{ "cells": [ { "cell_type": "code", "execution_count": null, "id": "de09d1f3-ae75-4e4a-b4c0-dbb1339a439c", "metadata": {}, "outputs": [], "source": [ "#This Jupyter File contains the following scripts for Granite3.2-2B-FP16:\n", "\n", "#1)The Training Code used to train the LoRA adapters for the model and the output losses (if available).\n", "#->The model can be ran again to check the for the losses.\n", "\n", "#2)The Testing Code used to test the 5 variants of the model at different base precisions using the same FP16 LoRA Adapters.\n", "\n", "#3) The Evaluation Code used to evaluate the responses of the combined model and LoRA Adapters." ] }, { "cell_type": "code", "execution_count": null, "id": "82ec46fe-ad4c-4185-ad18-92d5287feada", "metadata": {}, "outputs": [], "source": [ "#TRAINING CODE FOR Granite3.2-2B-FP16" ] }, { "cell_type": "code", "execution_count": null, "id": "1ad9c7f0-77a2-4bc3-a79d-dd2bbe58cc03", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "/home/jovyan/Falcon1B/lib/python3.11/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n", " from .autonotebook import tqdm as notebook_tqdm\n", "[nltk_data] Downloading package punkt to /home/jovyan/nltk_data...\n", "[nltk_data] Package punkt is already up-to-date!\n", "Map: 100%|██████████| 15000/15000 [00:26<00:00, 568.70 examples/s]\n", "Map: 100%|██████████| 1500/1500 [00:02<00:00, 648.03 examples/s]\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "\n", "Sample 0:\n", "Input:\n", " Question: I have written a canny edge detection algorithm for a project. I want to know is there any method to link the broken segments of an edge, since i am getting a single edge as a conglomeration of a few segments. I am getting around 100 segments, which i am sure can be decreased with some intelligence. Please help.\n", "Answer: You can use a method named dynamic programming. A very good intro on this can be found on chapter 6 of Sonka's digital image processing book\n", "Label mask:\n", " [-100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, 2448, 883, 793, 312, 1411, 8189, 7094, 16031, 32, 399, 5029, 4644, 8642, 544, 458, 883, 526, 2431, 544, 18471, 225, 40, 432, 28903, 3795, 1182, 18452, 1778, 8202, 7618, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]\n", "\n", "Sample 1:\n", "Input:\n", " Context: I have a dataset of book reviews:\n", "```\n", "| user_id | ISBN | vote | votes_for_user | average_user_vote | ISBN_categ |\n", " 213 3242X 4.5 12 3.4 1 \n", " 563 1245X 3.2 74 2.3 2\n", "```\n", "\n", "where \n", "```\n", " vote = rating given by user to a certain book\n", " votes_for_user = number of votes the user has in the dataset (nr of rows)\n", " average_user_vote = average of a user's votes\n", " ISBN_categ = integer categorical of the ISBN (since that is a string).\n", "```\n", "\n", "I want to apply a clustering algorithm such as DBSCAN to see how many clusters I can form with this dataset. \n", "My question is: \n", "Should I apply the clustering on the dataframe as it is (minus the ISBN column) or should I construct more features for every user and construct a dataframe where every user appears only once, together with their features, and cluster that? \n", "Remember, the intent here is to cluster users (by user_id), not data points (votes).\n", "Question: Clustering of users in a dataset\n", "Answer: If your objective is to find clusters of users, then you are interested in finding groups of \"similar\" reviewers.\n", "Therefore you should:\n", "\n", "- Retain information which relates to the users in a meaningful way - e.g. votes_for_user.