     "iopub.status.idle": "2025-02-20T15:27:24.762521Z",
     "shell.execute_reply": "2025-02-20T15:27:24.762088Z",
     "shell.execute_reply.started": "2025-02-20T15:27:24.642800Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "pos_diff_preference\n",
       "1.0    39854\n",
       "2.0    17915\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = df.sort_values(by=[\"request_id\", \"preference\", \"diff_preference\"])\n",
    "df[\"pos_diff_preference\"] = df[\"diff_preference\"].diff()\n",
    "df[df[\"preference\"]][\"pos_diff_preference\"].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:24.763197Z",
     "iopub.status.busy": "2025-02-20T15:27:24.763019Z",
     "iopub.status.idle": "2025-02-20T15:27:25.050072Z",
     "shell.execute_reply": "2025-02-20T15:27:25.049652Z",
     "shell.execute_reply.started": "2025-02-20T15:27:24.763184Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjkAAAHNCAYAAAD8AGr/AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8hTgPZAAAACXBIWXMAAA9hAAAPYQGoP6dpAACEXElEQVR4nO3dd3gUZdfA4d/uJiGBFNLoPZCgkkAABUIAKVJElKZiAxEpggjSQaQ3EZTeEaUooICfWHgRsYBUFQi99xKSQBppm935/ohZWRIgy252k51zXxfXa2ZnZ89zTt7kZGaeZzSKoigIIYQQQjgZraMDEEIIIYTID9LkCCGEEMIpSZMjhBBCCKckTY4QQgghnJI0OUIIIYRwStLkCCGEEMIpSZMjhBBCCKckTY4QQgghnJI0OUIIIYRwStLkONDcuXN54YUXHB2GEEII4ZSkybGBuXPnEhISYvavdevWZvuEhISwbds2u8bUunVratWqxZNPPsmbb77JoUOH8vz+JUuWEBISwuTJk822p6enM378eOrVq0d4eDj9+/cnNjbW1uELIYQQVnNxdADOolq1aqxYscL0tU6nc2A0UKlSJcaMGUP58uVJS0vj888/56233uLnn3/Gz8/vge+Niopi7dq1hISE5HhtypQp/P7778yaNQsvLy8mTpzIu+++y9q1a/NrKEIIIcQjkTM5NqLT6QgMDDT9u7uRaNasGQD9+vUjJCTE9HW2b7/9lmbNmlGnTh3ef/99kpOTrY6nXbt2REREUL58eapVq8bIkSNJTk7m5MmTD3zfnTt3GDp0KJMmTcLHx8fstaSkJDZs2MCIESNo0KABNWrUYMqUKRw4cICDBw9aHbMQQghhS9Lk2MjFixeJjIykefPmDB48mGvXrple++abbwCYOnUqO3fuNH0NcOnSJX755RcWLVrE4sWL2b9/P0uXLjW9vnHjxlzPqFgiIyODdevW4eXl9dBjTZgwgSZNmhAREZHjtSNHjqDX681eCwoKokyZMtLkCCGEKHDkcpUNhIWFMXXqVCpXrkxMTAzz58/ntddeY/PmzXh6eprO6nh7exMYGGj2XkVRmDp1Kp6engA8//zz7N69m/fffx8ALy8vKleu/Ehx/frrrwwaNIjU1FQCAwP57LPPHnip6ocffuDYsWNmTdjdYmNjcXV1xdvb22y7v78/MTExjxSjEEIIkV/kTI4NNGnShDZt2lC9enUaNWrEkiVLSExM5Keffnroe8uWLWtqcABKlChBXFyc6etnnnmGLVu23Pf93333HeHh4aZ/f/31l+m1evXq8e2337J27VoaNWrEwIEDzY59t+vXrzN58mQ+/vhjihQpkpdhCyGEEAWanMnJB97e3lSqVIlLly49dF8Xl5wlUBQlz5/VrFkzatasafq6ZMmSpv8uWrQoFStWpGLFitSqVYuWLVvyzTff0Lt37xzHOXr0KHFxcXTs2NG0zWAwsH//ftasWcPhw4cJCAhAr9eTmJhodjYnLi4uxxkqIYQQwtGkyckHd+7c4fLly2a/+F1dXTEYDDb/LE9PT7MzQQ9iNBrJyMjI9bX69euzefNms20jR46kSpUq9OzZE51OR40aNXB1dWX37t20atUKgHPnznHt2jVq1apl1TiEEEIIW5MmxwY++ugjmjZtSpkyZbh58yZz585Fq9Xy3HPPmfYpW7Ysu3fvpnbt2ri5ueWYuXQ/P//8MzNnznzgJat7paSksGjRIpo1a0ZgYCC3b99mzZo1REdHm63f061bN5555hlef/11PD09CQ4ONjtO0aJFKV68uGm7l5cXnTp1Ytq0afj4+ODp6cmkSZMIDw+XJkcIIUSBI02ODdy4cYNBgwYRHx+Pn58fderUYf369WY3+Q4fPpxp06bx9ddfU7JkSbZv356nYyclJXH+/HmL4tHpdJw7d45NmzZx+/ZtihcvTmhoKGvWrKFatWqm/S5fvszt27ctOvaoUaPQarW89957ZGRkEBkZydixYy06hhBCCGEPGsWSG0CEEEIIIQoJmV0lhBBCCKckTY4QQgghnJI0OUIIIYRwStLkCCGEEMIpSZNTwI0YMYK+ffs6OgwhhBCi0JEmxwb2799Pnz59iIyMJCQkhG3btuXYR1EUZs+eTWRkJGFhYbz55ptcuHDB9PqVK1cICQnh+PHjdot73bp1vPHGG9SuXZuQkBASExMf+p7k5GQmT55M06ZNCQsLo0uXLkRFRZntExISkuu/ZcuW5ddQhBBCiBykybGBlJQUQkJCHrhezNKlS1m1ahXjxo1j/fr1eHh40KNHD9LT0+0YqbnU1FQaNWpEnz598vye0aNHs2vXLqZPn87mzZtp2LAh3bt3Jzo62rTPzp07zf5NmTIFjUZjWiVZCCGEsAdZJ8fGQkJCmD9/Pi1atDBtUxSFRo0a0b17d3r06AFkLfIXERHBtGnTaNu2LSEhIWbHeeqpp1i1ahUjRowgMTGROnXqsGLFCvR6Pc8++yyjRo3C1dXVJjHv3buXrl27sn///hxPGL9bWloatWvXZsGCBTz99NOm7R07dqRRo0amJ6ffq2/fvty5c4cvvvjCJvEKIYQQeSFncuzgypUrxMTEEBERYdrm5eVFzZo1OXDgAABff/01AJ9//jk7d+5k7ty5pn337t3LpUuX+OKLL5g2bRqbNm1i06ZNptfnzp1Ls2bN8n0cmZmZGAyGHE8pL1KkCP/880+u74mNjeX333+nc+fO+R6fEEIIcTdpcuwgJiYGAH9/f7Pt/v7+xMbGApgeAVG8eHECAwMpXry4aT8fHx/GjBlDUFAQTZs2pUmTJuzevdv0uq+vL+XLl8/nUWQ9DDQ8PJwFCxYQHR2NwWDg//7v/zh48CA3b97M9T2bNm2iWLFitGzZMt/jE0IIIe4mTU4hULVqVXQ6nenrwMBA4uLiTF+//vrrD7wUtGjRIsLDw03/rl279sixTJ8+HUVRaNy4MaGhoaxatYq2bdui1eb+rbRhwwbatWuX4+yPEEIIkd/kAZ12EBgYCEBcXBwlSpQwbY+Li6N69eoPfb+Li3mZNBoNltxK1aVLF9q0aWP6+u4YLFWhQgVWr15NSkoKycnJlChRgoEDB+Z6Jumvv/7i/PnzzJo165E/TwghhHhUcibHDsqVK0dgYKDZJabk5GQOHTpEeHg4gOkmYoPBYPPPL168OBUrVjT9u7dpehRFixalRIkSJCQksHPnTpo3b55jn2+++YYnnngiT42cEEIIYWvS5NjAnTt3OH78uGmNmytXrnD8+HHTZSGNRkPXrl1ZuHAhv/zyCydPnmTYsGGUKFHCNAvL398fd3d3duzYQWxsLElJSXn+/NWrV9OtWzeL446JieH48eNcunQJgFOnTnH8+HHi4+NN+3Tr1o3Vq1ebvt6xYwd//PEHly9f5s8//6Rr165UqVKFjh07mh07OTmZLVu28OKLL1oclxBCCGEL0uTYwJEjR2jfvj3t27cHYOrUqbRv3545c+aY9unZsyevv/46Y8aMoXPnzqSkpLBs2TLTvSouLi6MHj2adevW0ahRI4tWOb59+zaXL1+2OO61a9fSvn17Ro8eDcBrr71G+/bt2b59u2mfy5cvc/v2bdPXSUlJTJgwgTZt2jB8+HDq1KnD8uXLc0xn/+GHH1AUheeee85s+9y5c/n9998fGtvGjRsJCQnhypUrFo9LFDwhISFmMwbvV99ly5bRvHlzHnvsMV544QUga1bf9OnTadKkCdWrV5cVwFVi7ty5hISEcOvWLUeHIgoxWSdH2ERsbCz79+8nOjqalJQU3N3d8fPzo3LlytSsWdO039y5cwkLC6NJkyYPPN7GjRsZOXIkv/zyC+XKlcvv8EU+CwkJ4d1336V///5A7vXduXMnPXr04Pnnn6dhw4b4+vrSpEkT1q1bx5gxY+jWrRuPP/44ZcqU4amnnnLkcFRLr9ezePFiNm3aRHR0NCVLlqRTp0706tXL7DJ49tpbuVm3bh21atUyfb1o0SKqVq1qtrYYZP2smDdvHrt37zbNPrVWdHQ0H3/8MYcPH+bmzZvodDoqVapk+gNPo9E88P1RUVF8++237N27l6tXr1K8eHFq1qzJwIEDqVy5co79z549y5QpU/jnn39wdXWlSZMmjBw50mbjEQ8nNx4Lq12/fp1Nmzbh5eXFE088QdGiRUlOTubGjRscOnTIrMnJqxdeeIG2bdvi5uaWDxELR8utvnv27EGr1TJ58uQc20uWLMmoUaMcEaq4y9ChQ9myZQudOnWiRo0aHDp0iNmzZ3P9+nUmTpyYY/833niD0NBQs20VKlQw+3rx4sW0atUqR5OTH27fvk10dDStW7emdOnSZGZm8ueffzJixAjOnz/PoEGDHvj+ZcuW8c8//9C6dWtCQkKIiYlhzZo1dOzYkXXr1hEcHGza98aNG7z22mt4eXnx/vvvk5KSwmeffcapU6f4+uuv5WebnUiTI6z2119/4ebmxksvvZRjqnhKSsojHVOn05lNmy/sMjMzMRqNBfYHm9FoRK/X222qf271jYuLw93dPUeO4uLiHrgSt6UURSE9PR13d3ebHbMgO3HihE1u/o+KiuKnn36ib9++DBgwAIBXXnkFX19fVqxYwWuvvZbjc+rWrUvr1q2t/mxbqV69OqtWrTLb9vrrr9OnTx9WrVrFgAEDHvhz580332TGjBlm36PPPvss7dq1Y8mSJcyYMcO0fdGiRaSmprJx40bKlCkDQFhYGN27d2fTpk28/PLLNh6dyI3ckyOslpCQgJ+fX66/IIsWLZrre86dO8eXX37JggUL2LBhAzdu3DB7Pbd7Npo1a0bv3r3Zu3cvHTt2JCwsjHbt2rF3714Atm7dSrt27QgNDaVjx44cO3bM7JgjRowwrRPUu3dvwsPDadSoEWvWrAHg5MmTdO3alVq1atG0aVM2b96cI+7ExEQmT55MkyZNqFGjBs888wxLlizBaDSa9sl+2Ory5cv5/PPPadGiBaGhoZw9e/a+Ofzzzz955ZVXqFu3LuHh4bRq1YpPPvnEbJ/09HTmzp1Lq1atCA0NJTIyknfffdd04zhkNZXTpk0zxdeqVSuWL1+eY8mBkJAQJkyYwHfffUfbtm0JDQ1lx44dQNYp/ZEjRxIREUGNGjVo27Yt33zzzX1jv1tGRgZTpkyhfv36hIeH06dPnxy1hZz1DQkJYePGjabnwGV/HRISwt69ezl9+rRpe3a9jUYjn3/+uSn+iIgIxowZQ0JCgtlnZX/f7Nixw/R9s3btWsDyeq5bt44WLVpQo0YNOnXqlOPhtJB1iWLAgAHUr1+fsLAwWrVqxaeffmq2jzU5ttQLL7xA586dWbt2LcnJyY98nL///huAtm3bmm1/9tlnURSFn376Kdf3JScnk5mZmetrISEhpKSksGnTJlN9R4wYYbZPUlISI0aMoG7dutSpU4eRI0eSmpr6yOPITdmyZUlNTUWv1z9wv9q1a+dowitVqkS1atU4d+6c2fatW7fy9NNPmxocgIiICCpVqnTfXAnbkzM5wmpeXl7cuHGDuLi4HKs65+batWucPXuW0NBQXF1dOX36NPv27aNx48ZmKz3n5uLFiwwePJguXbrw/PPP89lnn9GnTx/Gjx/Pp59+yiuvvALAkiVLGDhwIFu2bDFbqNBgMNCzZ0/q1q3LkCFD2Lx5MxMmTMDDw4NPP/2Udu3a0bJlS9auXcvw4cOpVauWaQ2g1NRUXn/9daKjo+nSpQulS5fmwIEDfPLJJ8TExPDBBx+Yxbpx40bS09N56aWXcHNzw8fHJ9cxnT59mt69exMSEsJ7772Hm5sbFy9eNHtUhsFgoHfv3uzevZu2bdvStWtX7ty5w59//smpU6eoUKECiqLwzjvvsHfvXjp37sxjjz3Gjh07mD59OtHR0Tku9+zZs4effvqJ1157DV9fX8qWLUtsbCwvvfQSGo2G1157DT8/P/744w8++OADkpOTefPNNx9Ynw8++IDvvvuO5557jtq1a7Nnzx569er1wPdA1iKT69evJyoqikmTJgHw+OOPM336dBYtWkRKSorpUkJQUBAAY8aMYdOmTXTs2JE33niDK1eusGbNGo4dO8ZXX31ldjP8+fPnGTx4MC+//DIvvfQSlStXtrie33//PXfu3OHll19Go9GwbNky+vfvz7Zt20yfdeLECV577TVcXFx4+eWXKVu2LJcuXWL79u2mZ7tZm2NLTZw4kQ0bNjB27FimTZtG69at6dy5M3Xr1rXoOBkZGQA5/pjx8PAAsiZg3GvkyJGkpKSg0+moU6cOw4YNM7t8NX36dEaPHk1YWBgvvfQSkPNy1sCBAylXrhyDBg3i2LFjfP311/j5+TF06FCL4r9bWloaKSkppKSksH//fjZu3EitWrUe6eyeoijExsZSrVo107bo6Gji4uKoUaNGjv3DwsL4448/Hjl2YSFFFGivv/66MmnSJEeH8UAXL15U5s2bp8ybN0/5+uuvlZ07dyoXL15UMjMzc+w7Z84cZf78+Up8fLxpW2xsrLJhwwblm2++MW3bsGGDEhwcrFy+fNm0rWnTpkpwcLDyzz//mLbt2LFDCQ4OVsLCwpSrV6+atq9du1YJDg5W9uzZY9o2fPhwJTg4WFm0aJFpW0JCghIWFqaEhIQoP/zwg2n72bNnleDgYGXOnDmmbfPnz1dq1aqlnD9/3mxMM2bMUB577DHl2rVriqIoyuXLl5Xg4GCldu3aSlxc3EPzt2LFCiU4OPiB+37zzTdKcHCwsmLFihyvGY1GRVEU5eeff1aCg4OVBQsWmL3ev39/JSQkRLl48aJpW3BwsFK9enXl9OnTZvuOGjVKadiwoXLr1i2z7e+//75Sp04dJTU19b4xHj9+XAkODlbGjRtntn3QoEE5cplbfYcPH67UqlUrx3Fff/11pW3btmbb9u/frwQHByvfffed2fY//vgjx/bs75s//vjDbF9L6/nUU0+Zfd9u27ZNCQ4OVrZv327a9tprrynh4eFm34uK8l+NFMW6HFvjzJkzyrRp05SIiAglODhYadmypbJ48WLl5s2beXr///73PyU4OFj59ttvzbZ/9dVXSnBwsPLcc8+Ztv39999K//79la+//lrZtm2bsnjxYuWpp55SQkNDlaNHj5q9v1atWsrw4cNzfN6cOXOU4OBgZeTIkWbb+/Xrpzz11FN5HXauFi9erAQHB5v+devWzVRvS3377bdKcHCw8vXXX5u2RUVFKcHBwcqmTZty7P/RRx8pwcHBSnp6+qOGLywgl6tsYPHixXTq1Inw8HAaNGhA3759c5y6TE9PZ/z48dSrV4/w8HD69+9vem4VZM1GCAkJITEx0W5xx8fHM3jwYGrXrk3dunUZNWoUd+7ceeB7YmJiGDp0KA0bNqRWrVp06NCB48eP07lzZypXrkxsbCwTJkygQ4cOhIaGUr9+fYYOHUp0dLTpGOXLlzc7q+Hv74+rqytXr159aMxVq1Y1LaAImG5qrl+/vtlp4eztuU2tv3vtHm9vbypXroyHh4fZqtBVqlTB29vb7P1btmyhTp06eHt7c+vWLdO/iIgIDAYD+/fvN/ucli1b5mkWRfb9Jr/88ovZZZK7bd26FV9fX15//fUcr2XPCPnjjz/Q6XS88cYbZq+/9dZbKIqS46/HJ598kqpVq5q+VhSFrVu30qxZMxRFMRtjZGQkSUlJHD169L7jyF4a4N7Pf5Q1nB5my5YteHl50bBhQ7M4s298z76kla1cuXI0atQoxzEsqeezzz5r9n2bfSYk+3vk1q1b7N+/n06dOpl9L8J/NbI2x9YICgpi+PDh/P777yxYsICgoCDmzJnD008/Td++fTlx4sQD39+kSRPKli3L9OnT2bp1K1evXuXHH3/k008/xcXFhbS0NNO+tWvXZs6cOXTu3JnmzZvTq1cv1q9fj0ajYebMmRbF3aVLF7Ov69atS3x8vFWX3tq2bcuKFSuYOXOmaZmLu+PPq7NnzzJhwgTCw8Pp0KGDaXt6ejpArvfgZZ8Je5TPE5aTy1U2sG/fPl577TVCQ0MxGAx88skn9OjRgx9++MF0T8qUKVP4/fffmTVrFl5eXkycOJF3333XdG+AIwwZMoSYmBhWrFiBXq9n1KhRjBkz5oE/hIYPH05iYiILFy7E19eXzZs3M3DgQDZs2MCzzz6LwWBAr9dTvHhxrl27RlJSEsePH2fAgAGmsXp6euY4rqura55+aJUuXdrsay8vLwBKlSpltj37M+5tGosUKZKj8fDy8qJUqVI5po96eXmZvf/ixYucPHmSBg0a5Brbvet55HXq+7PPPsvXX3/N6NGjmTlzJg0aNOCZZ56hdevWpkttly5donLlyg9crfrq1auUKFEiR36zL+/c20TeG9+tW7dITExk3bp1rFu3Lk9jvPfztVptjssNVapUue97HtXFixdJSkq6by3ufrYb5F4LS+t57/dedsOT/T2S3ezcPcMmt2Nak+P4+Hiz+0bc3d3x8vLi1q1bZqulFy1alGLFiuV6DBcXF5o3b06zZs346aef+PDDD/nll1946qmnHniDcpEiRVi8eDEDBw40LQXg5ubG0KFDWbRo0X3vv8tWsWJFmjdvztatWzEYDHmeWHBvw5j9R0FCQkKuP0vyomzZspQtWxaA5557jg8//JDu3buzZcuWPF+yiomJoXfv3nh5eTF79myz8WQ3MtmX+O6W3QCp5cZ3R5MmxwaWL19u9vW0adNo0KABR48e5cknnyQpKYkNGzYwY8YM0w/UKVOm8Oyzz3Lw4EECAgJMa0o8+eSTAHTo0IFp06YBWX/9TZ8+nW+++QZXV1e6dOli+iHzqM6ePcuOHTv45ptvTNfIR48eTa9evRg2bBglS5bM9X0HDhxg7NixhIWFAdC3b1+++OILjh49yuOPP45OpzPFdvz4cbZt20a5cuWYNWuW6Yfz/R7mqeRhyab7/WC83/Z7j2nN+41GIw0bNuTtt9/Odd9KlSqZfZ3XH2Lu7u6sWbOGvXv38ttvv7Fjxw5+/PFH1q1bx2effZZvs8zujS/7LNLzzz9v9lfp3UJCQvIlFksZjUb8/f3NZrPc7d5GNrdaWFrPvH6PPYi1Oe7fvz/79u0zfZ39c6Jz585mTezdaxLd6+rVq2zatIlNmzZx5coVypYtS/fu3XMs3JmbatWq8f3333PmzBkSEhKoWrUq7u7uTJ061fSz60FKlSqFXq8nNTU1zw2KNT8v8qpVq1asX7+e/fv35zjjl5ukpCR69uxJUlISa9asyfHzMvv5gDExMTneGxMTQ/HixQvsTEtnI01OPsh+JEP2X3pHjhxBr9cTERFh2icoKIgyZcpw8OBB3njjDebOnUv//v3ZsmULnp6eZj+UN23aRPfu3Vm/fj0HDx5kxIgR1K5dm4YNGwJZs4auXr2aY2rkgxw4cABvb2+zmwAjIiLQarVERUXxzDPP5Pq+8PBwfvrpJ55++mm8vb356aefSE9Pz3VxthIlSpCWlsbu3bsJDw/PsSpyYVOhQgVSUlLM6mgrWq2WBg0a0KBBA0aOHMmiRYv49NNP2bt3LxEREVSoUIFDhw6h1+vvm8eyZcuye/dukpOTzX6BZF86zf7L9X78/PwoVqwYRqPxkcZYtmxZjEYjly5dMjt7c++lW1uoUKECu3fvpnbt2o/8F7Gt65l9g/qpU6fuu4+1Oc4+k5ot+5fpxx9/bDpDcHcs2dLS0vj555/ZsGEDe/bswdXVlRYtWjBhwgQiIiIeugje3TQajdlNtr///nuex3PlyhWKFCny0LM+9pZ96Sgvj9NJT0+nT58+XLhwgRUrVphd8s1WsmRJ/Pz8cr0ZOyoqSp7nZ0dyT46NGY1GpkyZQu3atU2nrWNjY3F1dc2x1oe/vz8xMTHodDpTQ+Tv709gYKDpMgz8t1pspUqVaN++PTVq1DB72GdgYGCOU+kPExsbm+OvXRcXF3x8fHL96yPbrFmzyMzMpF69eoSGhjJmzBjGjRuX4xLFxx9/TNOmTVm2bBmJiYksWLDAovgKojZt2nDgwAHTVOu7JSYm3nea7MPc/aywbI899hjw3+nuli1bcvv2bdN097tl/0XbuHFjDAZDjn0+//xzNBoNjRs3fmAcOp2OVq1a8b///S/XX9QPW14/+/j3NttffPHFA9/3KNq0aYPBYMj1+yozMzNP97bZup5+fn48+eSTbNiwwfTcumzZNbI2xzVq1CAiIsL0L/sXbJ06dcy2393kjBkzhsjISIYMGUJcXBwjRozgjz/+4NNPP6Vhw4YWNTj3SktLY/bs2QQGBppNLc9tHCdOnGD79u00bNjQ7OxM0aJF7XYv4v3y+80336DRaHjiiSfM9j179qzZdHWDwcDAgQM5ePAgs2fPNrs/8F4tW7bkt99+4/r166Ztu3fv5sKFCwVq7SBnJ2dybGz8+PGcPn2aL7/80mbHvPf0dWBgoNk9B4MHD37g+8eMGWO25suBAwceOZbZs2eTmJjI559/jq+vL9u2bWPMmDGcPn2aBg0a4Ovri8FgoEqVKrz00ktkZmZy4sQJhg8fzuLFix/5cwuCHj16sH37dvr06UOHDh144oknSE1N5dSpU/zvf//jl19+eaTl2ufPn89ff/1lurEzLi6OL7/8klKlSlGnTh0A2rdvz7fffsvUqVOJioqiTp06pKamsnv3bl555RVatGhBs2bNqFevHp9++ilXr14lJCSEP//8k19++YVu3brlaERzM3jwYPbu3ctLL73Eiy++SNWqVUlISODo0aPs3r3b7FLJvR577DGee+45vvzyS5KSkggPD2fPnj1cvHjR4pw8zFNPPcXLL7/M4sWLOX78OA0bNsTV1ZULFy6wZcsWPvjgg4f+IsmPeo4ePZpXXnmFDh068PLLL1OuXDmuXr3Kb7/9xv/93/8B1uX4UXz//fc8++yzvPjii4+0+vjdBgwYQIkSJahatSrJycls2LCBy5cvs2TJErOzhwMHDsTd3Z3w8HD8/f05c+YM69evx93dnSFDhpgd84knnmD37t2sWLGCEiVKUK5cOYvjzH4ExMqVK6lXr95991u4cCH//PMPjRo1okyZMsTHx7N161YOHz7MG2+8QcWKFU37rlmzJscxp02bxvbt22natCnx8fGmmmbLft4aQJ8+fdiyZQtdu3ala9eupKSksHz5coKDg+nUqZNF4xOPTpocG5owYQK//fYbq1evNrsRNiAgAL1eT2JiotnZnLi4OAIDAx963HtvNtVoNBZdjx4wYAA9evQw2xYQEJDjr5rMzEwSEhLuG9OlS5dYvXo133//vel0dfXq1dm5cycnTpygdOnSHDlyBKPRiJeXF40bN6Zu3bokJibSpEkTDh48mOeYCyIPDw9WrVrF4sWL2bJlC99++y2enp5UqlSJ/v37m519s0SzZs24evUqGzZs4Pbt2/j6+vLUU0+ZHVOn07F06VIWLlzI999/z9atWylevDi1a9c2NcFarZaFCxcyZ84cfvzxRzZu3EjZsmUZNmwYb731Vp5iCQgI4Ouvv2b+/Pn8/PPPfPXVVxQvXpyqVavm+OWUmylTpphuSP/ll1+oV68eS5Yseeizyh7FhAkTqFGjBmvXruXTTz9Fp9NRtmxZnn/+eWrXrv3Q9+dHPatXr8769euZPXs2X331Fenp6ZQpU8Zs5p61ObbUzp07bXZ5qEaNGmzcuJF169bh7u5OnTp1mDlzpunMY7YWLVqwefNmPv/8c5KTk/H19eWZZ57h3XffNWskIOty+5gxY5g1axZpaWl06NDB4iYnJSUFjUZDQEDAA/d7+umnuXz5sun/a25uboSEhDB16tT73iN1t+wZaL/++iu//vprjtfvbnJKly7N6tWrmTZtGjNnzjQ9u2rEiBFyP449OWLeurMxGo3K+PHjlcjIyBxrbiiKoiQmJipPPPGEsmXLFtO27HVYDhw4oChK1roSwcHBOdbOyG2dnHfeeSfXdSUscebMGSU4OFg5fPiwaduOHTuUkJAQ5caNG7m+58SJE0pwcLBy5swZs+1vvfWWMnr06Pt+1tWrV3OsWSOEELbSqVMnpX///o4OQxRAck+ODYwfP57vvvuOmTNnUqxYMWJiYoiJiTHdzObl5UWnTp2YNm0ae/bs4ciRI4waNYrw8HDT03jLli2LRqPht99+49atWw9dr+ZuM2fOZNiwYRbFHBQURKNGjfjwww+Jiori77//ZuLEibRt29Y0UyD7QXbZS9dXqVKFihUrMmbMGKKiorh06RKfffYZf/75p+nheocOHWL16tUcP36cq1evsnv3bgYNGkSFChUeeP1aCCEeRXJyMidOnDA9T0uIu8nlKhv46quvgJyLoE2dOpWOHTsCMGrUKLRaLe+99x4ZGRlERkYyduxY074lS5akf//+zJw5k5EjR9K+fXvTFPKHiYmJMbu5La9mzJjBxIkT6datG1qtlpYtWzJ69GjT63q9nvPnz5tuvHN1dWXJkiXMnDmTPn36kJKSQoUKFUzPSoKsqbpbt25l7ty5pKSkEBgYSKNGjejbt6+cohVC2Jynp2eus5iEANAoig0XGxBCCCGEKCDkcpUQQgghnJI0OUIIIYRwStLkCCGEEMIpSZMjhBBCCKckTY4QQgghnJJMIQfi4pJ40BwzjUZDsWJFTE8Qfqg7KbicPE5myGNQzLqVRt9++y1CQkIYOnS4Vccp6LRaLXfupNv0ycKamBiK/N8m0l/ogJKHlaWFOnLmzGN05rHZiuTIfvIz1xoN+Ps/fFVymUIOxMY+uMnRav9rcnLbb9GiBSxZsshsW6VKldi48TvT1+np6XzyyQy2bt1CRkYGDRpEMHLkaPz9/QH466/99OrVg99/34mX13+PfujZ8y2Cg61vch72+blRFIVFixawadMGkpKSqFmzFqNGjaZChf+WZU9ISGD69Kn88cfvaDRamjdvwdChwy1aRl6j+a/JMRpV/+0ohBDiITQaCAh4eJMjl6tsJCgoiK1bt7N163Z+3vwjn48cjebfFY8BZs6czo4dv/PRRzNYunQFMTExDBnyvt3ie5TP/+KLFXz11ZeMGvUhX3yxBg8PD/r160N6erppnw8+GMHZs2dZsGAxs2fP5Z9//mbSpPH5PZw80cTfxu27TWjibzs6lEJDDTlz5jE689hsRXJkPwUh19Lk2IhO50JAQAABAQEEurhQ+rdf0CTEA5CUlMS3325i0KAhPPVUPR5//HHGjZvIoUMHiYo6xLVrV+nVK+sBmk2aRFK7dhhjx/638rCiKMya9QlPPx3JM880ZdGiBRbF9rDPz42iKHz55WrefrsnTz/dlODgYCZMmExMTAy//bYdgHPnzrFr15+MGTOO0NAwwsNrM2zYCP73vy3ExNx8hCzalu7SRXze7obuku2fgu2s1JAzZx6jM4/NViRH9lMQci1Njo1cunSRli2b065dG0Z+NJVrdz176vjxY2RmZlKvXn3TtsqVK1OqVGmioqIoWbIUH3/8CQCbNn3H1q3bGTLkv8tT33//HR4eHqxcuYYBA95n6dLF7Nmz2/T62LGj6dnz/k+Zftjn5+bq1avExsaavcfLy4saNUJNjVFU1CG8vLx4/PEnTPvUq1cfrVbL4cOHH5ozIYQQIj/Jjcc2EBoayvjxk6hYsRKxsTEsmT+X1/bt5etub+EBxMXF4urqanavDYC/vz9xcbHodDp8fHwA8PPzy7Ff1arV6N37HQAqVKjIunVr2bdvL/XrNwAgICDwgTdFP+zz7/eerHjM79nx9/cnNjbOtI+fn5/Z6y4uLnh7e9/3uEKIgsloNGIwZDo6jHxn1GpIq1iRDK0Ggz7D0eE4NWtyrdVq0Wp1aDQaq2KQJscGGjZsZPrv4OBgwkqW5NlXX2brjt95oXIVq49frVqw2dcBAQHcunXL9HX//vL0XSHEo0tPT+X27RhABTf+excldsFCjN5FIc7yBxsLC1iZazc3d7y9/XBxcX3kEKTJyQfexX2p5OfH5RvRAPj7B6DX60lKSjQ7mxIXF4e/f8BDj+fiYl4mjUaDouRxOvsjfn729lu34gi8a+pfXFwcISEhpn3ubrYAMjMzSUxMzNO48pvi7oE+tCaKu4ejQyk01JAzZx7jo4zNaDRy+3YMbm7ueHr6WP2Xc4GXno42zYDRrxQUKeLoaJzbI+ZaURQMhkySk+OJi7tBiRLlHvn7UpqcfJDs4cGl1FTalC8PwGOPPY6Liwv79u2lefNnALhw4Tw3blwnLCwMAFfXrE7VYMh785JXefn8e5UtW5aAgAD27dtLSEj1rHElJ3PkyGFefPElAMLCapKUlMSxY8d4/PHHAdi/fx9Go5HQ0FCbj8NShuAQ4n/Z4egwChU15MyZx/goY8u6RKXg6emDm5sKfum7ukFwdXSOjkMNrMp1EXQ6HbduRZOZqcfV1e2RjiJNjg18+ukMGjd+mtKlSxMTE8OiRQvQanW0bt0GyLpht337DsycOQNvbx+KFfNk+vSphIXVJCysJgClS5dGo9GwY8fvREY2okgR9zyvNTN37mxu3oxm4sQpub6el88H6Njxed59dwDNmjVHo9Hw6quvs2zZEipUqECZMmVZuHA+gYGBPP10MwCqVKlCRERDJk0ax6hRH5KZmclHH02lVavWBAaWsCalQgg7c/ozOKLQ0Wisnxsls6tsIDr6JiNHDqdDh+cZPnwIxd3dWd+kMf56vWmfwYOH0ahRY4YOHcTbb7+Jv38AM2Z8anq9RImS9OnTl7lzZ9OiRVM++ij3hiU3sbEx3Lhx44H7POzzAS5cuEBycrLp627dutOly6tMmjSBN954lZSUFObNW0iRu047Tp48jUqVKtOnT0/ee68ftWqFM3r02DzHnp9cDh8ioFwALodznyYvclJDzpx5jM48NptJScHl0EFISXF0JM6vAORaVjzG+hWPc+wffQOPlStI7dodY8lStgvUSeXXiscuUQfxbdGY29v+IDOsls2O68zUkDNnHuOjjE2vzyAu7jr+/qXNLgloNPY9u6MoSp5+vlotJQWXUyfIDK4OFqzMLh6Blbm+3/cm5H3FY7lcJYQQwoxGA3qNljsZ9ptSXszNBVfy9oekyF83btxg5syp/PPPX3h4FKVNm+fo3btfjkkwd0tMTODTTz/mzz93oNVqaNKkGQN69cU7l32vXLlM9+6vodNp2bLlt3wbB0iTI4QQ4h4ajYY7GZn8euImKen53+gULeJC0+ol8HXT2fQhvcJyBoOBYcMG4Ofnz6JFnxEbG8vkyWNxcXGhd+9+933f+PEfEhcXy6efziczM5OpU8czfdYMJnV6yWy/zMxMxo37gJo1a3HkSO6L0dqSNDlCCCFylZKeSbIdmpxH8e67vahSJQiA//3vR1xcXGjfvjNvv93HdJktMTGR2bNn8OefO9DrM6hVqw4D+/Sj8r/HuHHjOp98Mp2oqINkZuopVaoM/fq9R4MGkXmKYfnyxezY8TudO7/MZ58tISkpkVat2vL++0NZu3Y169Z9idFo5MUXu9CtWw/T+5KSkpg/fxY7d/5ORoae6tUfo3//QaY10a5evcLcuZ9w9OgR0tJSqVixMr179+PJJ+uZjtG5czuef74DV65c5tdff8HLy4tu3Xrwwgsdrcrrvn17uHDhPLNmLcDPz59q1UJ4++0+LFw4l7fe6mWaCXy3CxfOs3fvLpYtW0n16lkzbQcOHMrQoQMY0Lwlvnftu2TJAipWrEidOk/ZpcmRG4/zgdE/gNTuPTEWgLVi1CyzWgi3/thLZrUQR4dSaKghZ5nVQri9Yy+GYOcboxrqd7effvoBnc6FpUu/YMCAIaxbt4bNm781vT5lyjhOnjzORx99wqJFK1AUhaGjR5AWVA3c3fnkk4/Q6zOYP38pX3yxlnfe6Y+Hh2X3jly9eoU9e3Yxc+Zcxo6dzA8//B9Dhw4kJuYm8+Yt5p13+rN06UKOHj1ies+HHw7n9u1bzJgxh+XLVxEcXJ2BA98hMTEBgJSUFOrXb8js2Qv47LM11KvXgOHDB+WYYLJ27RqqV3+cFSvW0KHDi8ycOY1Lly6YXn/33V5MnjzOovEcPXqYKlWqmq12/9RTDbhz5w7nz5/N9T1HjkTh6ellanAA6tZ9Cq1WS5ReD+7uAPz9935+/fUXBg0anutx8oM0OfnBxQVjQAA84PplXo0dO5pBg2RF40fi4YGh+mPg4XyLvuUbFeRMU9SDtMeeQF+0GE43a1oF9btbyZIlee+9QVSoUImWLdvQqdPLrF//JQCXL19i584/GD58NDVrhlOtWjBjx04kJvYmf+zfC1ot0dE3CA2tSVBQVcqWLUfDho2oVau2RTEoipFRo8ZQuXIVIiMbEx5el8uXL/Lee4OpUKESbds+T4UKFfnnn78AOHToIMePH2XixI+oXv1xypevwLvvDsTT04tff/0FyFrlvn37TlSpUpXy5SvQs+c7lC1blj///N3ssxs0iKBjxxcpV648r7/eDR+f4qbPycpPKYsXZo2Li8vxuJ7shicuLi7X99y6FYevr6/ZNhcXF7y8vLmVnARaLQkJ8UyePI4PPhhLsWKeFsVkDWlybMBgMLBgwTyee641DRo8yfPPteazYYPh36eQQ9bMgYUL59OyZTMaNHiSPn16cumuJ7Neu3aV2rXDOHnyRL7E+LDPz9O4nn+WpUsXm10zf5Tj2ov28iU8338X7eVLjg6l0FBDznRXLlO0/zuknzvndGvDqKF+d3v88RpmNaxRI5TLly9hMBi4ePE8Op2Oxx+vYXrdx6c4FcpX4FLUQchIp3PnLnzxxXLeeectli9fzJkzpy2OoVSpMhQtWsz0tZ+fH5UqVUar1d61zZ/4+KzV4c+cOUVqaipt2zbnmWcamf5dv36Nq1evAPy7XMcsXnutM61bP80zzzTi4sULREebn8kJCqpm+m+NRoOfnz+3b982bfvwwwn06fPufWMfPPg90+e//vpL993v0Slobt+CjHQ++mgyzzzT2uIm0lpyT44NfP75Z3zzzXrGj59EUFAQx3f9ydjp0yhatSpdemU9WPOLL1bw1VdfMmHCpH8X1ptHv359+Oabb83Wnckvj/L5947r2LGjjBs3Bk9PT1555bUCMa4H0d6+hcealaR1fxtj+QoOjaWwUEPONLdu4fPVKhK79YAqlR/+hkJEDfWzmgKa1BTINNCuXXueeqo+u3fvZN++vaxatYJ33x1I585d8ny43B67k9sspOzlMVJTU/D3D2Du3MU59vH0zJoSPX/+LPbv30u/fgMpV648RYoUYfTo4ej15vdH5fbZD3pY871GjBhNenq62bH8/f05fvyo2X63bsWZXsvNvc0VZN1gnJSYRIBOB5kG/vlnP3/++Qdr164Gsv5ANhqNNGlSj6FDR/Hccy/kOW5LWHwmZ//+/fTp04fIyEhCQkLYtm2b2euKojB79mwiIyMJCwvjzTff5MKFC2b7xMfHM3jwYGrXrk3dunUZNWoUd+7cMdvnxIkTvPrqq4SGhtKkSROWLl2aI5affvqJ1q1bExoaSrt27fj9999z7GMPhw4dokmTpjRq1JgyZcryTKPGRJYqxZGTJ4GsnHz55WrefrsnTz/dlODgYCZMmExMTAy//bYdgOeey1od+ZVXXqJ27TB69nzL7DNWrvycli2b0bRpI6ZOnYz+roUGHyYvn5+XcbVo0ZL69Rtw5MgRq44rhBC2cOyY+S/jo0ePUL58BXQ6HRUrVsZgMHDs2H/3wiQkxHPpyiUqlypt2layZCnat+/MlCkf06XL62b39OSHkJDq3LoVh06no1y58mb/ihcvDsDhw4d49tl2NGnSlKCgrPtjbty4ZvNYAgNLmD671L85eeKJUM6dO8Pt2/89l3D//r0UK1aMSpVyf+B0jRphJCcnceLEcdO2f/75C6Ni5ImKWX9ILFq0ghUr1pj+9ejRm6JFi7FixRoaN25q87Fls7jJSUlJISQkhLFjc1/VdunSpaxatYpx48axfv16PDw86NGjh6lbBBgyZAhnzpxhxYoVLFq0iL/++osxY8aYXk9OTqZHjx6UKVOGjRs3MmzYMObNm8e6detM+/zzzz8MHjyYzp078+2339K8eXP69evHqVOnLB2S1WrWrMm+fXu5ePECACfPneXvmBga1n0SgKtXrxIbG0u9evVN7/Hy8qJGjVCiorJWJl21Kus68sKFS9i6dbvZasR//bWfK1cus3jxcsaPn8Tmzf/H5s3/Z3p90aIFtG3b+r7x5eXz8zKuU6dOcvDgARo2jLTquEIIYQvR0TeYO/cTLl26wM8/b2HDhnWmszDly1egUaMmfPTRZA4dOsjp06eYMGEMgf4BNP73cTazZ89k797dXLt2lZMnT/DPP39RsWL+nt2rW7ceTzwRysiRQ9i3bw/Xr1/j8OFDLF48nxMnjgFQrlwFfv99O6dPn+T06VOMH//BIy2UOnHiGBYtmmfRe556qj6VKlVm4sQxnD59ir17d7N06UI6dnwJN7esBfmOHTvCq692IibmJgCVKlWmXr0Ipk+fxLFjR4iKOsgnn0yn+dPNCfy3catUqTJVqlQ1/QsMLIFWq6FKlap4e+e2mo5tWHy5qkmTJjRp0iTX1xRFYeXKlbzzzju0aNECgOnTpxMREcG2bdto27YtZ8+eZceOHXzzzTemhziOHj2aXr16MWzYMEqWLMl3332HXq9nypQpuLm5Ua1aNY4fP86KFSt4+eWXAVi5ciWNGjXi7bffBmDgwIHs2rWL1atXM2HChEdKxqPq3r0Hd+7coWPHF9DpdBgMBt4PDaVts+YYgbi4WACzu9Uh69RfbGzWacDsm7aKFy9OQID5jWJeXt4MHz4KnU5H5cqVadSoMfv27aNjx87/vseXcuXK3Te+vHx+XsfVr19/nn22rVXHFUIUDkWL2OeOhkf9nNat25Kenk7Pnt3QanV07tzFbAr1yJFjmT17BsOHD0Sv11OzZm0+njwdlzt3yASMRgOffPIRMTE3KVq0GPXqNeC99waZ3t+5czvatHmOHj16WztEE41Gw4wZs1myZAFTpownPv42fn7+1KpVG1/frBt++/d/n6lTJ9Cnz1v4+BTntde65bjakRfR0TfM7g3KC51Ox/Tps5gxYyp9+nTHw8OD1q3Nc5CWlsalSxfJzPzv8tnYsRP55JPpDBjQ17QY4MDe/cDB94fZ9Dv4ypUrxMTEEBERYdrm5eVFzZo1OXDgAG3btuXAgQN4e3ubPaU6IiIia6pZVBTPPPMMBw8epG7duqauESAyMpKlS5eSkJCAj48PBw8e5M033zT7/MjIyByXz+zh55//x08//cCUKdOoUiWIU1GHmPHpTHz++J12L75s9fGDgoLQ6f57jmtAQACnT/93g1yXLq/QpcsrVn/Ove4d18mTJ5k5czqBgYG0a5c/109tyRhYgpT3BmGUh4XmmRpyppQowe133yfTCcdoq/opikIxt6wF+uylmJsLipL3+0kg6z6SAQMGM2TIyFxf9/b25sMP7/mjNyMDY2wsuLjw/vvD7nvstLQ0bt26RXh4nfvu06NH7xwN0AcfjMux37x5S8y+Llq0GAMHDmXgwKG5Hrd06TLMmbPIbFunexbV++abzTne9/nnXz7wc/OqVKnSzJgx576v165dl507/zLb5u3tw7hxk813zMjAWKJUrjONn322Hc8+2+6R4rOETZucmJgYIOfNSVl/2Wf91R8bG5tjepqLiws+Pj6m98fGxuY4M5F9diM2NhYfHx9iY2NznPG4+3PsadasT3jzzR60apV1X021asFcv32bFV+tod2LL5um8N26FUdgYKDpfXFxcYSEPHw9i5w3sWksWhX0UT8/t3HduHGdFSuW067dC1aPK78ZS5fhzuhxjg6jUFFDzoyly3B7VMF4iKyt2ap+igKuGPF10z18ZxtRFDs90sHNDWOZMg/d7Z9//qJOnbrUrl3XDkE5qTzmOj/JFHIbSEtLQ6v9bxqjJiMDl6REjIasv0rKli1LQEAA+/btNe2TnJzMkSOHCfv32nD2KpIGg2V/yeRFXj4/N/eOC7IepJl9bfhRj2svmuQkXP/cgSY5ydGhFBqqyFlSEh67nHOMtqyfomTNCLLXP7s9zcFgyMqPwfDA3SIiIvn449l2CspJ5THX+cmmTU72X/P3LhgUFxdnOusSEBDArVu3zF7PzMwkISHB9P6AgIAcZ2Syv777OPfuc/fn2FPjxk1YvnwpO3b8wbVrV/n1++9YtX4dzf69IVej0fDqq6+zbNkSfv/9V06fPsWYMR8QGBjI0083A8DX1w93d3d27dpJXFwcSUl5/yG1du1X9O799n1fz8vnA/Tu/TZr135133Ft3/4Lq1evomnTZhYd11F0585SvENbdOdyX6VT5KSGnOnOnaVs53a4OeEY1VC/bPPmLWHAgMGWvzE9Hd2Z03DXZBiRTwpArm16uapcuXIEBgaye/duHnvsMSDrL/tDhw7xyitZ94yEh4eTmJjIkSNHqFEja5GmPXv2YDQaCQsLA6BWrVrMmjULvV5vOsOxa9cuKleujI+Pj2mfPXv2mN2Xs2vXLmrVqmXLIeXJsGEjWbBgHlOnTub27VsE+vrxctWq9OjazbRPt27dSU1NZdKkCSQlJVGrVjjz5i00rSXj4uLC0KHDWbp0MYsWLSA8vDZLl36Wp8+Pj7/NlStXHrjPwz4fsu6pio//b62DHOMKDKRTp8706tXHouMKIYQQjqBRLHzk6507d7h0Ketu6fbt2zNy5Ejq1auHj48PZcqUYcmSJSxdupRp06ZRrlw5Zs+ezcmTJ/nxxx9Nv/jefvtt4uLiGD9+PHq9nlGjRlGjRg1mzpwJZD28rHXr1jRs2JCePXty+vRpRo0axciRI02zq/755x/eeOMNBg8eTJMmTfjxxx9ZvHgxGzduJDg42KIkxMYmPfBUqVaroVixIhiNebtmrI2+gcfKFaR27Y6xZCmLYlEjjSbrMtidO+mPNE3yflyiDuLbojG3t/1BZlgtmx3XmakhZ66HD1G8eSMub/mNonXr2PR7ztEepX56fQZxcdfx9y+Nq6vbw99Q2KWk4HLqBJnB1aGoZc+pEhayMtcP+t7UaCAgwOuhx7D4TM6RI0fo2rWr6eupU6cC0KFDB6ZNm0bPnj1JTU1lzJgxJCYmUqdOHZYtW2b2l/2MGTOYOHEi3bp1Q6vV0rJlS0aPHm163cvLi+XLlzNhwgQ6duyIr68vffv2NTU4ALVr12bGjBnMmjWLTz75hEqVKjF//nyLGxwhhBBYNJlBCHuwxfekxU1OvXr1OPnvSr650Wg0DBgwgAED7v9QyeLFi5vO2txP9erV+fLLLx+4T5s2bWjTps2DA3YErQ6jlxdo7TczQeSkuLhiKF0GxcXV0aEUGqrImasrmaXLoLg63xgfpX7Z66gYDJmACi4zazQorm4439NZCyArc52RkXUvj0736HfWWHy5yhnZ+nKVsEx+Xa4SIjdarYbbGVmzPXzddKr/nlMUhVu3bmI0ZuLj449GI5NuhWMpikJGRjrJybfx8PDExyfnM7Py7XKVEEII56HRaPDx8SMu7ga3bkU7OhwhTDw8PPH29nv4jg8gTU4+0MbEUGTDOtI7vYzxrkXyHsXYsaNJSkrik09kvQZL6Y4dxeeVTiR8tQHD4084OpxCQQ050x07SqWXO3Jt9ddQM8zR4djUo9bPxcWVEiXKkZmZ9wf/FlbaM6fxGvIeSTPmYKxazdHhODVrcq3TuVj8SIrcSJNjA3fu3GHBgnn8+ut2bt++RUiVID4sX45qxv8WQFIUhUWLFrBp0waSkpKoWbMWo0aNpkKFigBcu3aV555rw1dfrSckpLrNY/zll21s2PA1x48fIyEhIc+fs2bNKr75Zj03btygePHiNG/+DP37DzDdSN62bWuuX8/5dNwXX3yZkSM/sPk4LKHJ1KO7fg2NCn5w24oqcqbX43L9Ghq9843RmvppNBpVzK5y0esptn8/GXo9mSoYryMVhFzLxVcbmDBhHHv37mHixMmsW7eBBrVr0/3XX4m+a7HCL75YwVdffcmoUR/yxRdr8PDwoF+/PmZPZ89Pqamp1KoVznvvDczze3766Qfmzp1Nr1592LDhW8aMGc/Wrf9j3rz/nmmyevWXbN263fRv4cKsZ6U880xLWw9BCCGEsIicybFSWloa27dv45NPZlOnTtYzTt55vSs7f/yBr3/YTN8naqAoCl9+uZq33+7J0083BWDChMk880xTfvttO61ateG557Jmib3yStZD2OrUqWu2GODKlZ+zevVK9Ho9LVu2ZsiQYaaFEvPiueeyHoR27drVPL/n0KFD1KxZizZtsp46XqZMWVq3bsORI4dN+2Q/NTfbihXLKVeuvCkXQgghhKPImRwrGQwGDAaD2RPTAYrodBw4ehSAq1evEhsbS71/H/MAWWsB1agRSlTUIQBWrcqaLr9w4RK2bt3OjBmfmvb966/9XLlymcWLlzN+/CQ2b/4/Nm/+P9PrixYtoG3b1jYfW82aNTl+/Lipqbly5Qo7d+6gYcPIXPfX6/X89NMPvPBCezQyPVMIIYSDyZkcKxUrVoywsJosW7aEKlWq4Ofnz4979nDw1i3K/7vCY1xc1mUrP7/cns6e9ZwvX19fIGsNoXufv+Xl5c3w4aPQ6XRUrlyZRo0as2/fPjp27Pzve3xzPLXdFtq0aUt8fDxvvZX1eIrMzEw6d36RHj165rr/r79uJykpieeff8HmsTwKQ5Ug4jf9gKFKkKNDKTTUkDNDlSCufrOZjCpBONt6t2qon7UkR/ZTEHItZ3JsYOLEKSiKQqtWLahfvy5ffbOeVq3aoLHRYoBBQUHodP8dK+shp/89BLVLl1dYvHiZTT7rbn/9tZ/PPlvGyJEfsGbNWmbM+JSdO3ewdOniXPf/9ttNREQ0JDCwhM1jeRSKpxf6ho1QPB++loLIooqceXmRGuGcY1RF/awkObKfgpBraXJsoHz58ixbtoI//9zDjz9uZfXCxRivX6NcqaznVvn7Z52Zubsxgeynpudc5OheLi73nnDT2GUJ9gUL5vHss8/RoUMnqlULplmz5vTr9x4rVizHaDSa7Xvt2jX27dtDhw6d8j2uvNJev0axSePQ5jL7S+RODTnTXr+G/5Tx6JxwjGqon7UkR/ZTEHItTY4NeXgUJTAwkKQbN9h19AhPP/kUAGXLliUgIIB9+/aa9k1OTubIkcOEhdUEMN1EbDAYcx7YQdLS0nKsU6DTZX19b5P13Xff4ufnR2RkI7vF9zDamJsUnfMJ2pibjg6l0FBDzjQ3b+I771NcnHCMaqiftSRH9lMQci335NjArl1/oigKlSpV4vLly8ya8RFVvL15oWUrIGv9iVdffZ1ly5ZQoUIFypQpy8KF8wkMDOTpp5sBWbOU3N3d2bVrJyVLlsTNzQ0vr7yd4lu79it+/fWXB16ySkhI4MaN68TExABw4cIFIOssU/Y9QB9+OIoSJUrSv3/Wc8caN27CmjWrqF69OjVqhHL58mUWLJhPo0ZNzC6fGY1Gvvvu/3juuedzOeskhBBCOIb8RrKB5ORk5s2bTXR0ND4+PjRvEMFQN1dcXFzIPi/TrVt3UlNTmTRpAklJSdSqFc68eQtNi+q5uLgwdOhwli5dzKJFCwgPr202hfxB4uNvc+XKlQfu8/vvvzFu3Iemr0eOHAZAr1596NOnLwA3btwwO3Pz9tu90Gg0zJ8/j5iYm/j6+tKoURPefbe/2bH37t3DjRvXeeGF9nmKVwghhLAHeUAntn9Apzb6Bh4rV5DatTvGkqVsF6iTyq8HdLpEHcS3RWNub/uDzLBaNjuuM1NDzlwPH6J480Zc3vIbRevWcaoHdKqhftaSHNlPfuY6rw/olHty8oHi4YE+rCaKh4ejQ1E1o68fqa91xehr3QPe1EQNOVP8/Eh45Q0Mfs43RjXUz1qSI/spCLmWMznY/kyOsEx+nckRIjdarYbbGVnPlfN108n3nBCFkJzJcaTMTLSxsZCZ6ehI1C01Fd2J45Ca6uhICg815Cw1FbeTx9E44xjVUD9rSY7spwDkWpqcfKCNi8VjxVK0cbEP31nkG5fTJ/FrXA+X0ycdHUqhoYac6U6dpELTBrg54RjVUD9rSY7spyDkWpqcAm7s2NEMGjTA0WEIIYQQhY5MIbeBO3fusGDBPH79dTu3b98ipEoQH5YvR7W79hk7djSbN39n9r4GDSKYP38RkPV08Oeea8NXX60nJKS6zWNctGgBW7du4caNG7i6uvLYY4/Tr19/QkPD7vuezz5bxvbtv3DhwnmKFClCzZq1eO+9gVSqVNm0T3p6Op98MoOtW7eQkZFBgwYRjBw5Gn//h6/kLIQQQuQnaXJsYMKEcZw9e4aJEycTGFiCn75eS/f169jwxpsE3jWFPCKiIePGTTR9fe+Ty/NTxYoVGT58FGXLliM9PY01a1bRr18f/u//vsf3Pne+//33X7z0UheeeOIJDAYD8+bNoW/fPmzYsAkPj6xHG86cOZ2dO3fw0Ucz8PT04qOPpjBkyPusWLHSbmMTQgghciOXq6yUlpbG9u3bGDDgferUqUuFChV4541uVPDy4usfvjfb183NjYCAANM/b29v02vPPdcGgFdeeYnatcPo2fMts/euXPk5LVs2o2nTRkydOhm9Xm9RnG3atKVevfqUK1eOoKCqDBo0lOTkZE6dOnXf98yfv4jnn3+BoKCqBAeHMH78RG7cuM6xY8cASEpK4ttvNzFo0BCeeqoejz/+OOPGTeTQoYNERR2yKL58odGguLll3YYv8kYNOft3jIozjlEN9bOW5Mh+CkCu5UyOlQwGAwaDweysjLFESdwqVOTAafMG4q+//qJ58yZ4e3vz5JNP0bdvf4oXLw7AqlVf8sYbr7Jw4RKCgqqanmWV9b79BAQEsHjxci5fvsSIEUMJCQmhY8fOQNalqM2bv+OHH7bkKWa9Xs/Gjd/g6elFcHBInsealJQMgI+PDwDHjx8jMzOTevXqm/apXLkypUqVJioqyvRcLkfJDK1J7BW5+dsSasiZIawmZy9kPUunmINjsTU11M9akiP7KQi5libHSsWKFSMsrCbLli2hSpUq+Pn5s2XLT0RFHaJ8+fKm/SIiGtKsWXPKlCnLlStXmDdvDv379+Xzz1eh0+nw9fUFoHjx4qZnSWXz8vJm+PBR6HQ6KleuTKNGjdm3b5+pySle3Jdy5co9NNY//vidkSOHkZaWRkBAIAsXLjZ97sMYjUZmzJhOrVrhVK2adbdRXFwsrq6ueHl5m+3r7+9PnMwsE0II4WByucoGJk6cgqIotGrVgvr167J29UqeDQ5Ge9ciY61ataFJk6ZUqxZM06bNmD17HkePHuGvv/Y/9PhBQUFmD8QMCAjg1q0409ddurzywIdzZnvyySf56quvWbFiJRERDRk+fIjZcR5k2rTJnD17hqlTP8rT/gWB7tRJijdvhO6UTBXNKzXkTHfqBOVaNsbVCacQq6F+1pIc2U9ByLU0OTZQvnx5li1bwZ9/7uHHH7eyZtYcDCkplC11/+dWlStXjuLFfbl8+fJDj5/zyd4aHmWhag+PolSoUIGwsJqMHTsenc6Fb7/d9ND3TZs2hR07/mDJkmWUvOtGan//APR6PUlJiWb7x8XF4e8fcO9h7E6Tlorr4UNo0mTRr7xSRc5S03A/EoXWCReDU0X9rCQ5sp+CkGtpcmzIw6MogYGBJCYlsfP6dZ6u3+C++0ZH3yAhIZ7AwKxmIPseHIPBeN/32JqiGMnIyHjA6wrTpk3h11+3s3jxMsqWNb8k9thjj+Pi4sK+fXtN2y5cOM+NG9cJC7v/1HQhhBDCHuSeHBvYtetPFEWhUqVKXL58mVkzPqKKtzcvtGwFQEpKCosXL6R58xYEBARw+fJlZs/+lPLlK9CgQUMAfH39cHd3Z9eunZQsWRI3Nze8vB7+XA6AtWu/4tdff7nvJavU1BSWLVtKkyZPExAQSHx8POvXr+XmzZs880xL0369e79N06bN6dLlFSDrEtVPP/3Ep5/OpmjRYsTGZt1n4+npibu7O15eXrRv34GZM2fg7e1DsWKeTJ8+lbCwmg6/6VgIIYSQJscGkpOTmTdvNtHR0fj4+NC8QQRD3VxxcXHBSNbDJ0+fPs33339HUlISgYElqF+/AX37vmualeXi4sLQocNZunQxixYtIDy8NkuXfpanz4+Pv82VK1fu+7pWq+PChQt8//1g4uNv4+NTnCeeeILlyz8nKKiqab8rV64QH3/b9PXXX68HyDGdfdy4iTz//AsADB48DI1Gy9Chg/5dDLAhI0d+kKe4hRBCiPwkTyHH9k8h16Slob14AWPFSiju7rYL1Enl11PINfG3cf3jN/SNn0YpnrdZZGqnhpzpEuNJ/2U7qY2exqeEv1M9hVwN9bOW5Mh+8jPXeX0KuTQ52L7JEZbJryZHiNxotRpuZxgA8HXTyfecEIVQXpscufE4H2ju3MH1r31o7txxdCiqprl5E4+F89DcvOnoUAoNNeRMc/MmxRfPQxfjfGNUQ/2sJTmyn4KQa2ly8oEmOQm3X39Bk5zk6FBUTXfjGp5jR6G7cc3RoRQaasiZ9vo1AsaPxuW6841RDfWzluTIfgpCrqXJEUIIIYRTkiZHCCGEEE5JmhwhhBBCOCVZJ8ciGjSah8/EUDw8yAwNRfHwcOQT5guR/EmS0cub9FZtMN7zAFFxf2rImeLtzZ1nWmPw9nF0KDanhvpZS3JkPwUh1zKFnIdPIddooGhRN7RaOfGVX4xGIykpGTJFX+Q7mUIuROGX1ynkciYnDxQFUlIy0OT1tIxeD0lJ4OUF/z6TSjyYoii2b3D0ejQJCSg+PlKHvFJDzvR6tHG3MHr7gJvO0dHYlhrqZy3Jkf0UgFzLqYk8UhQwGpU8/dMePYJ/cEW0R4/k+T1q/5cfZ3Bcjh8l4PEquBw/avuDOyk15Ex37ChVQqtSxAnHqIb6WUtyZD8FIdfS5AghhBDCKUmTI4QQQginJE2OEEIIIZySNDlCCCGEcEoyhZyHTyG3mMGAJuUOStFioHOy2RuFidTBcirImVYxEh+fiFK0GL4ebs41hVwF9bOa5Mh+8jHXeZ1CLk0O+dDkCCEKLFknR4jCL69Njlyuyge6c2fweak9unNnHB2KqkkdLKeGnGnPnqHMKx1xPXfW0aHYnBrqZy3Jkf0UhFxLk5MPNMnJuP22HU1ysqNDUTWpg+XUkDNNcjJFf9+ONjnJ0aHYnBrqZy3Jkf0UhFxLkyOEEEIIpyRNjhBCCCGckjQ5QgghhHBK0uTkA0OZciRNnYGhTDlHh6JqUgfLqSFnxrJliZn8MfqyzjdGNdTPWpIj+ykIuZYp5MgUciHURKaQC1H4yRRyB9LcvkWRr9eiuX3L0aGomtTBcmrImeb2LTw3rEN7+7ajQ7E5NdTPWpIj+ykIuZYmJx/oLl/Cu18vdJcvOToUVZM6WE4NOdNeukSp/r1xvXzR0aHYnBrqZy3Jkf0UhFxLkyOEEEIIpyRNjhBCCCGckjQ5QgghhHBKNm9yDAYDs2bNolmzZoSFhdGiRQvmz5/P3ZO4FEVh9uzZREZGEhYWxptvvsmFCxfMjhMfH8/gwYOpXbs2devWZdSoUdy5c8dsnxMnTvDqq68SGhpKkyZNWLp0qa2H80iUosXQ13ky68mrwmGkDpZTQ86UokVJq/MkRiccoxrqZy3Jkf0UhFzbfAr5okWLWLFiBR999BFVq1blyJEjjBw5kvfff5+uXbsCsGTJEpYsWcK0adMoV64cs2fP5tSpU/z4448UKVIEgLfffpuYmBgmTJiAXq9n1KhRhIaGMnPmTACSk5Np1aoVDRo0oHfv3pw6dYpRo0YxatQoXn75ZYtilinkQqiHTCEXovBz2BTyAwcO0Lx5c55++mnKlStH69atiYyMJCoqCsg6i7Ny5UreeecdWrRoQfXq1Zk+fTo3b95k27ZtAJw9e5YdO3YwadIkatasSd26dRk9ejQ//PAD0dHRAHz33Xfo9XqmTJlCtWrVaNu2LW+88QYrVqyw9ZCEEEIIUQjZvMkJDw9nz549nD9/Hsi6pPT333/TuHFjAK5cuUJMTAwRERGm93h5eVGzZk0OHDgAZDVK3t7ehIaGmvaJiIhAq9WamqWDBw9St25d3NzcTPtERkZy/vx5EhISbD0si7hEHSSwhDcuUQcdGofaSR0sp4ac6Q4dpGqZ4hRxwjGqoX7WkhzZT0HItYutD9irVy+Sk5Np06YNOp0Og8HA+++/z/PPPw9ATEwMAP7+/mbv8/f3JzY2FoDY2Fj8/PzMA3VxwcfHx/T+2NhYypUzXyo6ICDA9JqPj4+thyaEEEKIQsTmTc5PP/3E5s2bmTlzJlWrVuX48eNMnTqVEiVK0KFDB1t/nBBCCCFErmze5EyfPp1evXrRtm1bAEJCQrh27RqLFy+mQ4cOBAYGAhAXF0eJEiVM74uLi6N69epA1hmZW7fMl4HOzMwkISHB9P6AgADTmZ9s2V9nn9ERQgghhHrZ/J6ctLQ0NBqN2TadTmeaQl6uXDkCAwPZvXu36fXk5GQOHTpEeHg4kHVfT2JiIkeOHDHts2fPHoxGI2FhYQDUqlWLv/76C71eb9pn165dVK5cWS5VCSGEEML2TU7Tpk1ZtGgRv/32G1euXOHnn39mxYoVtGjRAgCNRkPXrl1ZuHAhv/zyCydPnmTYsGGUKFHCtE9QUBCNGjXiww8/JCoqir///puJEyfStm1bSpYsCUC7du1wdXXlgw8+4PTp0/z444+sXLmS7t2723pIFssMrk7cngNkBld3dCiqJnWwnBpyZgipzsU//yHDCceohvpZS3JkPwUh1zZfJyc5OZnZs2ezbds20yWptm3b0q9fP9NMKEVRmDNnDuvXrycxMZE6deowduxYKleubDpOfHw8EydOZPv27Wi1Wlq2bMno0aMpVuy/RYVOnDjBhAkTOHz4ML6+vrz++uv06tXL4phlnRwh1EPWyRGi8MvrOjk2b3IKI1s3OdqLFyg2bRJ3RozGWLGS7Q4sLCJ1sJwacuZy+SIukycSN2wUXlWDnKrJUUP9rCU5sp/8zLXDFgMUoE2Ix33DerQJ8Y4ORdWkDpZTQ8408fF4bVyPLj7e0aHYnBrqZy3Jkf0UhFxLkyOEEEIIpyRNjhBCCCGckjQ5QgghhHBK0uTkA2PJUtwZMgJjyVKODkXVpA6WU0POjCVLcWvQcDKdcIxqqJ+1JEf2UxByLbOrkCnkQqiJTCEXovCT2VUOpElKxHX7NjRJiY4ORdWkDpZTQ840SYkU/e0XpxyjGupnLcmR/RSEXEuTkw90589RvEtHdOfPOToUVZM6WE4NOdOeO0eZVzvh5oRjVEP9rCU5sp+CkGtpcoQQQgjhlKTJEUIIIYRTkiZHCCGEEE5Jmpx8oLgVwVCpMopbEUeHompSB8upImdFipBRqTJKEecboyrqZyXJkf0UhFzLFHJkCrkQaiJTyIUo/GQKuRBCCCFUTZqcfKA7egT/xyqjO3rE0aGomtTBcmrIme7oESrXCMLtmPONUQ31s5bkyH4KQq6lyckHGkMm2rg4NIZMR4eialIHy6kiZ5mZ6G7Focl0vjGqon5WkhzZT0HItTQ5QgghhHBK0uQIIYQQwilJkyOEEEIIpyRTyMmHKeTJybgcP0rmY0+Ap6cNDywsInWwnApypk25Q2rUYdIfe5zivj7ONYVcBfWzmuTIfvIx13mdQi5NDrJOjhBqIuvkCFH4yTo5DqS9dpViH45Ee+2qo0NRNamD5dSQM+21qwSMG4XOCceohvpZS3JkPwUh19Lk5ANtbAxFF89HGxvj6FBUTepgOTXkTBMTQ/ElC3BxwjGqoX7WkhzZT0HItTQ5QgghhHBK0uQIIYQQwilJkyOEEEIIpyRNTj4w+vmT2v1tjH7+jg5F1aQOllNDzhR/f+K7vY3BCceohvpZS3JkPwUh1zKFHJlCLoSayBRyIQo/mULuSCkpuEQdhJQUR0eiblIHy6khZykpFIk6iMYZx6iG+llLcmQ/BSDX0uTkA5czp/Bt0RiXM6ccHYqqSR0sp4ac6U6fonzrp3FzwjGqoX7WkhzZT0HItTQ5QgghhHBK0uQIIYQQwilJkyOEEEIIpyRNTj5QNFqMnl4oGkmvI0kdLKeKnGn/HaPW+caoivpZSXJkPwUh1zKFHJlCLoSayBRyIQo/mUIuhBBCCFWTJicf6E6ewLfRU+hOnnB0KKomdbCcGnKmO3mC8k/Xx/WU841RDfWzluTIfgpCrqXJyQea9DRcTp5Ak57m6FBUTepgOVXkLC2NIqdOoE1zvjGqon5WkhzZT0HItTQ5QgghhHBK0uQIIYQQwilJkyOEEEIIpyRTyLH9FHJNQjyuu3ehbxCB4lPcdgcWFpE6WE4NOdMlJZD2xw7S6kfgE+jvVFPI1VA/a0mO7Cc/c53XKeTS5CDr5AihJrJOjhCFn6yT40Ca6Gg8Zs9EEx3t6FBUTepgOTXkTBMdje/cT9DddL4xqqF+1pIc2U9ByLU0OflAF30dz8nj0UVfd3QoqiZ1sJwacqa9cR3/qRNwueF8Y1RD/awlObKfgpBraXKEEEII4ZSkyRFCCCGEU5ImRwghhBBOSZqcfGD09iG9XXuM3j6ODkXVpA6WU0POFB8fkp97AYMTTh9WQ/2sJTmyn4KQa5lCjkwhF0JNZAq5EIWfTCF3pIwMtNeuQkaGoyNRN6mD5dSQs4wMdM46RjXUz1qSI/spALmWJicfuJw4hn+tx3A5cczRoaia1MFyasiZ7vgxKtd9giJOOEY11M9akiP7KQi5liZHCCGEEE5JmhwhhBBCOCVpcoQQQgjhlKTJEUIIIYRTkink5MMUcqMR9HpwdQWt9JEOI3WwnApypkXh9p00cHXF193VuaaQq6B+VpMc2U8+5jqvU8hdbPqpIotWC0WKODoKIXWwnBpy5sxjdOax2YrkyH4KQK7zpY2Njo5myJAh1KtXj7CwMNq1a8fhw4dNryuKwuzZs4mMjCQsLIw333yTCxcumB0jPj6ewYMHU7t2berWrcuoUaO4c+eO2T4nTpzg1VdfJTQ0lCZNmrB06dL8GI7FdGdP49P+WXRnTzs6FFWTOlhODTnTnjlN2U5tcT17xtGh2Jwa6mctyZH9FIRc27zJSUhI4JVXXsHV1ZWlS5fyww8/MHz4cHx8/lvWeenSpaxatYpx48axfv16PDw86NGjB+np6aZ9hgwZwpkzZ1ixYgWLFi3ir7/+YsyYMabXk5OT6dGjB2XKlGHjxo0MGzaMefPmsW7dOlsPyWKaO3dw27UTzT1NmbAvqYPl1JAzzZ07eOz+E+2dZEeHYnNqqJ+1JEf2UxBybfPLVUuXLqVUqVJMnTrVtK18+fKm/1YUhZUrV/LOO+/QokULAKZPn05ERATbtm2jbdu2nD17lh07dvDNN98QGhoKwOjRo+nVqxfDhg2jZMmSfPfdd+j1eqZMmYKbmxvVqlXj+PHjrFixgpdfftnWwxJCCCFEIWPzMznbt2+nRo0avPfeezRo0ID27duzfv160+tXrlwhJiaGiIgI0zYvLy9q1qzJgQMHADhw4ADe3t6mBgcgIiICrVZLVFQUAAcPHqRu3bq4ubmZ9omMjOT8+fMkJCTYelhCCCGEKGRs3uRcvnyZr776ikqVKrF8+XJeeeUVJk2axKZNmwCIiYkBwN/f3+x9/v7+xMbGAhAbG4ufn5/Z6y4uLvj4+JjeHxsbS0BAgNk+2V9nH0cIIYQQ6mXzy1WKolCjRg0GDRoEwOOPP87p06dZu3YtHTp0sPXHFUiGsuVJ+mQuhrLlH76zyDdSB8upIWfGcuW4+fEc9E44RjXUz1qSI/spCLm2eZMTGBhIUFCQ2bYqVarwv//9z/Q6QFxcHCVKlDDtExcXR/Xq1YGsMzK3bt0yO0ZmZiYJCQmm9wcEBOQ4Y5P99b1neOxN8fcn7fVuDo1BSB0ehRpypvgHkPhaV0eHkS/UUD9rSY7spyDk2uaXq2rXrs358+fNtl24cIGyZcsCUK5cOQIDA9m9e7fp9eTkZA4dOkR4eDgA4eHhJCYmcuTIEdM+e/bswWg0EhYWBkCtWrX466+/0Ov1pn127dpF5cqVzWZyOYImLg731V+giYtzaBxqJ3WwnBpypomLxXvNSrROOEY11M9akiP7KQi5tnmT061bNw4dOsSiRYu4ePEimzdvZv369bz66qsAaDQaunbtysKFC/nll184efIkw4YNo0SJEqbZVkFBQTRq1IgPP/yQqKgo/v77byZOnEjbtm0pWbIkAO3atcPV1ZUPPviA06dP8+OPP7Jy5Uq6d+9u6yFZTHf1Ml6D+qO7etnRoaia1MFyasiZ9soVSgx9D1cnHKMa6mctyZH9FIRc2/xyVVhYGPPmzeOTTz5h/vz5lCtXjlGjRvH888+b9unZsyepqamMGTOGxMRE6tSpw7Jlyyhy18qIM2bMYOLEiXTr1g2tVkvLli0ZPXq06XUvLy+WL1/OhAkT6NixI76+vvTt21emjwshhBACyKfHOjRt2pSmTZve93WNRsOAAQMYMGDAffcpXrw4M2fOfODnVK9enS+//PKR4xRCCCGE85KnkwkhhBDCKUmTkw+UYsXIiIhEKVbM0aGomtTBcmrImVKsGKkNGmIs5unoUGxODfWzluTIfgpCrjWKoigO+/QCIjY2CcmCEOqg1Wq4nWEAwNdNh9Eo/+cXorDRaCAgwOuh+8mZnPxgNEJ6etb/CseROlhODTlz5jE689hsRXJkPwUg19Lk5AOXI1EElg/E5UiUo0NRNamD5dSQM93hKKpWLkkRJxyjGupnLcmR/RSEXEuTI4QQQginJE2OEEIIIZySNDlCCCGEcErS5AghhBDCKckUcvJhCnlGBtrYGIwBgeDmZsMDC4tIHSyngpxpM/UkXruBISAQX08P55pCroL6WU1yZD/5mOu8TiGXJgdZJ0cINZF1coQo/GSdHAfSXjiPd4+uaC+cd3QoqiZ1sJwacqa9cJ5SvbrhcvGCo0OxOTXUz1qSI/spCLmWJicfaBMTKLL5W7SJCY4ORdWkDpZTQ840CQl4fv9/6BLiHR2KzamhftaSHNlPQci1NDlCCCGEcErS5AghhBDCKUmTI4QQQginJE1OPjCULE3yB2MxlCzt6FBUTepgOTXkzFiqNHEjx5BZyvnGqIb6WUtyZD8FIdcyhRyZQi6EmsgUciEKP5lC7kCahHjctvyIxglnbxQmUgfLqSFnmoR4iv7vR7ROOEY11M9akiP7KQi5liYnH+guXsCnaxd0TrgOR2EidbCcGnKmvXCBMt1fxdUJx6iG+llLcmQ/BSHX0uQIIYQQwilJkyOEEEIIpyRNjhBCCCGckjQ5+UAp4k5mSHWUIu6ODkXVpA6WU0XO3N1JD66O0d35xqiK+llJcmQ/BSHXMoUcmUIuhJrIFHIhCj+ZQi6EEEIIVZMmJx/oDkfhX6UsusNRjg5F1aQOllNDznSHo6gSXB63I843RjXUz1qSI/spCLmWJicfaBQj2uQkNIrR0aGomtTBcqrImfHfMRqdb4yqqJ+VJEf2UxByLU2OEEIIIZySNDlCCCGEcErS5AghhBDCKckUcvJhCnlKCi5nTpFZNRiKFrXhgYVFpA6WU0HOtGmppBw7TkbVYIoX93KuKeQqqJ/VJEf2k4+5zusUcmlykHVyhFATWSdHiMJP1slxIO2Vy3gOH4T2ymVHh6JqUgfLqSFn2iuXCRg5BBcnHKMa6mctyZH9FIRcS5OTD7S34vBYsQztrThHh6JqUgfLqSFnmrg4in+xDJ0TjlEN9bOW5Mh+CkKupckRQgghhFOSJkcIIYQQTkmaHCGEEEI4JWly8oExIJCU3v0wBgQ6OhRVkzpYTg05UwIDie/Vl0wnHKMa6mctyZH9FIRcyxRyZAq5EGoiU8iFKPxkCrkjJSfjsn8vJCc7OhJ1kzpYTg05S07G/a99aO444RjVUD9rSY7spwDkWpqcfOBy7gy+bZ/B5dwZR4eialIHy6khZ7qzZyj3fEvczjrfGNVQP2tJjuynIORamhwhhBBCOCVpcoQQQgjhlKTJEUIIIYRTkiYnHyg6F4z+/ig6F0eHompSB8upImcuLhj8/FFcnG+MqqiflSRH9lMQci1TyJEp5EKoiUwhF6LwkynkQgghhFA1aXLyge7EcfyeqonuxHFHh6JqUgfLqSFnuhPHqRARjttJ5xujGupnLcmR/RSEXEuTkw80GenoLpxHk5Hu6FBUTepgOVXkLD0dtwvn0aQ73xhVUT8rSY7spyDkWpocIYQQQjglaXKEEEII4ZSkyRFCCCGEU5Ip5Nh+CrkmKRGX/fvIfPIpFC9v2x1YWETqYDk15Ex3J4m03XtIrfMkxf19nWoKuRrqZy3Jkf3kZ67zOoVcmhxknRwh1ETWyRGi8JN1chxIG32DotOnoI2+4ehQVE3qYDk15Exz4wZ+M6aic8IxqqF+1pIc2U9ByLU0OflAG32DYjOmyf+JHEzqYDk15EwbfQO/Tz7CxQnHqIb6WUtyZD8FIdfS5AghhBDCKeV7k7NkyRJCQkKYPHmyaVt6ejrjx4+nXr16hIeH079/f2JjY83ed+3aNXr16kXNmjVp0KABH330EZmZmWb77N27lw4dOlCjRg2eeeYZNm7cmN/DEUIIIUQhka9NTlRUFGvXriUkJMRs+5QpU/j111+ZNWsWq1at4ubNm7z77rum1w0GA71790av17N27VqmTZvGpk2bmDNnjmmfy5cv07t3b+rVq8f//d//0a1bN0aPHs2OHTvyc0hCCCGEKCTyrcm5c+cOQ4cOZdKkSfj4+Ji2JyUlsWHDBkaMGEGDBg2oUaMGU6ZM4cCBAxw8eBCAnTt3cubMGT7++GMee+wxmjRpwoABA1izZg0ZGRkArF27lnLlyjFixAiCgoJ4/fXXadWqFZ9//nl+DSnPjD7FSev0Ekaf4o4ORdWkDpZTQ86U4sVJ6vgShuLFHR2KzamhftaSHNlPQch1vjU5EyZMoEmTJkRERJhtP3LkCHq93mx7UFAQZcqUMTU5Bw8eJDg4mICAANM+kZGRJCcnc+bMGdM+DRo0MDt2ZGSk6RiOZKxYiaSFyzBWrOToUFRN6mA5NeTMWLES0fOWkFmhkqNDsTk11M9akiP7KQi5zpcm54cffuDYsWMMHjw4x2uxsbG4urri7W2+MJC/vz8xMTGmfe5ucADT1w/bJzk5mbS0NJuN5ZGkpaE9dxYcHYfaSR0sp4acpaXhev4cGmccoxrqZy3Jkf0UgFzbvMm5fv06kydP5uOPP6ZIkSK2Pnyh4HLqBP71w3E5dcLRoaia1MFyasiZ7uQJKjasjZsTjlEN9bOW5Mh+CkKuXWx9wKNHjxIXF0fHjh1N2wwGA/v372fNmjUsX74cvV5PYmKi2dmcuLg4AgMDgawzMlFRUWbHzZ59dfc+987Iio2NxdPTE3d3d1sPSwghhBCFjM2bnPr167N582azbSNHjqRKlSr07NmT0qVL4+rqyu7du2nVqhUA586d49q1a9SqVQuAWrVqsWjRIuLi4vD39wdg165deHp6UrVqVdM+f/zxh9nn7Nq1y3QMIYQQQqibzZscT09PgoODzbYVLVqU4sWLm7Z36tSJadOm4ePjg6enJ5MmTSI8PNzUoERGRlK1alWGDRvG0KFDiYmJYdasWbz22mu4ubkB0KVLF9asWcP06dPp1KkTe/bs4aeffmLx4sW2HpIQQgghCiGbNzl5MWrUKLRaLe+99