\n", "\n", "- Discard information which has no meaningful relationship to a user - e.g. user_id (unless perhaps it contains some information such as time / order).\n", "\n", "- Be mindful of fields which may contain implicit relationships involving a user - e.g. vote may be a result of the interaction between user and ISBN.\n", "Label mask:\n", " [-100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, 1670, 1370, 25110, 438, 372, 2290, 18063, 432, 4250, 30, 1615, 844, 884, 16574, 328, 17340, 8351, 432, 313, 18310, 20, 37499, 32, 203, 8309, 1444, 844, 1395, 44, 203, 203, 31, 9687, 504, 2471, 1510, 1875, 1196, 372, 322, 4250, 328, 312, 33081, 3352, 429, 484, 32, 89, 32, 34751, 81, 979, 81, 496, 32, 203, 203, 31, 3645, 2294, 2471, 1510, 1401, 1289, 33081, 12112, 372, 312, 1256, 429, 484, 32, 89, 32, 1256, 81, 314, 308, 23437, 19368, 561, 4304, 1629, 2471, 3751, 619, 1133, 517, 2532, 547, 203, 203, 31, 4261, 12204, 2790, 432, 3829, 1510, 1631, 4799, 10353, 25041, 14907, 7172, 312, 1256, 429, 484, 32, 89, 32, 20424, 1631, 526, 312, 1056, 432, 322, 15994, 3733, 1256, 461, 2756, 14282, 32, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]\n", "\n", "Sample 2:\n", "Input:\n", " Question: What's a common technical challenge when using logistic regression?\n", "Answer: Dealing with class imbalance, which is when the number of observations in one class is significantly lower than the number of observations in the other class.\n", "Label mask:\n", " [-100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, 1513, 23959, 623, 443, 3960, 10401, 30, 1510, 438, 1412, 322, 1451, 432, 25285, 328, 1591, 443, 438, 32323, 7216, 2784, 322, 1451, 432, 25285, 328, 322, 1604, 443, 32, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Loading checkpoint shards: 100%|██████████| 2/2 [00:23<00:00, 11.74s/it]\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "trainable params: 34,078,720 || all params: 2,567,610,368 || trainable%: 1.3273\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "/tmp/ipykernel_65109/1516336251.py:195: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `Trainer.__init__`. Use `processing_class` instead.\n", " trainer = Trainer(model=model, args=training_args, train_dataset=train_dataset, eval_dataset=test_dataset, tokenizer=tokenizer,\n" ] }, { "data": { "text/html": [ "\n", "
\n", " \n", " \n", " [ 691/4220 35:02 < 23:09:42, 0.04 it/s, Epoch 3.27/20]\n", "
\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
StepTraining LossValidation Loss
6106.0043000.846677
6202.6903000.721633
6301.8920000.662078
6401.8555000.638113
6501.8781000.642582
6601.8788000.658247
6701.8367000.670157
6801.9419000.676019

\n", "

\n", " \n", " \n", " [ 437/1498 00:55 < 02:15, 7.80 it/s]\n", "
\n", " " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ "Saved checkpoint at step 650 to ./Granite3.2-2B-NF4-lora-FP16/checkpoint-650\n" ] } ], "source": [ "#Granite-2B-Instruct code\n", "#Does BNB quantisation via NF4 on base and FP16 on Adapters\n", "#Code with custom preprocess_function()\n", "\n", "import os\n", "import torch\n", "import time\n", "import json\n", "import pandas as pd\n", "import evaluate\n", "import nltk\n", "import gc\n", "import math\n", "from transformers import AutoModelForCausalLM, AutoTokenizer, TrainingArguments, Trainer, EarlyStoppingCallback, TrainerCallback\n", "from transformers.trainer_utils import get_last_checkpoint\n", "from peft import LoraConfig, get_peft_model, prepare_model_for_kbit_training\n", "from datasets import Dataset\n", "from sklearn.metrics