x4ZGRlERkYyduxY0+s6nY5FixYxbtw4Xn75ZTw8POjQoQPvvfeeaZ/y5cuzePFipk6dysqVKylVqhSTJk2iUaNGjhiSEEIIIQoYeQo58hRyIdREnkIuROEnTyEXQgghhKpJk5MPdGdOU7xNc3RnTjs6FFWTOlhODTnTnj5FuXbP4OqEY1RD/awlObKfgpBraXLygSblDq5/70eTcsfRoaia1MFyasiZJiUF97/3o3XCMaqhftaSHNlPQci1NDlCCCGEcErS5AghhBDCKUmTI4QQQginJE1OPjCUr0Di/CUYyldwdCiqJnWwnBpyZqxQgRtzF6MvX9HRodicGupnLcmR/RSEXMs6Ocg6OUKoiayTI0ThJ+vkOJAmNhb35UvQ3PMAUWFfUgfLqSFnmtgYfFYsRRvnfGNUQ/2sJTmyn4KQa2ly8oHu2hW8Rg5Bd+2Ko0NRNamD5dSQM+3VqwR+MBTXq843RjXUz1qSI/spCLmWJkcIIYQQTkmaHCGEEEI4JWlyhBBCCOGUpMnJB4qnJxlPN0Px9HR0KKomdbCcGnKmeHqS0qQZRs+Hz8wobNRQP2tJjuynIORappAjU8iFUBOZQi5E4SdTyB3JYECTlAgGg6MjUTepg+XUkDNnHqMzj81WJEf2UwByLU1OPnA5epiAoHK4HD3s6FBUTepgOTXkTHfkMEEhFSjihGNUQ/2sJTmyn4KQa2lyhBBCCOGUpMkRQgghhFOSJkcIIYQQTkmaHCGEEEI4JZlCTj5MIdfr0SQkoPj4gKurDQ8sLCJ1sJwKcqY1ZJIQewujtw++xdydawq5CupnNcmR/eRjrvM6hVyaHGSdHCHURNbJEaLwk3VyHEh7/hzeb7yM9vw5R4eialIHy6khZ9rz5yjdrQsuF847OhSbU0P9rCU5sp+CkGtpcvKBNimRIv/7CW1SoqNDUTWpg+XUkDNNYiLFft6CLjHB0aHYnBrqZy3Jkf0UhFxLkyOEEEIIpyRNjhBCCCGckjQ5QgghhHBK0uTkA0OpMiSPn4KhVBlHh6JqUgfLqSFnxtJliB07iczSzjdGNdTPWpIj+ykIuZYp5MgUciHURKaQC1H4yRRyB9LE38btu01o4m87OhRVkzpYTg0508Tfptjmb9HGxzs6FJtTQ/2sJTmyn4KQa2ly8oHu0kV83u6G7tJFR4eialIHy6khZ9qLFynd+01cL11wdCg2p4b6WUtyZD8FIdfS5AghhBDCKUmTI4QQQginJE2OEEIIIZySNDn5QHH3QB9aE8Xdw9GhqJrUwXKqyJmHO2k1wjB6ON8YVVE/K0mO7Kcg5FqmkCNTyIVQE5lCLkThJ1PIhRBCCKFq0uTkA5fDhwgoF4DL4UOODkXVpA6WU0POdFGHCKpUAjcnHKMa6mctyZH9FIRcS5OTHxQFTUYGcg3MwaQOllNDzv4do8YZx6iG+llLcmQ/BSDX0uQIIYQQwilJkyOEEEIIpyRNjhBCCCGckkwhJx+mkKemort4AUPFSuCEa3EUGlIHy6kgZ9r0NO6cOYu+QiWK+3g61xRyFdTPapIj+8nHXOd1Crk0Ocg6OUKoiayTI0ThJ+vkOJD28iU8338X7eVLjg5F1aQOllNDzrSXLxE4uD8uV5xvjGqon7UkR/ZTEHItTU4+0N6+hcealWhv33J0KKomdbCcGnKmuXULn69WobvlfGNUQ/2sJTmyn4KQa2lyhBBCCOGUpMkRQgghhFOSJkcIIYQQTkmanHxgDCxBynuDMAaWcHQoqiZ1sJwacqaUKMHtd98n0wnHqIb6WUtyZD8FIdcyhRyZQi6EmsgUciEKP5lC7kCa5CRc/9yBJjnJ0aGomtTBcqrIWVISHrucc4yqqJ+VJEf2UxByLU1OPtCdO0vxDm3RnTvr6FBUTepgOTXkTHfuLGU7t8PNCceohvpZS3JkPwUh19LkCCGEEMIpSZMjhBBCCKckTY4QQgghnJI0OflAcXHFULoMiouro0NRNamD5VSRM1dXMkuXQXF1vjGqon5WkhzZT0HItc2nkC9evJitW7dy7tw53N3dCQ8PZ8iQIVSpUsW0T3p6OtOmTePHH38kIyODyMhIxo4dS0BAgGmfa9euMW7cOPbu3UvRokVp3749gwcPxsXFxbTP3r17mTZtGqdPn6Z06dK88847dOzY0eKYZQq5EOohU8iFKPwcNoV83759vPbaa6xfv54VK1aQmZlJjx49SElJMe0zZcoUfv31V2bNmsWqVau4efMm7777rul1g8FA79690ev1rF27lmnTprFp0ybmzJlj2ufy5cv07t2bevXq8X//939069aN0aNHs2PHDlsPSQghhBCFkM2bnOXLl9OxY0eqVatG9erVmTZtGteuXePo0aMAJCUlsWHDBkaMGEGDBg2oUaMGU6ZM4cCBAxw8eBCAnTt3cubMGT7++GMee+wxmjRpwoABA1izZg0ZGRkArF27lnLlyjFixAiCgoJ4/fXXadWqFZ9//rmth2Qx3bGj+NWsju7YUUeHompSB8upIWe6Y0epVOdx3I473xjVUD9rSY7spyDkOt/vyUlKyloEyMfHB4AjR46g1+uJiIgw7RMUFESZMmVMTc7BgwcJDg42u3wVGRlJcnIyZ86cMe3ToEEDs8+KjIw0HcORNJl6dNevocnUOzoUVZM6WE4VOdPrcbl+DY3e+caoivpZSXJkPwUh1/na5BiNRqZMmULt2rUJDg4GIDY2FldXV7y9vc329ff3JyYmxrTP3Q0OYPr6YfskJyeTlpaWL+MRQgghROHh8vBdHt348eM5ffo0X375ZX5+jBBCCCFEDvl2JmfChAn89ttvfPHFF5QqVcq0PSAgAL1eT2Jiotn+cXFxBAYGmvaJjY01ez3764ft4+npibu7u83HI4QQQojCxeZNjqIoTJgwgZ9//pkvvviC8uXLm71eo0YNXF1d2b17t2nbuXPnuHbtGrVq1QKgVq1anDp1iri4ONM+u3btwtPTk6pVq5r22bNnj9mxd+3aZTqGIxmqBBG/6QcMVYIcHYqqSR0sp4acGaoEcfWbzWQ44RjVUD9rSY7spyDk2ubr5IwbN47vv/+eBQsWULlyZdN2Ly8v0xmWsWPH8scffzB16lQ8PT2ZNGkSkDVjCrKmkLdv354SJUowdOhQYmJiGDZsGC+++CKDBg0CsqaQt2vXjldffZVOnTqxZ88eJk+ezOLFi2nUqJFFMcs6OUKoh6yTI0Thl9d1cmze5ISEhOS6ferUqaaF+rIXA/zhhx/MFgPMvhQFcPXqVcaNG8e+ffvw8PCgQ4cOuS4GOHXqVM6cOUOpUqXo27dvgVgMUHv9Gh7Ll5DaoxfG0mVsd2BhEamD5dSQM5fo67B4EfHde+JdsbxTNTlqqJ+1JEf2k5+5dliTUxjZuslxiTqIb4vG3N72B5lhtWx3YGERqYPl1JAz18OHKN68EZe3/EbRunWcqslRQ/2sJTmyn/zMtcNWPBZCCCGEKAikyRFCCCGEU5ImRwghhBBOSZqcfGD09SP1ta4Yff0cHYqqSR0sp4acKX5+JLzyBgY/5xujGupnLcmR/RSEXMuNx8gUciHURKaQC1H4yY3HjpSaiu7EcUhNdXQk6iZ1sJwacpaaitvJ42iccYxqqJ+1JEf2UwByLU1OPnA5fRK/xvVwOX3S0aGomtTBcmrIme7USSo0bYCbE45RDfWzluTIfgpCrqXJEUIIIYRTkiZHCCGEEE5JmhwhhBBCOCVpcvKDRoPi5pZ1+7dwHKmD5dSQs3/HqDjjGNVQP2tJjuynAORappAjU8iFUBOZQi5E4SdTyIUQQgihatLk5APdqZMUb94I3SmZouhIUgfLqSFnulMnKNeyMa5OOIVYDfWzluTIfgpCrqXJyQeatFRcDx9CkyaLTTmS1MFyqshZahruR6LQOuFicKqon5UkR/ZTEHItTY4QQgghnJI0OUIIIYRwStLkCCGEEMIpSZOTDwwVKpKw7AsMFSo6OhRVkzpYTg05M1asyPXFn6OvUMnRodicGupnLcmR/RSEXMs6Ocg6OUKoiayTI0ThJ+vkOJDm5k08Fs5Dc/Omo0NRNamD5dSQM83NmxRfPA9djPONUQ31s5bkyH4KQq6lyckHuhvX8Bw7Ct2Na44ORdWkDpZTQ860168RMH40Ltedb4xqqJ+1JEf2UxByLU2OEEIIIZySNDlCCCGEcErS5AghhBDCKUmTkw+MXt6kt2qD0cvb0aGomtTBcmrImeLtzZ1nWmPw9nF0KDanhvpZS3JkPwUh1zKFHJlCLoSayBRyIQo/mULuSHo9mthY0OsdHYm6SR0sp4ac6fVo45x0jGqon7UkR/ZTAHItTU4+cDl+lIDHq+By/KijQ1E1qYPl1JAz3bGjVAmtShEnHKMa6mctyZH9FIRcS5MjhBBCCKckTY4QQgghnJI0OUIIIYRwStLkCCGEEMIpyRRy8mEKucGAJuUOStFioNPZ8MDCIlIHy6kgZ1rFSHx8IkrRYvh6uDnXFHIV1M9qkiP7ycdc53UKuTQ5yDo5QqiJrJMjROEn6+Q4kO7cGXxeao/u3BlHh6JqUgfLqSFn2rNnKPNKR1zPnTVt02iy/t1No8lqiO7dXpCpoX7WkhzZT0HItTQ5+UCTnIzbb9vRJCc7OhRVkzpYrrDn7O5m5b7/nZxM0d+3o01OMr2m12jRa7Rm++s1Wm5nGMy2F3SFvX72IDmyn4KQa2lyhBBOQauFzH+bFa32v8bl7v/OrVnRaDTcycjkTkYmGo3m34Yoa9uec3Gm7UKIwsfF0QEIIYS1NBrI1OhITM9Ei0KRIi7cSc8EMP23VgNF3HQ53gcas68z0KLVaAAD6XoD2nvOCAFyD58QhYScyRFCFCi53R/z8PdoSNFnkmk0cnfTkk2rAa1OR0aOhkZDgt6A0bQ96wxOij4T0OCq05re97AzQkKIgkeanHxgKFOOpKkzMJQp5+hQVE3qYDlH5+x+98fcr6nIa0Ok0WhIycgkOd2AUq4cMZM/xlC2PMnpBv65ePu+zZGL7r/3aTRa02UtrVZjdlNybjcpP0qzZi1H168wkBzZT0HItUwhR6aQC1FQ3Du9W1EU9Jqsv8VcFSOKgumeGVDI+PfvNDcUEjONpGYacdNq8Sui49a/l6v8iriQoDfkeE2n1WAwwr7zsdSq6Jfn17QaKOrmQnJ6JsXcXHDDSAZZDVAxNxdcFSNAjriFELYjU8gdSHP7FkW+Xovm9i1Hh6JqUgfLFbSc3XtTcPbNxbczDGRqdGav5fmYt2/huWEd2kcco0ajITndwJ5zcaTqM9FqtaablLO+1uSI214KWv0KIsmR/RSEXEuTkw90ly/h3a8XusuXHB2KqkkdLOeonGVf7smmNZ2tMd8nU6MjMSOrwci+b8ZS2kuXKNW/Ny5WjtFoVNDqdOj/jSH76wxyxm2v9Xbke/7hJEf2UxByLU2OEMKh7l6TJlOjNd0kfCs96+u79jTdXJyeaXRYvNmy79e5k2EANPfcv5PV0Wj/vbn5doaBTG3WdHYhhP3I/+WEEA5195o0KfpMNBotKRmZ7D3/6GdrCorsS1t/X7xFmkGRmVlC2Jk0OUIIu7j7ss3d/zI1WoxoSNcbzPZPu+frwsxgVHKc5YHC+egIIQoTaXLygVK0GPo6T2Y9eVU4jNTBcvmVM7MbhrVa08rEZuvb2Ok3vVK0KGkO/L7Ibmyy192x5aMj5Hv+4SRH9lMQci0rHucDQ9VqxP/0i6PDUD2pg+VsmbO7f2lnr0a871wcDasGYDAC/65MbG/GasFc2fwzOq0G7HxrT/Y9OskZmXgVceVOup495+KICPKniJsOg8G6uebyPf9wkiP7KQi5ljM5QgibyDpDYX6W4t6zNQXhhmFHunv6efb9RnfPyrp3McH7XcpyxEKDQhRG0uTkA5eogwSW8MYl6qCjQ1E1qYPl8pKze++tyW5qMrVa0q1cw8YedIcOUrVMcdwc+H1x9/1Huc3KMptx9u+srHsvc2VtN2+C5Hv+4SRH9lMQci1NjhAih+yF9+5uZLLP1Nx7b012U5OqNxCfqmfv+VuFflaUo/yX66wZZ9mzsnLmOpM0g8LtDON97+eRG5qFkCZHCFW6+6zAvY0MQHKmgXSNFsPdv1y1Wgx3LcaXqjeQnMvCfM40K8qezNbU+Xd9oOxZWSkZRrNc3z3N/u5naWUz3HVDc3aNhVAjaXKEUIHc7pdJ0OdsZAz//nI9eSOZ+FQ9Kfr/frmm6g3c0Rty3Ftz79Rv8Whyu1/nXrlNs7+7OTL8u9rgHf1/j524+4Gnd38f3D2NXwhnJU2OEE5Oo8n9fpkDl+JzNDJpmVm/RDMM5r9MpZGxH0tzfXdzlP5v/QyKQnqm0fRa9tkes+8Ds6n8+TESIRxPppDng8zg6sTtOYCxTFlHh6JqUocsGo2G1H+fwn3g4m0igvwBjekXYrZ0vQF9cHUu/vkPCenujgnWDgwhWWNUyjrX90V2/S7vPkB6YCm4knzPHjm/DwxG0GoUiri5YDSq42np8nPBfgpCrqXJyQ/u7hirBDk6CiF1yOFh98so7u5kVg7CeD7WThE5gLs7+spVHLJOTn7Lrp+SaQDubXL+k/19cPdzwoq56XBFBY2O/FywnwKQa7lclQ+0Fy/g9c7baC9ecHQoqiZ1sJzLxQuUfLcnxa5ddnQo+UZ78QIl3+2FixN+X7hcvEBgv7dxvXQxT/trNBqzG5gL2nT//CA/F+ynIOS60Dc5a9asoVmzZoSGhvLiiy8SFRXl6JDQJsTjvmE92oR4R4eialIHy2kT4vHauB63pARHhwJknWmw9e9dTXzWGJ3x+0KbEI/nhvXoLBybmmbEyc8F+ykIuS7UTc6PP/7I1KlT6devH5s2baJ69er06NGDuLg4R4cmhLCSq06LVqejsP36zWrMnP+MiBCFQaFuclasWMFLL71Ep06dqFq1KuPHj8fd3Z0NGzY4OjQhhJVcdBpSMzLJNCgUloUFs+5x0RKfYcBYSGIWwpkV2huPMzIyOHr0KL179zZt02q1REREcODAAYuOZfM/unQ68PICnU6mZjqS1AG4a7E/DXi6u6DVglGBYm4uWb+U//3a090FrZKVMw/3Ijlf04Ki5DzOg17T/HtWI7fPv99r2ccxxfdvHHk9zt375vqaS9YYtS5Z3xd358G0by45etBr2V+7aDWk6Q1ciEmmUmAxtFodRkXJPdcPqsNDXrtvHf6tn0ansyjXnu4u6lk3R34u2E8+5jqvx9MoSuG8lz46OprGjRuzdu1awsPDTdunT5/O/v37+frrrx0YnRBCCCEcrVBfrhJCCCGEuJ9C2+T4+vqi0+ly3GQcFxdHQECAg6ISQgghREFRaJscNzc3nnjiCXbv3m3aZjQa2b17t9nlKyGEEEKoU6G98Rige/fuDB8+nBo1ahAWFsYXX3xBamoqHTt2dHRoQgghhHCwQt3kPPvss9y6dYs5c+YQExPDY489xrJly+RylRBCCCEK7+wqIYQQQogHKbT35AghhBBCPIg0OUIIIYRwStLkCCGEEMIpSZMjhBBCCKckTY6T6dOnD08//TShoaFERkYydOhQoqOjHR2WQ1y5coVRo0bRrFkzwsLCaNGiBXPmzCEjI8PRoTnMwoUL6dKlCzVr1qRu3bqODsch1qxZQ7NmzQgNDeXFF18kKirK0SE5zP79++nTpw+RkZGEhISwbds2R4fkUIsXL6ZTp06Eh4fToEED+vbty7lz5xwdlsN8+eWXtGvXjtq1a1O7dm1efvllfv/9d0eHZRFpcpxM/fr1mTVrFlu2bGHOnDlcvnyZAQMGODoshzh37hyKojBhwgR++OEHRo4cydq1a/n0008dHZrD6PV6WrduzSuvvOLoUBzixx9/ZOrUqfTr149NmzZRvXp1evTokWPldLVISUkhJCSEsWPHOjqUAmHfvn289tprrF+/nhUrVpCZmUmPHj1ISUlxdGgOUapUKYYMGcLGjRvZsGED9evXp1+/fpw+fdrRoeWdIpzatm3blJCQECUjI8PRoRQIS5cuVZo1a+boMBxuw4YNSp06dRwdht117txZGT9+vOlrg8GgREZGKosXL3ZgVAVDcHCw8vPPPzs6jAIlLi5OCQ4OVvbt2+foUAqMJ598Ulm/fr2jw8gzOZPjxOLj49m8eTPh4eG4uro6OpwCISkpCR8fH0eHIRwgIyODo0ePEhERYdqm1WqJiIjgwIEDDoxMFFRJSUkA8jMDMBgM/PDDD6SkpBSqRycV6hWPRe4+/vhj1qxZQ2pqKrVq1WLRokWODqlAuHjxIqtXr2b48OGODkU4wO3btzEYDPj7+5tt9/f3V/V9FyJ3RqORKVOmULt2bYKDgx0djsOcPHmSLl26kJ6eTtGiRZk/fz5Vq1Z1dFh5Jk1OITBjxgyWLl36wH1+/PFHgoKCAOjRowedO3fm2rVrzJs3j+HDh7N48WI0Go09ws13luYDIDo6mrfffpvWrVvz0ksv5XeIdvUo+RBCPNj48eM5ffo0X375paNDcajKlSvz7bffkpSUxP/+9z+GDx/O6tWrC02jI01OIfDWW2/RoUOHB+5Tvnx503/7+fnh5+dH5cqVCQoKokmTJhw8eLBQnWJ8EEvzER0dTdeuXQkPD2fixIn5HZ7dWZoPtfL19UWn0+W4yTguLk6edyfMTJgwgd9++43Vq1dTqlQpR4fjUG5ublSsWBGAGjVqcPjwYVauXMmECRMcHFneSJNTCGQ3LY/CaDQCONW0aUvykd3gPPHEE0ydOhWt1vluQ7Pm+0NN3NzceOKJJ9i9ezctWrQAsv7/sXv3bl5//XUHRycKAkVRmDhxIj///DOrVq2SPw5yYTQaC9XvE2lynMihQ4c4fPgwderUwdvbm0uXLjF79mwqVKjgNGdxLBEdHc0bb7xBmTJlGD58OLdu3TK9FhgY6MDIHOfatWskJCRw7do1DAYDx48fB6BChQoUK1bMwdHlv+7duzN8+HBq1KhBWFgYX3zxBampqXTs2NHRoTnEnTt3uHTpkunrK1eucPz4cXx8fChTpowDI3OM8ePH8/3337NgwQKKFStGTEwMAF5eXri7uzs4OvubOXMmjRs3pnTp0ty5c4fvv/+effv2sXz5ckeHlmfyFHIncvLkSSZPnszJkydJSUkhMDCQRo0a0bdvX0qWLOno8Oxu48aNjBw5MtfXTp48aedoCoYRI0awadOmHNtXrlxJvXr1HBCR/a1evZrly5cTExPDY489xujRo6lZs6ajw3KIvXv30rVr1xzbO3TowLRp0xwQkWOFhITkun3q1KmqbIRHjRrFnj17uHnzJl5eXoSEhNCzZ08aNmzo6NDyTJocIYQQQjgl57tBQQghhBACaXKEEEII4aSkyRFCCCGEU5ImRwghhBBOSZocIYQQQjglaXKEEEII4ZSkyRFCCCGEU5ImRwghhBBOSZocIYQQQjglaXKEEEII4ZSkyRFCCCGEU5ImRwghhBBO6f8BTOTZwWvcRjwAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "lookup_percentiles = [5, 10, 20, 50, 80, 90, 95]\n",
    "df[\"shimmer_score_diff\"] = df[\"total_shimmer_score\"].diff()\n",
    "percentiles = np.percentile(\n",
    "    df[df[\"preference\"]][\"shimmer_score_diff\"].dropna(), lookup_percentiles\n",
    ")\n",
    "plt.hist(\n",
    "    df[df[\"preference\"]][\"shimmer_score_diff\"],\n",
    "    label=f\"pos, mean: {np.mean(df[df['preference']]['shimmer_score_diff']):.2f}\",\n",
    "    bins=np.linspace(-1, 1, 100),\n",
    "    alpha=0.5,\n",
    ")\n",
    "textstr = \"\\n\".join(\n",
    "    [\n",
    "        f\"{lookup_percentiles[i]}th: {percentile:.2f}\"\n",
    "        for i, percentile in enumerate(percentiles)\n",
    "    ]\n",
    ")\n",
    "plt.gcf().text(\n",
    "    0.15,\n",
    "    0.98,\n",
    "    textstr,\n",
    "    fontsize=10,\n",
    "    verticalalignment=\"top\",\n",
    "    horizontalalignment=\"left\",\n",
    "    bbox=dict(facecolor=\"white\", alpha=0.5),\n",
    ")\n",
    "for percentile in percentiles:\n",
    "    plt.axvline(x=percentile, color=\"r\", linestyle=\"dashed\", linewidth=1)\n",
    "plt.title(\n",
    "    f\"Shimmer score difference --> {lookup_percentiles[-1]}th, {percentiles[-1]:.2f}\"\n",
    ")\n",
    "# plt.yscale(\"log\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:25.050717Z",
     "iopub.status.busy": "2025-02-20T15:27:25.050546Z",
     "iopub.status.idle": "2025-02-20T15:27:25.304365Z",
     "shell.execute_reply": "2025-02-20T15:27:25.303938Z",
     "shell.execute_reply.started": "2025-02-20T15:27:25.050703Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "lookup_percentiles = [5, 10, 20, 50, 80, 90, 95]\n",
    "percentiles = np.percentile(\n",
    "    df[df[\"preference\"]][\"pair_quality\"].dropna(), lookup_percentiles\n",
    ")\n",
    "plt.hist(\n",
    "    df[df[\"preference\"]][\"pair_quality\"],\n",
    "    label=f\"pos, mean: {np.mean(df[df['preference']]['pair_quality']):.2f}\",\n",
    "    bins=np.linspace(0, 1, 100),\n",
    "    alpha=0.5,\n",
    ")\n",
    "textstr = \"\\n\".join(\n",
    "    [\n",
    "        f\"{lookup_percentiles[i]}th: {percentile:.2f}\"\n",
    "        for i, percentile in enumerate(percentiles)\n",
    "    ]\n",
    ")\n",
    "plt.gcf().text(\n",
    "    0.15,\n",
    "    0.98,\n",
    "    textstr,\n",
    "    fontsize=10,\n",
    "    verticalalignment=\"top\",\n",
    "    horizontalalignment=\"left\",\n",
    "    bbox=dict(facecolor=\"white\", alpha=0.5),\n",
    ")\n",
    "for percentile in percentiles:\n",
    "    plt.axvline(x=percentile, color=\"r\", linestyle=\"dashed\", linewidth=1)\n",
    "plt.title(f\"Pair quality --> {lookup_percentiles[0]}th, {percentiles[0]:.2f}\")\n",
    "# plt.yscale(\"log\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:25.305024Z",
     "iopub.status.busy": "2025-02-20T15:27:25.304858Z",
     "iopub.status.idle": "2025-02-20T15:27:25.727649Z",
     "shell.execute_reply": "2025-02-20T15:27:25.727218Z",
     "shell.execute_reply.started": "2025-02-20T15:27:25.305011Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# CS said -1 means stero to mono; 1 means mono to stereo\n",
    "# only cut off the left side\n",
    "df[\"stereo_width_diff\"] = df[\"stereo_width\"].diff()\n",
    "lookup_percentiles = [5, 10, 20, 50, 80, 90, 95]\n",
    "percentiles = np.percentile(\n",
    "    df[df[\"preference\"]][\"stereo_width_diff\"].dropna(), lookup_percentiles\n",
    ")\n",
    "plt.hist(\n",
    "    df[df[\"preference\"]][\"stereo_width_diff\"],\n",
    "    label=f\"pos, mean: {np.mean(df[df['preference']]['stereo_width_diff']):.2f}\",\n",
    "    bins=np.linspace(-1, 1, 400),\n",
    "    alpha=0.5,\n",
    ")\n",
    "textstr = \"\\n\".join(\n",
    "    [\n",
    "        f\"{lookup_percentiles[i]}th: {percentile:.2f}\"\n",
    "        for i, percentile in enumerate(percentiles)\n",
    "    ]\n",
    ")\n",
    "plt.gcf().text(\n",
    "    0.15,\n",
    "    0.98,\n",
    "    textstr,\n",
    "    fontsize=10,\n",
    "    verticalalignment=\"top\",\n",
    "    horizontalalignment=\"left\",\n",
    "    bbox=dict(facecolor=\"white\", alpha=0.5),\n",
    ")\n",
    "for percentile in percentiles:\n",
    "    plt.axvline(x=percentile, color=\"r\", linestyle=\"dashed\", linewidth=1)\n",
    "plt.title(\n",
    "    f\"Stereo Width Difference --> {lookup_percentiles[0]}th, {percentiles[0]:.2f}\"\n",
    ")\n",
    "# plt.yscale(\"log\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:25.728322Z",
     "iopub.status.busy": "2025-02-20T15:27:25.728154Z",
     "iopub.status.idle": "2025-02-20T15:27:26.898230Z",
     "shell.execute_reply": "2025-02-20T15:27:26.897796Z",
     "shell.execute_reply.started": "2025-02-20T15:27:25.728308Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# CS said no tails is good\n",
    "lookup_percentiles = [5, 10, 20, 50, 80, 90, 95]\n",
    "percentiles = np.percentile(\n",
    "    df[df[\"preference\"]][\"spectral_centroid\"].dropna(), lookup_percentiles\n",
    ")\n",
    "percentiles = np.percentile(\n",
    "    df[~df[\"preference\"]][\"spectral_centroid\"].dropna(), lookup_percentiles\n",
    ")\n",
    "plt.hist(\n",
    "    df[df[\"preference\"]][\"spectral_centroid\"],\n",
    "    label=f\"pos, mean: {np.mean(df[df['preference']]['spectral_centroid']):.2f}\",\n",
    "    bins=np.linspace(0, 8000, 400),\n",
    "    alpha=0.5,\n",
    ")\n",
    "plt.hist(\n",
    "    df[~df[\"preference\"]][\"spectral_centroid\"],\n",
    "    label=f\"neg, mean: {np.mean(df[~df['preference']]['spectral_centroid']):.2f}\",\n",
    "    bins=np.linspace(0, 8000, 400),\n",
    "    alpha=0.5,\n",
    ")\n",
    "textstr = \"\\n\".join(\n",
    "    [\n",
    "        f\"{lookup_percentiles[i]}th: {percentile:.2f}\"\n",
    "        for i, percentile in enumerate(percentiles)\n",
    "    ]\n",
    ")\n",
    "plt.gcf().text(\n",
    "    0.15,\n",
    "    0.98,\n",
    "    textstr,\n",
    "    fontsize=10,\n",
    "    verticalalignment=\"top\",\n",
    "    horizontalalignment=\"left\",\n",
    "    bbox=dict(facecolor=\"white\", alpha=0.5),\n",
    ")\n",
    "for percentile in percentiles:\n",
    "    plt.axvline(x=percentile, color=\"r\", linestyle=\"dashed\", linewidth=1)\n",
    "plt.title(\"Spectral Centroid\")\n",
    "# plt.yscale(\"log\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:26.898919Z",
     "iopub.status.busy": "2025-02-20T15:27:26.898730Z",
     "iopub.status.idle": "2025-02-20T15:27:27.320843Z",
     "shell.execute_reply": "2025-02-20T15:27:27.320407Z",
     "shell.execute_reply.started": "2025-02-20T15:27:26.898905Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# CS said no tails is good\n",
    "# take the relative centroid diff\n",
    "df[\"spectral_centroid_diff\"] = df[\"spectral_centroid\"].diff() / df[\"spectral_centroid\"]\n",
    "lookup_percentiles = [5, 10, 20, 50, 80, 90, 95]\n",
    "percentiles = np.percentile(\n",
    "    df[df[\"preference\"]][\"spectral_centroid_diff\"].dropna(), lookup_percentiles\n",
    ")\n",
    "plt.hist(\n",
    "    df[df[\"preference\"]][\"spectral_centroid_diff\"],\n",
    "    label=f\"pos, mean: {np.mean(df[df['preference']]['spectral_centroid_diff']):.2f}\",\n",
    "    bins=np.linspace(-1, 1, 400),\n",
    "    alpha=0.5,\n",
    ")\n",
    "textstr = \"\\n\".join(\n",
    "    [\n",
    "        f\"{lookup_percentiles[i]}th: {percentile:.2f}\"\n",
    "        for i, percentile in enumerate(percentiles)\n",
    "    ]\n",
    ")\n",
    "plt.gcf().text(\n",
    "    0.15,\n",
    "    0.98,\n",
    "    textstr,\n",
    "    fontsize=10,\n",
    "    verticalalignment=\"top\",\n",
    "    horizontalalignment=\"left\",\n",
    "    bbox=dict(facecolor=\"white\", alpha=0.5),\n",
    ")\n",
    "for percentile in percentiles:\n",
    "    plt.axvline(x=percentile, color=\"r\", linestyle=\"dashed\", linewidth=1)\n",
    "plt.title(\n",
    "    f\"Spectral Centroid Difference ratio --> {lookup_percentiles[-1]}th, {percentiles[-1]:.2f}\"\n",
    ")\n",
    "# plt.yscale(\"log\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:27.323430Z",
     "iopub.status.busy": "2025-02-20T15:27:27.323229Z",
     "iopub.status.idle": "2025-02-20T15:27:27.590087Z",
     "shell.execute_reply": "2025-02-20T15:27:27.589589Z",
     "shell.execute_reply.started": "2025-02-20T15:27:27.323416Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "good_continue_at\n",
      "True    115538\n",
      "Name: count, dtype: int64\n",
      "\n",
      " Check some basics... \n",
      " preference\n",
      "False    57769\n",
      "True     57769\n",
      "Name: count, dtype: int64 model_name\n",
      "chirp-v4-up-u-4    115538\n",
      "Name: count, dtype: int64 preference  model_name     \n",
      "False       chirp-v4-up-u-4    57769\n",
      "True        chirp-v4-up-u-4    57769\n",
      "Name: count, dtype: int64\n",
      "task\n",
      "upsample    115538\n",
      "Name: count, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "df = df.loc[:, ~df.columns.duplicated()].copy()\n",
    "# get the original duration of the clips, if they are concacted\n",
    "df[\"original_duration_s\"] = df[\"total_start_s\"] + df[\"duration\"]\n",
    "# classify the continue at behavoirs by the duration choice\n",
    "audio_prompt_id_to_continue_at = {}\n",
    "for _, row in df[~df[\"continued_parent\"].isna()].iterrows():\n",
    "    audio_prompt_id = row[\"continued_parent\"]\n",
    "    if audio_prompt_id not in audio_prompt_id_to_continue_at:\n",
    "        audio_prompt_id_to_continue_at[audio_prompt_id] = row[\"continue_at\"]\n",
    "    else:\n",
    "        # pick the max\n",
    "        audio_prompt_id = max(\n",
    "            audio_prompt_id_to_continue_at[audio_prompt_id], row[\"continue_at\"]\n",
    "        )\n",
    "print(len(audio_prompt_id_to_continue_at))\n",
    "df[\"has_continue_and_start_continue_at\"] = df[\"s3_id\"].apply(\n",
    "    lambda x: audio_prompt_id_to_continue_at.get(x)\n",
    ")\n",
    "# we want continue at to be at most of the clip...\n",
    "df[\"good_continue_at\"] = (\n",
    "    (df[\"has_continue_and_start_continue_at\"] / df[\"duration\"]) > 0.9\n",
    ") | df[\"has_continue_and_start_continue_at\"].isna()\n",
    "print(df[\"good_continue_at\"].value_counts())\n",
    "\n",
    "\n",
    "print(\n",
    "    \"\\n Check some basics... \\n\",\n",
    "    df[\"preference\"].value_counts(),\n",
    "    df[\"model_name\"].value_counts(),\n",
    "    df.groupby([\"preference\"])[\"model_name\"].value_counts(),\n",
    ")\n",
    "\n",
    "df = df.sort_values(by=[\"request_id\", \"preference\"])\n",
    "df[\"duration_rel_diff\"] = df[\"duration\"].diff()\n",
    "df[\"play_rel_diff\"] = df[\"reaction_play_count\"].diff()\n",
    "print(df[\"task\"].value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:27.590788Z",
     "iopub.status.busy": "2025-02-20T15:27:27.590601Z",
     "iopub.status.idle": "2025-02-20T15:27:27.939683Z",
     "shell.execute_reply": "2025-02-20T15:27:27.939254Z",
     "shell.execute_reply.started": "2025-02-20T15:27:27.590775Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(\n",
    "    df[df[\"preference\"]][\"total_shimmer_score\"],\n",
    "    label=f\"pos, mean: {np.mean(df[df['preference']]['total_shimmer_score']):.2f}\",\n",
    "    bins=np.linspace(0, 10, 100),\n",
    "    alpha=0.5,\n",
    ")\n",
    "plt.hist(\n",
    "    df[~df[\"preference\"]][\"total_shimmer_score\"],\n",
    "    label=f\"neg, mean: {np.mean(df[~df['preference']]['total_shimmer_score']):.2f}\",\n",
    "    bins=np.linspace(0, 10, 100),\n",
    "    alpha=0.5,\n",
    ")\n",
    "percentiles = np.percentile(df[df[\"preference\"]][\"total_shimmer_score\"].dropna(), [50, 75, 90])\n",
    "for percentile in percentiles:\n",
    "    # print(percentile)\n",
    "    plt.axvline(x=percentile, color=\"r\", linestyle=\"dashed\", linewidth=1)\n",
    "# plt.yscale(\"log\")\n",
    "plt.title(f\"Shimmer score\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:59:41.035167Z",
     "start_time": "2024-05-16T13:59:40.801098Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:27.940368Z",
     "iopub.status.busy": "2025-02-20T15:27:27.940193Z",
     "iopub.status.idle": "2025-02-20T15:27:28.040017Z",
     "shell.execute_reply": "2025-02-20T15:27:28.039515Z",
     "shell.execute_reply.started": "2025-02-20T15:27:27.940354Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "negative 53417 positive 34763\n",
      "total pair requests 57769 selected pair requests 31797 frac 0.550\n"
     ]
    }
   ],
   "source": [
    "normal_pos_play_count = 3\n",
    "# this is lower, cause a concat is probably already ensuring that it is good\n",
    "concat_pos_play_count = 2\n",
    "# this is a filter on the concated clip\n",
    "concat_total_play_count = 3\n",
    "\n",
    "neg_filter_selection_mask = (\n",
    "    (~df[\"preference\"])  # get basics aligned\n",
    "    & (df[\"reaction_play_count\"] >= 1)  # has to be played once\n",
    "    # & (df[\"play_count\"] <= 3)  # if it is actually bad, shouldn't be listened often\n",
    "    & (df[\"duration\"] >= 30)  # can't be too short, otherwise it is obvious\n",
    "    # & (df[\"duration\"] <= 60)  # can't be badly long\n",
    "    & (df[\"has_continue_and_start_continue_at\"].isna())  # won't have any continues\n",
    "    & (df[\"norm_play_frac\"] <= 2.1)\n",
    "    # & (df[\"sum_total_play_duration_5\"] >= 31)\n",
    "    # & (df[\"dislike_count\"] >= 1) # this is kinda strict\n",
    "    #     & (\n",
    "    #         (df_slice[\"is_in_playlist\"] == False)\n",
    "    #         & (df_slice[\"concat_in_playlist\"] == False)\n",
    "    #     )  # can't be part of a playlist -- otherwise there are some like signal in it?\n",
    ")\n",
    "pos_filter_selectin_mask = (\n",
    "    (df[\"preference\"])  # get basics aligned\n",
    "    & (\n",
    "        df[\"good_continue_at\"]\n",
    "    )  # if continue, needs to continue off a certain percentage\n",
    "    & (df[\"reaction_play_count\"] >= 1)\n",
    "    & (df[\"play_rel_diff\"] >= 0)  # this is more like quality assurance\n",
    "    & (df[\"duration\"] >= 30)  # can't be too short, otherwise it is obvious\n",
    "    & (df[\"dislike_count\"] == 0)  # can't have dislikes\n",
    "    & (df[\"flag_count\"] == 0)  # can't have issues\n",
    "    & (\n",
    "        (\n",
    "            (df[\"part_of_concat\"])\n",
    "            & (df[\"reaction_play_count\"] >= concat_pos_play_count)\n",
    "            & (df[\"concat_play_counts\"] >= concat_total_play_count)\n",
    "        )\n",
    "        | (\n",
    "            (~df[\"part_of_concat\"])\n",
    "            & (df[\"reaction_play_count\"] >= normal_pos_play_count)\n",
    "            & (df[\"norm_play_frac\"] >= 2.1)  # this is a bit of a luxury cut...\n",