import accuracy_score, precision_recall_fscore_support\n", "from nltk.translate.bleu_score import sentence_bleu\n", "from rouge_score import rouge_scorer\n", "from torch.utils.data import DataLoader\n", "from transformers import BitsAndBytesConfig\n", "\n", "#export FLASH_ATTENTION=1 # Add before training\n", "os.environ[\"TOKENIZERS_PARALLELISM\"] = \"false\" # Add this before any tokenizer calls\n", "os.environ[\"PYTORCH_CUDA_ALLOC_CONF\"] = \"expandable_segments:True\"\n", "\n", "nltk.download(\"punkt\")\n", "\n", "MODEL_NAME = \"ibm-granite/granite-3.2-2b-instruct\"\n", "\n", "# Load the tokenizer\n", "tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)\n", "tokenizer.pad_token = tokenizer.eos_token\n", "tokenizer.padding_side = \"right\" #set to left for Falcon\n", "\n", "# Load dataset from CSV files\n", "train_df = pd.read_csv(\"Training Dataset RE.csv\") #Using datasets with 40% less samples (25k -> 15k)\n", "test_df = pd.read_csv(\"Testing Dataset RE.csv\") #Using datasets with 40% less samples (2.5k -> 1.5k)\n", "\n", "def preprocess_function(examples):\n", " inputs = []\n", " labels = []\n", "\n", " for context, question, answer in zip(\n", " examples.get(\"Context\", [\"\"] * len(examples[\"Question\"])),\n", " examples[\"Question\"],\n", " examples[\"Answer\"]):\n", " \n", " context = context.strip() if context else \"\"\n", " question = question.strip()\n", " answer = answer.strip()\n", "\n", " if context:\n", " prompt = f\"Context: {context}\\nQuestion: {question}\\nAnswer:\"\n", " else:\n", " prompt = f\"Question: {question}\\nAnswer:\"\n", "\n", " full_text = prompt + \" \" + answer\n", "\n", " tokenized = tokenizer(full_text, padding=\"max_length\", truncation=True, max_length=512)\n", " prompt_ids = tokenizer(prompt, truncation=True, max_length=512, add_special_tokens=False)[\"input_ids\"]\n", "\n", " input_ids = tokenized[\"input_ids\"]\n", " attention_mask = tokenized[\"attention_mask\"]\n", " label_ids = input_ids.copy()\n", " label_ids[:len(prompt_ids)] = [-100] * len(prompt_ids) # mask prompt tokens\n", "\n", " # Fix #4: Skip samples with no answer tokens (i.e., all labels = -100)\n", " if all(id_ == -100 for id_ in label_ids):\n", " continue\n", "\n", " inputs.append({\"input_ids\": input_ids, \"attention_mask\": attention_mask, \"labels\": label_ids})\n", "\n", " return {\"input_ids\": [x[\"input_ids\"] for x in inputs], \"attention_mask\": [x[\"attention_mask\"] for x in inputs],\n", " \"labels\": [x[\"labels\"] for x in inputs]}\n", "\n", "#train_dataset = Dataset.from_pandas(train_df).map(preprocess_function, batched=True)\n", "#test_dataset = Dataset.from_pandas(test_df).map(preprocess_function, batched=True)\n", "\n", "train_dataset = Dataset.from_pandas(train_df).map(preprocess_function, batched=True, batch_size=32, # Match your batch size\n", " remove_columns=train_df.columns.tolist()) # Convert to list explicitly & Remove original columns\n", "\n", "test_dataset = Dataset.from_pandas(test_df).map(preprocess_function, batched=True, batch_size=32, \n", " remove_columns=test_df.columns.tolist())\n", "\n", "# After mapping, print a few samples\n", "for i in range(3):\n", " decoded_input = tokenizer.decode(train_dataset[i][\"input_ids\"], skip_special_tokens=True)\n", " labels = train_dataset[i][\"labels\"]\n", " print(f\"\\nSample {i}:\")\n", " print(\"Input:\\n\", decoded_input)\n", " print(\"Label mask:\\n\", labels)\n", "\n", "# Load model with BitsAndBytes NF4 quantization\n", "bnb_config = BitsAndBytesConfig(\n", " load_in_4bit=True, # Enables 4-bit