    "            & (df[\"sum_total_play_duration_5\"] >= 31)\n",
    "        )\n",
    "    )\n",
    "    # & (df[\"norm_play_frac\"] >= 1.9)\n",
    "    # & (df[\"user_n_clips\"] >= 100)  # user needs to have genereated at least 20\n",
    "    # & (df[\"duration_rel_diff\"] < 10) # positive isn't just longer\n",
    "    # & ((df[\"task\"] == \"\") | (df[\"task\"] == \"extend\"))\n",
    "    & (\n",
    "        (df[\"upvote_count\"] >= 1)\n",
    "        | (df[\"reaction_play_count\"] >= 5)\n",
    "        | (df[\"concat_play_counts\"] >= 5)\n",
    "    )\n",
    "    # & (df[\"pos_diff_preference\"] == 2)\n",
    "    # & ((0 < df[\"similarity\"]) &  (df[\"similarity\"] <= 0.99))\n",
    "    # & (\n",
    "    #     (df[\"cer_diff_preference\"] < 0.25) & (df[\"cer\"] < 0.8)\n",
    "    # )  # cut on hoot cer difference and abs cer\n",
    "    # & (df[\"pair_quality\"] > 0.31)  # bottom 5%\n",
    "    # & ((df[\"total_shimmer_score\"] < 1) | (df[\"shimmer_score_diff\"] < 0.4))\n",
    "    # & (df[\"stereo_width_diff\"] > -0.2)  # cut off bottom 5%\n",
    "    # & (df[\"spectral_centroid_diff\"] < 0.25)  # crop off the top 5%\n",
    ")\n",
    "print(\n",
    "    \"negative\",\n",
    "    sum(neg_filter_selection_mask),\n",
    "    \"positive\",\n",
    "    sum(pos_filter_selectin_mask),\n",
    ")\n",
    "\n",
    "neg_filter_requests = df[neg_filter_selection_mask][\"request_id\"].unique()\n",
    "pos_filter_requests = df[pos_filter_selectin_mask][\"request_id\"].unique()\n",
    "# looking for very strong signal here:\n",
    "# listen to the positive/negative more than once\n",
    "# disliked one of the clips\n",
    "unique_requests = set(pos_filter_requests).intersection(neg_filter_requests)\n",
    "print(\n",
    "    \"total pair requests\",\n",
    "    df[\"request_id\"].nunique(),\n",
    "    \"selected pair requests\",\n",
    "    len(unique_requests),\n",
    "    f\"frac {len(unique_requests) / df['request_id'].nunique():.3f}\",\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:59:41.250737Z",
     "start_time": "2024-05-16T13:59:41.036434Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:28.040703Z",
     "iopub.status.busy": "2025-02-20T15:27:28.040534Z",
     "iopub.status.idle": "2025-02-20T15:27:28.146357Z",
     "shell.execute_reply": "2025-02-20T15:27:28.145846Z",
     "shell.execute_reply.started": "2025-02-20T15:27:28.040690Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " requests 31797 clips 63594 total khrs 3.384; N gpus for 1000 iters 3.975; 4 gpus for x iters 993.656; n unique users 16398 n pro users 15325\n"
     ]
    }
   ],
   "source": [
    "df_slice = df[df[\"request_id\"].isin(set(unique_requests))].copy()\n",
    "print(\n",
    "    f\"{os.path.basename(OUT_DATA_DIR)} requests\",\n",
    "    df_slice[\"request_id\"].nunique(),\n",
    "    \"clips\",\n",
    "    df_slice.shape[0],\n",
    "    f\"total khrs {sum(df_slice['duration'] / 3600 / 1000):.3f};\",\n",
    "    f\"N gpus for 1000 iters {df_slice.shape[0] / 8 / 2 / 1000:.3f};\",\n",
    "    f\"4 gpus for x iters {df_slice.shape[0] / 8 / 2 / 4:.3f};\",\n",
    "    f\"n unique users {df_slice['user_id'].nunique()}\",\n",
    "    f\"n pro users {df_slice[df_slice['is_pro_user']]['user_id'].nunique()}\",\n",
    ")\n",
    "# up t7 requests 37735 clips 75470 total khrs 3.964; N gpus for 1000 iters 4.717; 4 gpus for x iters 1179.219; n unique users 19093 n pro users 16217\n",
    "# up t17 requests 49262 clips 98524 total khrs 5.187; N gpus for 1000 iters 6.158; 4 gpus for x iters 1539.438; n unique users 24093 n pro users 20326\n",
    "# up v2 t1 requests 6008 clips 12016 total khrs 0.642; N gpus for 1000 iters 0.751; 4 gpus for x iters 187.750; n unique users 3846 n pro users 3645\n",
    "# up v2 t2 requests 10772 clips 21544 total khrs 1.154; N gpus for 1000 iters 1.347; 4 gpus for x iters 336.625; n unique users 6527 n pro users 6027\n",
    "# up v3 t10 requests 27102 clips 54204 total khrs 2.938; N gpus for 1000 iters 3.388; 4 gpus for x iters 846.938; n unique users 13932 n pro users 12187\n",
    "# up v4 t1  requests 3201 clips 6402 total khrs 0.343; N gpus for 1000 iters 0.400; 4 gpus for x iters 100.031; n unique users 2363 n pro users 2321\n",
    "# up v4 t2  requests 12818 clips 25636 total khrs 1.371; N gpus for 1000 iters 1.602; 4 gpus for x iters 400.562; n unique users 7753 n pro users 7524\n",
    "# up v4 t3  requests 15332 clips 30664 total khrs 1.637; N gpus for 1000 iters 1.917; 4 gpus for x iters 479.125; n unique users 8960 n pro users 8656\n",
    "# up v4 t4  requests 18368 clips 36736 total khrs 1.962; N gpus for 1000 iters 2.296; 4 gpus for x iters 574.000; n unique users 10427 n pro users 10049\n",
    "# up v4 t5  requests 22037 clips 44074 total khrs 2.347; N gpus for 1000 iters 2.755; 4 gpus for x iters 688.656; n unique users 12121 n pro users 11551\n",
    "# up v4 t6  requests 27374 clips 54748 total khrs 2.915; N gpus for 1000 iters 3.422; 4 gpus for x iters 855.438; n unique users 14463 n pro users 13678\n",
    "# up v4 t7  requests 31797 clips 63594 total khrs 3.384; N gpus for 1000 iters 3.975; 4 gpus for x iters 993.656; n unique users 16398 n pro users 15325"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:59:41.277006Z",
     "start_time": "2024-05-16T13:59:41.252105Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:28.147061Z",
     "iopub.status.busy": "2025-02-20T15:27:28.146874Z",
     "iopub.status.idle": "2025-02-20T15:27:28.158294Z",
     "shell.execute_reply": "2025-02-20T15:27:28.157885Z",
     "shell.execute_reply.started": "2025-02-20T15:27:28.147047Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "positive in playlist (8682, 118)\n"
     ]
    }
   ],
   "source": [
    "test_mask = (df_slice[\"preference\"]) & (\n",
    "    (df_slice[\"is_in_playlist\"]) | (df_slice[\"concat_in_playlist\"])\n",
    ")\n",
    "print(\"positive in playlist\", df_slice[test_mask].shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:59:41.392244Z",
     "start_time": "2024-05-16T13:59:41.324472Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:28.158964Z",
     "iopub.status.busy": "2025-02-20T15:27:28.158748Z",
     "iopub.status.idle": "2025-02-20T15:27:28.210703Z",
     "shell.execute_reply": "2025-02-20T15:27:28.210283Z",
     "shell.execute_reply.started": "2025-02-20T15:27:28.158951Z"
    }
   },
   "outputs": [],
   "source": [
    "# save positive ids\n",
    "# positive_preference_ids = df_slice[df_slice[\"preference\"] == False][\"s3_id\"].to_json(orient='values')\n",
    "# with open('/home/tony/Data/Preference/7b_v2/7v_v20_full_recut_id_negative.json', 'w') as file:\n",
    "#     file.write(positive_preference_ids)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:28.211396Z",
     "iopub.status.busy": "2025-02-20T15:27:28.211170Z",
     "iopub.status.idle": "2025-02-20T15:27:28.283818Z",
     "shell.execute_reply": "2025-02-20T15:27:28.283355Z",
     "shell.execute_reply.started": "2025-02-20T15:27:28.211382Z"
    }
   },
   "outputs": [],
   "source": [
    "# def modify_model_name(model_name, metadata):\n",
    "#     if (\n",
    "#         model_name.startswith(\"chirp-v3p5-engine-t\")\n",
    "#         or model_name.startswith(\"chirp-v3p5-engine-s\")\n",
    "#         or model_name.startswith(\"chirp-v4\")\n",
    "#         or model_name.startswith(\"chirp-v3p5-h-s-31\")\n",
    "#     ):\n",
    "#         if \"param_experiment\" in metadata:\n",
    "#             exp = metadata.get(\"param_experiment\", \"\")\n",
    "#             if exp:\n",
    "#                 return f\"{model_name}_{exp}\"\n",
    "#     return model_name\n",
    "\n",
    "# metrics_check_df_slice = df_slice.copy()\n",
    "# metrics_check_df_slice[\"model_name\"] = metrics_check_df_slice.apply(\n",
    "#     lambda row: modify_model_name(row[\"model_name\"], row[\"metadata\"]), axis=1\n",
    "# )\n",
    "# get_preference_counts(metrics_check_df_slice)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:28.284463Z",
     "iopub.status.busy": "2025-02-20T15:27:28.284304Z",
     "iopub.status.idle": "2025-02-20T15:27:28.363906Z",
     "shell.execute_reply": "2025-02-20T15:27:28.363430Z",
     "shell.execute_reply.started": "2025-02-20T15:27:28.284450Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "source\n",
      "web        60220\n",
      "ios         2360\n",
      "android     1014\n",
      "Name: count, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print(df_slice[\"source\"].value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T14:00:20.866354Z",
     "start_time": "2024-05-16T14:00:12.443344Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:28.364640Z",
     "iopub.status.busy": "2025-02-20T15:27:28.364387Z",
     "iopub.status.idle": "2025-02-20T15:27:28.592939Z",
     "shell.execute_reply": "2025-02-20T15:27:28.592362Z",
     "shell.execute_reply.started": "2025-02-20T15:27:28.364626Z"
    }
   },
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'BREAK' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[29], line 2\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[38;5;66;03m# df_slice.to_csv(\"/home/tony/Data/Preference/13b_v0/interesting_clips_v3p5_s_8_20240828_slice.csv\")\u001b[39;00m\n\u001b[0;32m----> 2\u001b[0m \u001b[43mBREAK\u001b[49m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'BREAK' is not defined"
     ]
    }
   ],
   "source": [
    "# df_slice.to_csv(\"/home/tony/Data/Preference/13b_v0/interesting_clips_v3p5_s_8_20240828_slice.csv\")\n",
    "BREAK"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Need to kick out the ones has gpt prompt -- these are pairs with different text inputs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:59:41.932966Z",
     "start_time": "2024-05-16T13:59:41.932957Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:41.564580Z",
     "iopub.status.busy": "2025-02-20T15:27:41.564244Z",
     "iopub.status.idle": "2025-02-20T15:27:41.573413Z",
     "shell.execute_reply": "2025-02-20T15:27:41.573004Z",
     "shell.execute_reply.started": "2025-02-20T15:27:41.564565Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "31797\n"
     ]
    }
   ],
   "source": [
    "final_filtered_requests = df_slice[\"request_id\"].unique()\n",
    "print(len(final_filtered_requests))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:59:41.934277Z",
     "start_time": "2024-05-16T13:59:41.934268Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:41.574292Z",
     "iopub.status.busy": "2025-02-20T15:27:41.574068Z",
     "iopub.status.idle": "2025-02-20T15:27:41.897464Z",
     "shell.execute_reply": "2025-02-20T15:27:41.896951Z",
     "shell.execute_reply.started": "2025-02-20T15:27:41.574278Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "31479 318\n",
      "(62958, 118) (636, 118)\n"
     ]
    }
   ],
   "source": [
    "train_requests, val_requests = train_test_split(\n",
    "    sorted(list(final_filtered_requests)), test_size=0.01, random_state=42\n",
    ")\n",
    "print(len(train_requests), len(val_requests))\n",
    "\n",
    "train_df = df_slice[df_slice[\"request_id\"].isin(set(train_requests))].copy()\n",
    "val_df = df_slice[df_slice[\"request_id\"].isin(set(val_requests))].copy()\n",
    "train_df = train_df.sort_values(by=[\"request_id\", \"preference\"])\n",
    "train_df = train_df  # .reset_index()\n",
    "val_df = val_df.sort_values(by=[\"request_id\", \"preference\"])\n",
    "val_df = val_df  # .reset_index()\n",
    "\n",
    "print(train_df.shape, val_df.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:41.898133Z",
     "iopub.status.busy": "2025-02-20T15:27:41.897953Z",
     "iopub.status.idle": "2025-02-20T15:27:41.903563Z",
     "shell.execute_reply": "2025-02-20T15:27:41.903226Z",
     "shell.execute_reply.started": "2025-02-20T15:27:41.898119Z"
    }
   },
   "outputs": [],
   "source": [
    "# BREAK"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Actually make"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:59:41.934954Z",
     "start_time": "2024-05-16T13:59:41.934946Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:41.904041Z",
     "iopub.status.busy": "2025-02-20T15:27:41.903939Z",
     "iopub.status.idle": "2025-02-20T15:27:41.993239Z",
     "shell.execute_reply": "2025-02-20T15:27:41.992746Z",
     "shell.execute_reply.started": "2025-02-20T15:27:41.904030Z"
    }
   },
   "outputs": [],
   "source": [
    "# val_df[[\"request_id\", \"metadata\", \"updated_at\", \"user_id\", \"preference\"]].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:59:41.935620Z",
     "start_time": "2024-05-16T13:59:41.935613Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:41.994430Z",
     "iopub.status.busy": "2025-02-20T15:27:41.994257Z",
     "iopub.status.idle": "2025-02-20T15:27:43.797563Z",
     "shell.execute_reply": "2025-02-20T15:27:43.797055Z",
     "shell.execute_reply.started": "2025-02-20T15:27:41.994417Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████████████████████████████████████████████████████████████████| 62958/62958 [00:01<00:00, 36790.67it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3,350 hours of 62958 clips, 3.934875 nodes, 983.71875 steps\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "total_duration = 0\n",
    "for i, row in tqdm(train_df.iterrows(), total=len(train_df)):\n",
    "    # we need to alternate between preference: neg, pos\n",
    "    # print(i, row)\n",
    "    try:\n",
    "        assert row[\"preference\"] == (i % 2 == 1)\n",
    "    except:\n",
    "        print(i, row)\n",
    "    total_duration += row[\"duration\"]\n",
    "print(\n",
    "    f\"{round(total_duration / 60 / 60):,} hours of {train_df.shape[0]} clips, {train_df.shape[0] / 8 / 2 / 1000} nodes, {train_df.shape[0] / 8 / 2 / 4} steps\"\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:59:41.936268Z",
     "start_time": "2024-05-16T13:59:41.936260Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:43.798221Z",
     "iopub.status.busy": "2025-02-20T15:27:43.798048Z",
     "iopub.status.idle": "2025-02-20T15:27:52.981056Z",
     "shell.execute_reply": "2025-02-20T15:27:52.980556Z",
     "shell.execute_reply.started": "2025-02-20T15:27:43.798208Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "df shape: (636, 118)\n",
      "total chunks: 4\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 4/4 [00:09<00:00,  2.29s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Done! val: wrote 1935000 semantic tokens and 247680000 vae latents. \n",
      "Total slices of data: 2580. Per node: 80.6. \n",
      "Passed quality check: 2580, Failed quality check: 1198. \n",
      "Total chunks with prev chunk as vae ctx: 1654.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "make_dataset(\n",
    "    val_df,\n",
    "    OUT_DATA_DIR,\n",
    "    is_val=True,\n",
    "    npz_dir=NPZ_DIR,\n",
    "    do_extend_chunks=True,\n",
    "    clip_id_to_quality_scores=unpacked_pair_quality,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:59:41.936964Z",
     "start_time": "2024-05-16T13:59:41.936957Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:27:52.981756Z",
     "iopub.status.busy": "2025-02-20T15:27:52.981584Z",
     "iopub.status.idle": "2025-02-20T15:38:16.562695Z",
     "shell.execute_reply": "2025-02-20T15:38:16.562346Z",
     "shell.execute_reply.started": "2025-02-20T15:27:52.981743Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "df shape: (62958, 118)\n",
      "total chunks: 315\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████| 315/315 [10:23<00:00,  1.98s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Done! tr: wrote 185485500 semantic tokens and 23742144000 vae latents. \n",
      "Total slices of data: 247314. Per node: 7728.6. \n",
      "Passed quality check: 247314, Failed quality check: 124600. \n",
      "Total chunks with prev chunk as vae ctx: 158210.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "make_dataset(\n",
    "    train_df,\n",
    "    OUT_DATA_DIR,\n",
    "    is_val=False,\n",
    "    npz_dir=NPZ_DIR,\n",
    "    do_extend_chunks=True,\n",
    "    clip_id_to_quality_scores=unpacked_pair_quality,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-01-29T19:46:47.549860Z",
     "start_time": "2024-01-29T19:46:47.548015Z"
    }
   },
   "source": [
    "# Validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:59:41.937879Z",
     "start_time": "2024-05-16T13:59:41.937870Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:16.563288Z",
     "iopub.status.busy": "2025-02-20T15:38:16.563134Z",
     "iopub.status.idle": "2025-02-20T15:38:16.604041Z",
     "shell.execute_reply": "2025-02-20T15:38:16.603741Z",
     "shell.execute_reply.started": "2025-02-20T15:38:16.563275Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1290.0\n",
      "(2580, 750, 128)\n",
      "(2580, 750)\n"
     ]
    }
   ],
   "source": [
    "# verify\n",
    "metas_val = read_jsonl(os.path.join(OUT_DATA_DIR, \"metas_val.jsonl\"))\n",
    "print(len(metas_val) / 2)\n",
    "mm_semantic_val = np.memmap(\n",
    "    os.path.join(OUT_DATA_DIR, \"data_semantic_val.bin\"), dtype=np.uint16, mode=\"r\"\n",
    ")\n",
    "mm_vae_val = np.memmap(\n",
    "    os.path.join(OUT_DATA_DIR, \"data_vae_val.bin\"), dtype=np.float16, mode=\"r\"\n",
    ")\n",
    "\n",
    "\n",
    "mm_vae_val = mm_vae_val.reshape(-1, VAE_MEMMAP_SIZE, VAE_DIM)\n",
    "print(mm_vae_val.shape)\n",
    "\n",
    "mm_semantic_val = mm_semantic_val.reshape(-1, SEMANTIC_MEMMAP_SIZE)\n",
    "print(mm_semantic_val.shape)\n",
    "\n",
    "assert len(metas_val) == mm_vae_val.shape[0] == mm_semantic_val.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:16.604569Z",
     "iopub.status.busy": "2025-02-20T15:38:16.604377Z",
     "iopub.status.idle": "2025-02-20T15:38:16.617413Z",
     "shell.execute_reply": "2025-02-20T15:38:16.617158Z",
     "shell.execute_reply.started": "2025-02-20T15:38:16.604557Z"
    }
   },
   "outputs": [],
   "source": [
    "# # load codec for decoding\n",
    "# from suno_utils.tasks.dac_vae_100hz_peaq import (  # NOTE: works for 25hz as well\n",
    "#     preload_models as preload_codec_models,\n",
    "#     decode as codec_decode,\n",
    "#     encode as codec_encode,\n",
    "#     get_embedding_rate,\n",
    "#     load_model as load_codec_model,\n",
    "# )\n",
    "\n",
    "# CODEC_FILEPATH = \"s3://suno-data/christian/25hz_vae_peaq_kl_0.005.pth\"\n",
    "# preload_codec_models(CODEC_FILEPATH)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:16.617910Z",
     "iopub.status.busy": "2025-02-20T15:38:16.617722Z",
     "iopub.status.idle": "2025-02-20T15:38:16.661295Z",
     "shell.execute_reply": "2025-02-20T15:38:16.660978Z",
     "shell.execute_reply.started": "2025-02-20T15:38:16.617898Z"
    }
   },
   "outputs": [],
   "source": [
    "# # decode some audio\n",
    "idx = 108\n",
    "# # ensure even index\n",
    "assert idx % 2 == 0\n",
    "# print(metas_val[idx])\n",
    "# print(\"negative\")\n",
    "# audio = codec_decode(mm_vae_val[idx])\n",
    "# audio.normalize_volume().play()\n",
    "\n",
    "# print(metas_val[idx + 1])\n",
    "# print(\"positive\")\n",
    "# audio = codec_decode(mm_vae_val[idx + 1])\n",
    "# audio.normalize_volume().play()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:16.661823Z",
     "iopub.status.busy": "2025-02-20T15:38:16.661634Z",
     "iopub.status.idle": "2025-02-20T15:38:18.818187Z",
     "shell.execute_reply": "2025-02-20T15:38:18.817662Z",
     "shell.execute_reply.started": "2025-02-20T15:38:16.661812Z"
    }
   },
   "outputs": [],
   "source": [
    "import torch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:18.818902Z",
     "iopub.status.busy": "2025-02-20T15:38:18.818727Z",
     "iopub.status.idle": "2025-02-20T15:38:18.845298Z",
     "shell.execute_reply": "2025-02-20T15:38:18.844947Z",
     "shell.execute_reply.started": "2025-02-20T15:38:18.818889Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "torch.equal(torch.tensor(mm_semantic_val[idx]), torch.tensor(mm_semantic_val[idx + 1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:59:41.941167Z",
     "start_time": "2024-05-16T13:59:41.941159Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:18.845913Z",
     "iopub.status.busy": "2025-02-20T15:38:18.845684Z",
     "iopub.status.idle": "2025-02-20T15:38:18.866711Z",
     "shell.execute_reply": "2025-02-20T15:38:18.866356Z",
     "shell.execute_reply.started": "2025-02-20T15:38:18.845900Z"
    }
   },
   "outputs": [],
   "source": [
    "# original_npz_path = f\"/app/suno/data/dpo/7b_npz/{test_metas[idx]['id']}.npz\"\n",
    "# original_npz_path = \"/app/suno/data/dpo/7b_npz/729c3011-f672-4ccd-8d82-1cbf2b52ff69.npz\"\n",
    "# original_arr = np.load(original_npz_path)[\"v2_raw\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:59:41.941801Z",
     "start_time": "2024-05-16T13:59:41.941793Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:18.868545Z",
     "iopub.status.busy": "2025-02-20T15:38:18.868306Z",
     "iopub.status.idle": "2025-02-20T15:38:18.912989Z",
     "shell.execute_reply": "2025-02-20T15:38:18.912572Z",
     "shell.execute_reply.started": "2025-02-20T15:38:18.868532Z"
    }
   },
   "outputs": [],
   "source": [
    "def validation_on_metas(input_metas):\n",
    "    total_bad = 0\n",
    "    total_good = 0\n",
    "    for idx in range(len(input_metas)):\n",
    "        if idx % 2 == 0:\n",
    "            pos_idx = idx + 1\n",
    "            if input_metas[idx].get(\"tags\") != input_metas[pos_idx].get(\"tags\"):\n",
    "                # print(test_metas[idx].get(\"text\") == test_metas[pos_idx].get(\"text\"), test_metas[idx].get(\"tags\"), test_metas[pos_idx].get(\"tags\"))\n",
    "                total_bad += 1\n",
    "            else:\n",
    "                total_good += 1\n",
    "    print(total_good, total_bad)\n",
    "    return"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:18.913591Z",
     "iopub.status.busy": "2025-02-20T15:38:18.913480Z",
     "iopub.status.idle": "2025-02-20T15:38:23.612954Z",
     "shell.execute_reply": "2025-02-20T15:38:23.612419Z",
     "shell.execute_reply.started": "2025-02-20T15:38:18.913579Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "123657 0\n",
      "247314\n"
     ]
    }
   ],
   "source": [
    "metas_tr = read_jsonl(os.path.join(OUT_DATA_DIR, \"metas_tr.jsonl\"))\n",
    "validation_on_metas(metas_tr)\n",
    "print(len(metas_tr))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:23.613629Z",
     "iopub.status.busy": "2025-02-20T15:38:23.613459Z",
     "iopub.status.idle": "2025-02-20T15:38:23.702426Z",
     "shell.execute_reply": "2025-02-20T15:38:23.701984Z",
     "shell.execute_reply.started": "2025-02-20T15:38:23.613615Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10536"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum(len(meta[\"tags\"][0]) == 0 for meta in metas_tr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-05-16T13:59:41.945972Z",
     "start_time": "2024-05-16T13:59:41.945964Z"
    },
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:23.703075Z",
     "iopub.status.busy": "2025-02-20T15:38:23.702910Z",
     "iopub.status.idle": "2025-02-20T15:38:23.963798Z",
     "shell.execute_reply": "2025-02-20T15:38:23.963184Z",
     "shell.execute_reply.started": "2025-02-20T15:38:23.703062Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Submitted batch job 3482\n"
     ]
    }
   ],
   "source": [
    "!cd /home/tony/Work/tony/slurm/diffusion && sbatch run_diffusion.sh"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:23.964735Z",
     "iopub.status.busy": "2025-02-20T15:38:23.964519Z",
     "iopub.status.idle": "2025-02-20T15:38:23.989060Z",
     "shell.execute_reply": "2025-02-20T15:38:23.988639Z",
     "shell.execute_reply.started": "2025-02-20T15:38:23.964720Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cache kept!\n"
     ]
    }
   ],
   "source": [
    "import shutil\n",
    "\n",
    "# Basic file copy\n",
    "shutil.copy(\n",
    "    \"/home/tony/Work/tony/Preference/make_dataset_diff_upsample_v1_r3_comb.ipynb\",\n",
    "    os.path.join(OUT_DATA_DIR, \"make_dataset.ipynb\"),\n",
    ")\n",
    "print(\"Cache kept!\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Inspections "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:23.989799Z",