loading\n", " bnb_4bit_quant_type=\"nf4\", # Uses NormalFloat4 (NF4)\n", " bnb_4bit_compute_dtype=torch.float16, # Compute in FP16\n", " bnb_4bit_use_double_quant=True # Double quantization for efficiency\n", ")\n", "model = AutoModelForCausalLM.from_pretrained(MODEL_NAME, quantization_config=bnb_config, device_map=\"auto\") #removed flash attention\n", "model.config.pad_token_id = tokenizer.pad_token_id\n", "\n", "# Apply LoRA configuration\n", "lora_config = LoraConfig(r=64, lora_alpha=128, lora_dropout=0.2, bias=\"none\", task_type=\"CAUSAL_LM\",\n", " target_modules=[\"q_proj\", \"k_proj\", \"v_proj\", \"o_proj\"]) #Remove/add new target modules for Falcon\n", "#increased dropout from 0.2 to 0.25\n", "\n", "# Prepare model for k-bit training\n", "model = prepare_model_for_kbit_training(model)\n", "\n", "model.gradient_checkpointing_enable()\n", "model.config.use_cache = False # Required for checkpointing to work with transformers\n", "\n", "#for name, module in model.named_modules(): #Check target modules\n", " #print(name)\n", "\n", "model = get_peft_model(model, lora_config)\n", "model.print_trainable_parameters() # Optional: sanity check\n", "\n", "# Verify gradients\n", "assert any(p.requires_grad for p in model.parameters()), \"No trainable parameters found!\"\n", "\n", "#model = torch.compile(model) #ehh its risky\n", "\n", "class SaveAt650Callback(TrainerCallback):\n", " def __init__(self, tokenizer):\n", " self.tokenizer = tokenizer\n", " \n", " def on_step_end(self, args, state, control, **kwargs):\n", " if state.global_step == 650:\n", " checkpoint_dir = os.path.join(args.output_dir, f\"checkpoint-650\")\n", " os.makedirs(checkpoint_dir, exist_ok=True)\n", " \n", " # Save model\n", " kwargs[\"model\"].save_pretrained(checkpoint_dir)\n", " \n", " # Save tokenizer (using the instance we stored)\n", " self.tokenizer.save_pretrained(checkpoint_dir)\n", " \n", " # Save training state\n", " torch.save({\n", " 'optimizer_state_dict': kwargs[\"optimizer\"].state_dict(),\n", " 'scheduler_state_dict': kwargs[\"lr_scheduler\"].state_dict(),\n", " }, os.path.join(checkpoint_dir, \"training_state.pt\"))\n", " \n", " print(f\"Saved checkpoint at step 650 to {checkpoint_dir}\")\n", "\n", "# Initialize the callback with your tokenizer\n", "save_callback = SaveAt650Callback(tokenizer=tokenizer)\n", "\n", "#total_steps = (len(train_dataset) // training_args.per_device_train_batch_size) * training_args.num_train_epochs\n", "total_steps = (len(train_dataset) // 32) * 20 # Use per_device_train_batch_size=8 and num_train_epochs=4 directly\n", "training_args = TrainingArguments(\n", " output_dir=\"./Granite3.2-2B-NF4-lora-FP16\",\n", " per_device_train_batch_size=32, #set to 32,\n", " gradient_accumulation_steps=2, #set to 2, # Effective batch=64\n", " gradient_checkpointing=True, # Ensure this is enabled\n", " per_device_eval_batch_size=1, #reduce this to 2 or 1\n", " eval_accumulation_steps=4, # Process in smaller chunks\n", " num_train_epochs=20, #increased from 4 to 12 (after 7 epochs) to 20 to 25 (after 18 epochs)\n", " learning_rate=3e-5, # Set to 1e-5, Raised to 3e-5 at STEP 600\n", " #lr_scheduler_type=\"linear\", # Linear decay\n", " #lr_scheduler_type=\"cosine\",\n", " lr_scheduler_type=\"cosine_with_restarts\",\n", " lr_scheduler_kwargs={\"num_cycles\": 4}, #How to pass num_cycles with cosine decay, Set to 4 to allow more resets\n", " warmup_ratio=0.1, # Longer warmup\n", " max_grad_norm=1.0, # Gradient clipping, set to 1\n", " optim=\"paged_adamw_8bit\", # Adamw can be slow for lora\n", " warmup_steps=int(0.1 * total_steps), # Explicit steps (alternative to ratio) \n", " fp16=True,\n", " fp16_full_eval=True,\n", " weight_decay=0.02, # Regularization\n", " logging_steps=10,\n", " save_strategy=\"steps\",\n", " save_steps=200, # Save checkpoint every 200 steps\n", " eval_strategy=\"steps\",\n", " #eval_strategy=\"epoch\",\n", " save_total_limit=4,\n", " label_names=[\"labels\"],\n", " do_eval=True, # Ensure evaluation occurs\n", " dataloader_num_workers=4, #set to 4\n", " dataloader_pin_memory=True,\n", " remove_unused_columns=False,\n", " #load_best_model_at_end=True, # Required for EarlyStoppingCallback #H\n", " metric_for_best_model=\"eval_loss\", # Which metric to monitor\n", " greater_is_better=False, # For eval_loss, lower is better\n", " #predict_with_generate=True # For text generation tasks\n", ")\n", "\n", "#callbacks = [EarlyStoppingCallback(early_stopping_patience=5, # Stop if no improvement\n", " #early_stopping_threshold=0.1)] # Min delta to qualify as improvement\n", "\n", "trainer = Trainer(model=model, args=training_args, train_dataset=train_dataset, eval_dataset=test_dataset, tokenizer=tokenizer,\n", " callbacks=[save_callback])\n", " #callbacks=[EarlyStoppingCallback(early_stopping_patience=3, early_stopping_threshold=0.01)])\n", "\n", "# Add this before training\n", "torch.backends.cuda.enable_flash_sdp(True) # Enable FlashAttention\n", "torch.backends.cuda.enable_mem_efficient_sdp(True) # Enable memory-efficient attention\n", "\n", "if os.path.exists(\"./Granite3.2-2B-NF4-lora-FP16/checkpoint-600\"):\n", " trainer.train(resume_from_checkpoint=\"./Granite3.2-2B-NF4-lora-FP16/checkpoint-600\") #resume_from_checkpoint=\"./Phi-1-NF4-lora-FP16-2/checkpoint-500\"\n", "else:\n", " trainer.train() # Fresh start\n", " \n", "model.save_pretrained(\"Granite3.2-2B-lora_adapters-FP16\")\n", "\n", "def compute_metrics(predictions, references):\n", " predictions_text = tokenizer.batch_decode(predictions, skip_special_tokens=True)\n", " references_text = tokenizer.batch_decode(references, skip_special_tokens=True)\n", "\n", " # Exact Match (EM)\n", " def exact_match(pred, ref):\n", " return int(pred == ref)\n", "\n", " em_score = sum(exact_match(p, r) for p, r in zip(predictions_text, references_text)) / len(predictions_text)\n", " \n", " accuracy = accuracy_score(references_text, predictions_text)\n", " precision, recall, f1, _ = precision_recall_fscore_support(references_text, predictions_text, average=\"weighted\")\n", " \n", " bleu_scores = [sentence_bleu([ref.split()], pred.split()) for ref, pred in zip(references_text, predictions_text)]\n", " avg_bleu = sum(bleu_scores) / len(bleu_scores)\n", " \n", " rouge = rouge_scorer.RougeScorer([\"rouge1\", \"rouge2\", \"rougeL\"], use_stemmer=True)\n", " rouge_scores = [rouge.score(ref, pred) for ref, pred in zip(references_text, predictions_text)]\n", " avg_rouge = {key: sum(d[key].fmeasure for d in rouge_scores) / len(rouge_scores) for key in rouge_scores[0]}\n", " \n", " return {\"accuracy\": accuracy, \"precision\": precision, \"recall\": recall, \"f1\": f1, \"bleu\": avg_bleu, \"rouge\": avg_rouge, \n", " \"exact_match\": em_score,}\n", "\n", "# Inference latency and memory measurement\n", "def measure_inference_performance(model, dataset):\n", " latencies = []\n", " memory_usages = []\n", " \n", " for sample in dataset:\n", " input_ids = torch.tensor(sample[\"input_ids\"]).unsqueeze(0).to(\"cuda\")\n", " attention_mask = torch.tensor(sample[\"attention_mask\"]).unsqueeze(0).to(\"cuda\")\n", " \n", " torch.cuda.synchronize()\n", " start_time = time.time()\n", " memory_before = torch.cuda.memory_allocated() / (1024 ** 3)\n", " \n", " with torch.no_grad():\n", " #_ = model.generate(input_ids, attention_mask=attention_mask, max_new_tokens=50)\n", " output = model.generate(input_ids, attention_mask=attention_mask, max_new_tokens=50, \n", " do_sample=True, # Adds randomness for diversity\n", " top_p=0.9, # Nucleus sampling (prevents very unlikely words)\n", " top_k=50, # Limits token selection\n", " temperature=0.7, # Controls randomness (lower = more deterministic)\n", " early_stopping=True, # Prevents excessive token generation\n", " repetition_penalty=1.1, # Reduce nonsense repeats\n", " length_penalty=0.8 # Prefer shorter answers\n", " )\n", " \n", " torch.cuda.synchronize()\n", " end_time = time.time()\n", " memory_after = torch.cuda.memory_allocated() / (1024 ** 3)\n", " \n", " latencies.append((end_time - start_time) * 1000) # Convert to ms\n", " memory_usages.append(memory_after - memory_before)\n", " \n", " return {\"avg_latency_ms\": sum(latencies) / len(latencies), \"avg_memory_usage_gb\": sum(memory_usages) / len(memory_usages)}\n", "\n", "torch.cuda.empty_cache()\n", "eval_results = trainer.evaluate(test_dataset) #Measures loss/generalization (without generating outputs).\n", "\n", "# Compute Perplexity\n", "eval_loss = eval_results[\"eval_loss\"]\n", "perplexity = math.exp(eval_loss) # Perplexity = e^(loss)\n", "eval_results[\"perplexity\"] = perplexity\n", "\n", "torch.cuda.empty_cache()\n", "model.gradient_checkpointing_disable() # Before prediction\n", "predictions = trainer.predict(test_dataset).predictions\n", "model.gradient_checkpointing_enable() # Re-enable if needed\n", "references = test_dataset[\"Answer\"]\n", "performance_metrics = compute_metrics(predictions, references) #Measures accuracy, BLEU, ROUGE (based on generated outputs).\n", "inference_performance = measure_inference_performance(model, test_dataset) #Measures speed & memory usage of generating outputs.\n", "\n", "results = {\"evaluation_metrics\": eval_results, \"performance_metrics\": performance_metrics, \"inference_performance\": inference_performance}\n", "with open(\"evaluation_results-Granite3.2-2B-NF4-lora-FP16.json\", \"w\") as f:\n", " json.dump(results, f, indent=4)\n", "\n", "print(\"Evaluation complete! Results saved in evaluation_results-Granite3.2-2B-NF4-lora-FP16.json.\")\n" ] }, { "cell_type": "code", "execution_count": null, "id": "bc4b748b-51d5-466c-9db4-5a014b68bf18", "metadata": {}, "outputs": [], "source": [ "#TESTING CODE FOR Granite3.2-2B-FP16" ] }, { "cell_type": "code", "execution_count": null, "id": "6a59d5cd-9e60-442d-9373-ad912b015caa", "metadata": {}, "outputs": [], "source": [ "#1)#####################################################################################################################" ] }, { "cell_type": "code", "execution_count": 1, "id": "4c73787b-f6b3-453f-9b48-969c9fedb078", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "/home/jovyan/Falcon1B/lib/python3.11/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n", " from .autonotebook import tqdm as notebook_tqdm\n", "[nltk_data] Downloading package punkt to /home/jovyan/nltk_data...\n", "[nltk_data] Package punkt is already up-to-date!\n", "Map: 100%|██████████| 1500/1500 [00:02<00:00, 626.64 examples/s]\n", "Loading checkpoint shards: 100%|██████████| 2/2 [00:41<00:00, 20.82s/it]\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Generating predictions...\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "/home/jovyan/Falcon1B/lib/python3.11/site-packages/transformers/generation/configuration_utils.py:695: UserWarning: `num_beams` is set to 1. However, `length_penalty` is set to `0.8` -- this flag is only used in beam-based generation modes. You should set `num_beams>1` or unset `length_penalty`.\n", " warnings.warn(\n", "Featurizing p: 100%|██████████| 1498/1498 [00:42<00:00, 35.10it/s]\n", "Featurizing q: 100%|██████████| 1498/1498 [00:42<00:00, 35.25it/s]\n", "WARNING clustering 2996 points to 150 centroids: please provide at least 5850 training points\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Evaluation Complete. Results saved to Complete_Evaluation_Results/Granite3.2-2B/NF4/FP16/Granite3.2-2B-NF4-lora-FP16-Evaluation_Results.json\n" ] }, { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjAAAAHHCAYAAAChjmJTAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjEsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvc2/+5QAAAAlwSFlzAAAPYQAAD2EBqD+naQAAlUFJREFUeJztnXd4FFXbxu9N7wktCZ0ICgHpKERQUSIBsaDYEVBAXxAUUAH5VEQBQRQQlaKiCSJIUUQF6RBqaIFA6C0QWhIgpCe7m935/ogZdpNtMzuzc2b3+V1XLtiZM2ee0+85VcNxHAeCIAiCIAgV4aW0AQRBEARBEEIhAUMQBEEQhOogAUMQBEEQhOogAUMQBEEQhOogAUMQBEEQhOogAUMQBEEQhOogAUMQBEEQhOogAUMQBEEQhOogAUMQBEEQhOogAUMQVSgqKsLQoUMRHR0NjUaD0aNHK22SR3L27Fn07NkT4eHh0Gg0WL16tdImEQTBECRgCLcjKSkJGo0GBw8eFPX8559/jqSkJAwfPhyLFy/GgAEDJLZQvTRp0gQajYb/i4yMxIMPPog///xT8ncNGjQI6enpmDp1KhYvXoxOnTpJ/g5Po6CgAJ9++inatm2LkJAQBAYG4t5778X48eNx7do1pc0jCEFo6Cwkwt1ISkrC66+/jgMHDohq9Lp06QIfHx/s2rVLBuvUTZMmTVCjRg289957AIBr167h+++/x4ULFzB//nwMGzZMkveUlpYiKCgIH374IaZMmSKJn57OhQsXEB8fj8zMTDz//PPo1q0b/Pz8cPToUfz222+oWbMmzpw5o7SZBOEwPkobQBCskZOTg5YtW0rmn9FohE6nQ0BAgGR+Kkn9+vXx6quv8r8HDhyIZs2aYfbs2U4LmLKyMvj5+eHGjRsAgIiICKf8M6W4uBjBwcGS+acmysvL8eyzzyI7OxvJycno1q2b2f2pU6fiiy++kORdlWno5UUd/IS8UA4jPILXXnsNISEhuHr1Kvr27YuQkBDUqVMH77//PgwGAwAgOTkZGo0GGRkZWLt2LT9McvHiRQCAVqvFJ598gmbNmsHf3x8NGzbEuHHjoNVqzd6l0WgwcuRILFmyBK1atYK/vz/Wr18PALh69SoGDx6MqKgo+Pv7o1WrVvj555/Nnq+0Y8WKFZg6dSoaNGiAgIAA9OjRA+fOnasWtn379uHxxx9HjRo1EBwcjDZt2mDOnDlmbk6dOoXnnnsONWvWREBAADp16oS///5bkriNjo5GbGwsMjIy+GtCwrls2TJ89NFHqF+/PoKCgvDuu++icePGAICxY8dCo9GgSZMm/HOHDx9G7969ERYWhpCQEPTo0QN79+4187tyGHH79u146623EBkZiQYNGgAAunfvjnvvvRdHjx7Fww8/jKCgIDRr1gy///47AGD79u3o3LkzAgMD0bx5c2zevNnM70uXLuGtt95C8+bNERgYiFq1auH555/n80lVG3bv3o13330XderUQXBwMJ555hleoJmybt06PPzwwwgNDUVYWBjuu+8+LF261MzNvn370KtXL4SHhyMoKAgPP/wwdu/ebTeN/vjjDxw5cgQffvhhNfECAGFhYZg6dSr/u0mTJnjttdequevevTu6d+/O/7aWhocOHYJGo8GiRYuq+bFhwwZoNBqsWbOGv+ZIfiGIqlAPDOExGAwGJCQkoHPnzvjqq6+wefNmzJw5E02bNsXw4cMRGxuLxYsXY8yYMWjQoAE/TFKnTh0YjUY89dRT2LVrF958803ExsYiPT0ds2fPxpkzZ6pNMN26dStWrFiBkSNHonbt2mjSpAmys7PRpUsXXuDUqVMH69atw5AhQ1BQUFBtsvD06dPh5eWF999/H/n5+ZgxYwb69++Pffv28W42bdqEJ554AnXr1sWoUaMQHR2NkydPYs2aNRg1ahQA4Pjx4+jatSvq16+PDz74AMHBwVixYgX69u2LP/74A88884xT8arX63H58mXUqlULAASHc/LkyfDz88P7778PrVaLxx9/HE2aNMGYMWPw8ssv4/HHH0dISAgflgcffBBhYWEYN24cfH198f3336N79+688DDlrbfeQp06dTBx4kQUFxfz12/fvo0nnngCL730Ep5//nnMnz8fL730EpYsWYLRo0dj2LBheOWVV/Dll1/iueeew+XLlxEaGgoAOHDgAPbs2YOXXnoJDRo0wMWLFzF//nx0794dJ06cQFBQkJkNb7/9NmrUqIFPPvkEFy9exNdff42RI0di+fLlvJukpCQMHjwYrVq1woQJExAREYHDhw9j/fr1eOWVVwBU5KnevXujY8eO+OSTT+Dl5YXExEQ8+uij2LlzJ+6//36raVQpVuWaz1U1DVu2bIm77roLK1aswKBBg8zcLl++HDVq1EBCQgIA4fmFIHg4gnAzEhMTOQDcgQMH+GuDBg3iAHCfffaZmdv27dtzHTt2NLvWuHFjrk+fPmbXFi9ezHl5eXE7d+40u75gwQIOALd7927+GgDOy8uLO378uJnbIUOGcHXr1uVu3rxpdv2ll17iwsPDuZKSEo7jOG7btm0cAC42NpbTarW8uzlz5nAAuPT0dI7jOK68vJyLiYnhGjduzN2+fdvMT6PRyP+/R48eXOvWrbmysjKz+w888AB39913c0Jo3Lgx17NnT+7GjRvcjRs3uCNHjnAvvfQSB4B7++23RYXzrrvu4q9VkpGRwQHgvvzyS7Prffv25fz8/Ljz58/z165du8aFhoZyDz30EH+tMg9069aNKy8vN/Pj4Ycf5