     "iopub.status.busy": "2025-02-20T15:38:23.989553Z",
     "iopub.status.idle": "2025-02-20T15:38:24.013874Z",
     "shell.execute_reply": "2025-02-20T15:38:24.013481Z",
     "shell.execute_reply.started": "2025-02-20T15:38:23.989785Z"
    }
   },
   "outputs": [],
   "source": [
    "# df[df[\"preference\"] & (df[\"shimmer_score_diff\"] > 3)][\n",
    "#     [\n",
    "#         \"index\",\n",
    "#         \"s3_id\",\n",
    "#         \"total_shimmer_score\",\n",
    "#         \"shimmer_score_diff\",\n",
    "#         \"request_id\",\n",
    "#         \"preference\",\n",
    "#     ]\n",
    "# ].tail()\n",
    "\n",
    "# df[df[\"preference\"] & (df[\"pair_quality\"] < 0.1)][\n",
    "#     [\n",
    "#         \"index\",\n",
    "#         \"s3_id\",\n",
    "#         \"total_shimmer_score\",\n",
    "#         \"shimmer_score_diff\",\n",
    "#         \"pair_quality\",\n",
    "#         \"request_id\",\n",
    "#         \"preference\",\n",
    "#     ]\n",
    "# ].tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:24.014669Z",
     "iopub.status.busy": "2025-02-20T15:38:24.014368Z",
     "iopub.status.idle": "2025-02-20T15:38:24.053554Z",
     "shell.execute_reply": "2025-02-20T15:38:24.053145Z",
     "shell.execute_reply.started": "2025-02-20T15:38:24.014657Z"
    }
   },
   "outputs": [],
   "source": [
    "# test_pair_df = df[df[\"request_id\"] == \"621c8b02-a905-48f1-a2d5-a4e8423d1505\"]\n",
    "# print(\n",
    "#     test_pair_df[\n",
    "#         [\n",
    "#             \"s3_id\",\n",
    "#             \"total_shimmer_score\",\n",
    "#             \"pair_quality\",\n",
    "#             \"request_id\",\n",
    "#             \"preference\",\n",
    "#             \"prompt_text\",\n",
    "#         ]\n",
    "#     ]\n",
    "# )\n",
    "# negative_audio = Audio.from_s3(\n",
    "#     f\"s3://suno-data-uploads/studio/uploads/{test_pair_df['s3_id'].values[0]}.mp3\",\n",
    "#     n_channels=2,\n",
    "# )\n",
    "# print(\"negative\")\n",
    "# negative_audio.get_segment(0, 30).play()\n",
    "# positive_audio = Audio.from_s3(\n",
    "#     f\"s3://suno-data-uploads/studio/uploads/{test_pair_df['s3_id'].values[1]}.mp3\",\n",
    "#     n_channels=2,\n",
    "# )\n",
    "# print(\"positive\")\n",
    "# positive_audio.get_segment(0, 30).play()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:24.054144Z",
     "iopub.status.busy": "2025-02-20T15:38:24.054035Z",
     "iopub.status.idle": "2025-02-20T15:38:24.097039Z",
     "shell.execute_reply": "2025-02-20T15:38:24.096623Z",
     "shell.execute_reply.started": "2025-02-20T15:38:24.054133Z"
    }
   },
   "outputs": [],
   "source": [
    "# total_dict = {}\n",
    "# total_dict.update(pair_quality_dict)\n",
    "# total_dict.update(pair_quality_1_dict)\n",
    "# total_dict.update(pair_quality_2_dict)\n",
    "# total_dict.update(pair_quality_3_dict)\n",
    "# len(total_dict)\n",
    "# with open(\n",
    "#     os.path.join(\"/home/tony/Data/Preference/up_v1\", \"pair_quality.json\"), \"w\"\n",
    "# ) as fp:\n",
    "#     json.dump(total_dict, fp, indent=4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:24.097793Z",
     "iopub.status.busy": "2025-02-20T15:38:24.097537Z",
     "iopub.status.idle": "2025-02-20T15:38:24.140714Z",
     "shell.execute_reply": "2025-02-20T15:38:24.140286Z",
     "shell.execute_reply.started": "2025-02-20T15:38:24.097780Z"
    }
   },
   "outputs": [],
   "source": [
    "# import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:24.141424Z",
     "iopub.status.busy": "2025-02-20T15:38:24.141265Z",
     "iopub.status.idle": "2025-02-20T15:38:24.183145Z",
     "shell.execute_reply": "2025-02-20T15:38:24.182725Z",
     "shell.execute_reply.started": "2025-02-20T15:38:24.141411Z"
    }
   },
   "outputs": [],
   "source": [
    "# test_arr = np.load(\"/home/tony/Data/test_npz/diffusion_input_tensor([ 18, 182]).npy\")\n",
    "# test_arr.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:24.183816Z",
     "iopub.status.busy": "2025-02-20T15:38:24.183659Z",
     "iopub.status.idle": "2025-02-20T15:38:24.228867Z",
     "shell.execute_reply": "2025-02-20T15:38:24.228441Z",
     "shell.execute_reply.started": "2025-02-20T15:38:24.183804Z"
    }
   },
   "outputs": [],
   "source": [
    "# mm_vae_val = np.memmap(\n",
    "#     os.path.join(OUT_DATA_DIR, \"data_vae_val.bin\"), dtype=np.float16, mode=\"r\"\n",
    "# )\n",
    "\n",
    "# mm_vae_val = mm_vae_val.reshape(-1, VAE_MEMMAP_SIZE, VAE_DIM)\n",
    "# print(mm_vae_val.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:24.229619Z",
     "iopub.status.busy": "2025-02-20T15:38:24.229375Z",
     "iopub.status.idle": "2025-02-20T15:38:24.273105Z",
     "shell.execute_reply": "2025-02-20T15:38:24.272686Z",
     "shell.execute_reply.started": "2025-02-20T15:38:24.229605Z"
    }
   },
   "outputs": [],
   "source": [
    "# print(\"negative\")\n",
    "# audio = codec_decode(test_arr[0].T / 2.5)\n",
    "# audio.normalize_volume().play()\n",
    "\n",
    "# print(\"positive\")\n",
    "# audio = codec_decode(test_arr[1].T / 2.5)\n",
    "# audio.normalize_volume().play()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:24.273782Z",
     "iopub.status.busy": "2025-02-20T15:38:24.273619Z",
     "iopub.status.idle": "2025-02-20T15:38:24.321102Z",
     "shell.execute_reply": "2025-02-20T15:38:24.320673Z",
     "shell.execute_reply.started": "2025-02-20T15:38:24.273770Z"
    }
   },
   "outputs": [],
   "source": [
    "# print(\"negative\")\n",
    "# audio = codec_decode(mm_vae_val[18])\n",
    "# audio.normalize_volume().play()\n",
    "\n",
    "# print(\"positive\")\n",
    "# audio = codec_decode(mm_vae_val[19])\n",
    "# audio.normalize_volume().play()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:24.321892Z",
     "iopub.status.busy": "2025-02-20T15:38:24.321604Z",
     "iopub.status.idle": "2025-02-20T15:38:24.375583Z",
     "shell.execute_reply": "2025-02-20T15:38:24.375185Z",
     "shell.execute_reply.started": "2025-02-20T15:38:24.321879Z"
    }
   },
   "outputs": [],
   "source": [
    "import torch\n",
    "\n",
    "rng = torch.quasirandom.SobolEngine(1, scramble=True, seed=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:24.378210Z",
     "iopub.status.busy": "2025-02-20T15:38:24.377593Z",
     "iopub.status.idle": "2025-02-20T15:38:24.418944Z",
     "shell.execute_reply": "2025-02-20T15:38:24.418580Z",
     "shell.execute_reply.started": "2025-02-20T15:38:24.378195Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([0.4746, 0.5586, 0.9883, 0.0396], dtype=torch.bfloat16)\n",
      "tensor([1.0000, 1.0000, 0.9883, 0.0396], dtype=torch.bfloat16)\n",
      "tensor([1.0000, 1.0000, 1.0000, 1.0000, 0.9883, 0.9883, 0.0396, 0.0396],\n",
      "       dtype=torch.bfloat16)\n"
     ]
    }
   ],
   "source": [
    "t = rng.draw(4)[:, 0].to(torch.bfloat16)\n",
    "print(t)\n",
    "# Replace 1% of t with ones to ensure training on terminal SNR\n",
    "t = torch.where(torch.rand_like(t) < 0.5, torch.ones_like(t), t)\n",
    "print(t)\n",
    "t = torch.repeat_interleave(t, repeats=2, dim=0)\n",
    "print(t)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:24.419577Z",
     "iopub.status.busy": "2025-02-20T15:38:24.419299Z",
     "iopub.status.idle": "2025-02-20T15:38:24.462459Z",
     "shell.execute_reply": "2025-02-20T15:38:24.461866Z",
     "shell.execute_reply.started": "2025-02-20T15:38:24.419565Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([1.0000, 1.0000, 1.0000, 1.0000, 0.9688, 0.9688, 0.0312, 0.0312])"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(t * 32).to(int) / 32"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:24.463409Z",
     "iopub.status.busy": "2025-02-20T15:38:24.463094Z",
     "iopub.status.idle": "2025-02-20T15:38:24.506359Z",
     "shell.execute_reply": "2025-02-20T15:38:24.505810Z",
     "shell.execute_reply.started": "2025-02-20T15:38:24.463393Z"
    }
   },
   "outputs": [],
   "source": [
    "t[0] = 0.99"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:24.507082Z",
     "iopub.status.busy": "2025-02-20T15:38:24.506953Z",
     "iopub.status.idle": "2025-02-20T15:38:24.550634Z",
     "shell.execute_reply": "2025-02-20T15:38:24.550136Z",
     "shell.execute_reply.started": "2025-02-20T15:38:24.507070Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 0.0312, 0.0312],\n",
       "       dtype=torch.bfloat16)"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "torch.round(t * 32) / 32"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:24.551402Z",
     "iopub.status.busy": "2025-02-20T15:38:24.551201Z",
     "iopub.status.idle": "2025-02-20T15:38:35.775705Z",
     "shell.execute_reply": "2025-02-20T15:38:35.774994Z",
     "shell.execute_reply.started": "2025-02-20T15:38:24.551388Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "61011\n",
      "61011\n"
     ]
    }
   ],
   "source": [
    "import json\n",
    "with open(f\"/home/tony/Data/Preference/up_v4/full_pair_quality.json\", \"r\") as f:\n",
    "   result = json.load(f)\n",
    "# result = {}\n",
    "print(len(result))\n",
    "for job_idx in range(4):\n",
    "    with open(f\"/home/tony/Data/Preference/up_v4/full_pair_quality_{job_idx}.json\", \"r\") as fp:\n",
    "        current_result = json.load(fp)\n",
    "        result.update(current_result)\n",
    "print(len(result))\n",
    "# with open(f\"/home/tony/Data/Preference/up_v4/full_pair_quality.json\", \"w\") as f:\n",
    "#     json.dump(result, f, indent=4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:35.776760Z",
     "iopub.status.busy": "2025-02-20T15:38:35.776394Z",
     "iopub.status.idle": "2025-02-20T15:38:35.778891Z",
     "shell.execute_reply": "2025-02-20T15:38:35.778479Z",
     "shell.execute_reply.started": "2025-02-20T15:38:35.776744Z"
    }
   },
   "outputs": [],
   "source": [
    "# import torch\n",
    "# semantic_codes_chunk = torch.ones((1, 100))\n",
    "# semantic_skip_phase = 0\n",
    "# semantic_skip_factor = 4\n",
    "# mask = torch.ones_like(semantic_codes_chunk, dtype=torch.bool)\n",
    "# indices = (\n",
    "#     torch.arange(semantic_codes_chunk.size(1)) + semantic_skip_phase\n",
    "# ) % semantic_skip_factor == 0\n",
    "# mask[:, indices] = False\n",
    "# mask"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:35.779439Z",
     "iopub.status.busy": "2025-02-20T15:38:35.779330Z",
     "iopub.status.idle": "2025-02-20T15:38:35.828427Z",
     "shell.execute_reply": "2025-02-20T15:38:35.827937Z",
     "shell.execute_reply.started": "2025-02-20T15:38:35.779429Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['index', 'id_x', 'created_at', 'updated_at', 'time_used', 'metadata', 'user_id', 'status', 'discord_message_id', 'prompt_id',\n",
       "       ...\n",
       "       'clips_per_second', 'original_duration_s', 'has_continue_and_start_continue_at', 'good_continue_at', 'duration_rel_diff', 'play_rel_diff', 'pos_diff_preference', 'shimmer_score_diff', 'stereo_width_diff', 'spectral_centroid_diff'], dtype='object', length=118)"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_slice.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:35.829290Z",
     "iopub.status.busy": "2025-02-20T15:38:35.829011Z",
     "iopub.status.idle": "2025-02-20T15:38:35.869618Z",
     "shell.execute_reply": "2025-02-20T15:38:35.869115Z",
     "shell.execute_reply.started": "2025-02-20T15:38:35.829276Z"
    }
   },
   "outputs": [],
   "source": [
    "# df_slice[(df_slice[\"preference\"]) & (df_slice[\"total_shimmer_score\"] < 0.5)][[\"s3_id\", \"total_shimmer_score\"]].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-02-20T15:38:35.870527Z",
     "iopub.status.busy": "2025-02-20T15:38:35.870184Z",
     "iopub.status.idle": "2025-02-20T15:38:35.910861Z",
     "shell.execute_reply": "2025-02-20T15:38:35.910353Z",
     "shell.execute_reply.started": "2025-02-20T15:38:35.870512Z"
    }
   },
   "outputs": [],
   "source": [
    "# df_slice[(df_slice[\"preference\"]) & (df_slice[\"total_shimmer_score\"] > 2)][[\"s3_id\", \"total_shimmer_score\"]].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.10.15"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}