gBwS5cu5a+dOnWKT7e9e/fy1zds2MAB4BITE/lrVW3lOI5LSUnhAHC//PJLNRvi4+PN0mPMmDGct7c3l5eXx3Ecx+Xl5XGhoaFc586dudLSUjN/K58zGo3c3XffzSUkJJj5VVJSwsXExHCPPfZYNZtMad++PRceHm7TjSmNGzfmBg0aVO36ww8/zD388MP8b1tpOGHCBM7X15fLzc3lr2m1Wi4iIoIbPHgwf83R/EIQVaEhJMKjqDpH48EHH8SFCxfsPrdy5UrExsaiRYsWuHnzJv/36KOPAgC2bdtm5v7hhx82m0fDcRz++OMPPPnkk+A4zsyPhIQE5Ofn49ChQ2Z+vP766/Dz8zOzFQBv7+HDh5GRkYHRo0dXmyui0WgAALm5udi6dSteeOEFFBYW8u+8desWEhIScPbsWVy9etVu+E3ZuHEj6tSpgzp16qBt27ZYuXIlBgwYgC+++EJUOAcNGoTAwEC77zUYDNi4cSP69u2Lu+66i79et25dvPLKK9i1axcKCgrMnnnjjTfg7e1dza+QkBC89NJL/O/mzZsjIiICsbGxZr04lf83zSOmtur1ety6dQvNmjVDREREtbABwJtvvsmnB1CRjgaDAZcuXQJQ0YtWWFiIDz74oNo8qcrn0tLScPbsWbzyyiu4desWH6fFxcXo0aMHduzYAaPRaDXuCgoK+B4kObCUhi+++CL0ej1WrVrFX9u4cSPy8vLw4osvAhBXLgiiEhpCIjyGgIAA1KlTx+xajRo1cPv2bbvPnj17FidPnqz2fCU5OTlmv2NiYsx+37hxA3l5efjhhx/www8/OORHo0aNqtkKgLf3/PnzAIB7773Xqt3nzp0Dx3H4+OOP8fHHH1t9b/369a36UZXOnTtjypQp0Gg0CAoKQmxsLC+gcnJyBIezalxZ48aNGygpKUHz5s2r3YuNjYXRaMTly5fRqlUru343aNDATFQAQHh4OBo2bFjtGgCzPFJaWopp06YhMTERV69eBWeykDM/P7/au6RIx7NnzwJAteEYU/Lz83m/qxIWFuaQUBeLpXhu27YtWrRogeXLl2PIkCEAKoaPateuzQt/MeWCICohAUN4DJa+xB3FaDSidevWmDVrlsX7VRu+ql+jlV/Hr776qtVGqE2bNma/rdnLCdj5oPK977//Pj/noCrNmjVz2D8AqF27NuLj422+T0g4Hel9EYs1v63FrSNx/vbbbyMxMRGjR49GXFwcv9HeSy+9ZLEXRMp0/PLLL9GuXTuLbirnCVmiRYsWOHz4MC5fvlwtr1qiqrirxGAwWAyPtXh+8cUXMXXqVNy8eROhoaH4+++/8fLLL8PHp6LpEZNfCKISEjAE4QBNmzbFkSNH0KNHD6uVuy3q1KmD0NBQGAwGq42/GJsA4NixY1b9rBxq8fX1ley9tpAjnKZ+BwUF4fTp09XunTp1Cl5eXg41zs7y+++/Y9CgQZg5cyZ/raysDHl5eaL8M01Ha2Ky0k1YWJioeH3yySfx22+/4ddff8WECRPsuq9Ro4bF8Fy6dMls+M4eL774Ij799FP88ccfiIqKQkFBgdnQnZz5hXB/aA4MQTjACy+8gKtXr+LHH3+sdq+0tNRshYslvL290a9fP/zxxx84duxYtfuWltXao0OHDoiJicHXX39drbGp/LqPjIxE9+7d8f333+P69euSvNcWcoTT1O+ePXvir7/+MluynJ2djaVLl6Jbt24ICwsT7b8QO6r2nnz77bf8cnyh9OzZE6GhoZg2bRrKysrM7lW+p2PHjmjatCm++uorFBUVVfPDXrw+99xzaN26NaZOnYqUlJRq9wsLC/Hhhx/yv5s2bYq9e/dCp9Px19asWYPLly8LCltsbCxat26N5cuXY/ny5ahbty4eeugh/r6c+YVwf6gHhiAcYMCAAVixYgWGDRuGbdu2oWvXrjAYDDh16hRWrFiBDRs22N31d/r06di2bRs6d+6MN954Ay1btkRubi4OHTqEzZs3Izc3V5BNXl5emD9/Pp588km0a9cOr7/+OurWrYtTp07h+PHj2LBhAwBg7ty56NatG1q3bo033ngDd911F7Kzs5GSkoIrV67gyJEjouPFFeE0ZcqUKdi0aRO6deuGt956Cz4+Pvj++++h1WoxY8YMCUNhnSeeeAKLFy9GeHg4WrZsiZSUFGzevJlfRi6UsLAwzJ49G0OHDsV9992HV155BTVq1MCRI0dQUlKCRYsWwcvLCwsXLkTv3r3RqlUrvP7666hfvz6uXr2Kbdu2ISwsDP/884/Vd/j6+mLVqlWIj4/HQw89hBdeeAFdu3aFr68vjh8/jqVLl6JGjRr8XjBDhw7F77//jl69euGFF17A+fPn8euvv/I9QUJ48cUXMXHiRAQEBGDIkCHVNriTM78Q7g0JGIJwAC8vL6xevRqzZ8/GL7/8gj///BNBQUG46667MGrUKNxzzz12/YiKisL+/fvx2WefYdWqVZg3bx5q1aqFVq1aid4FNSEhAdu2bcOnn36KmTNnwmg0omnTpnjjjTd4Ny1btsTBgwfx6aefIikpCbdu3UJkZCTat2+PiRMninqvLeQIZyWtWrXCzp07MWHCBEybNg1GoxGdO3fGr7/+Wm0PGLmYM2cOvL29sWTJEpSVlaFr167YvHmz1TlGjjBkyBBERkZi+vTpmDx5Mnx9fdGiRQuMGTOGd9O9e3ekpKRg8uTJ+O6771BUVITo6Gh07twZ//vf/+y+o1mzZkhLS8Ps2bPx559/YvXq1TAajWjWrBmGDh2Kd955h3ebkJCAmTNnYtasWRg9ejQ6deqENWvW8HsjCeHFF1/ERx99hJKSEn71kSly5hfCvaGzkAiCIAiCUB00B4YgCIIgCNVBQ0gEQSArK8vm/cDAQH5PFIIgCBagISSCIOwuDR80aBCSkpJcYwxBEIQDUA8MQRDYtGmTzfv16tVzkSUEQRCOIWgOzKRJk6DRaMz+WrRowd8vKyvDiBEjUKtWLYSEhKBfv37Izs428yMzMxN9+vRBUFAQIiMjMXbsWJSXl5u5SU5ORocOHeDv749mzZrRlx9ByEx8fLzNP9NznQiCIFhA8CTeVq1a4fr16/zfrl27+HtjxozBP//8g5UrV2L79u24du0ann32Wf6+wWBAnz59oNPpsGfPHixatAhJSUlmSzkzMjLQp08fPPLII0hLS8Po0aMxdOhQfk8LgiAIgiAIQXNgJk2ahNWrVyMtLa3avfz8fNSpUwdLly7Fc889B6Bie+/Y2FikpKSgS5cuWLduHZ544glcu3YNUVFRAIAFCxZg/PjxuHHjBvz8/DB+/HisXbvWbFfGl156CXl5eVi/fr3DATMajbh27RpCQ0NFbf1OEARBEITr4TgOhYWFqFevXrWND00RPAfm7NmzqFevHgICAhAXF4dp06ahUaNGSE1NhV6vNzvPokWLFmjUqBEvYFJSUtC6dWtevAAVGyYNHz4cx48fR/v27ZGSklLtTIyEhASMHj3apl1arRZarZb/ffXqVer2JgiCIAiVcvnyZTRo0MDqfUECpnPnzkhKSkLz5s1x/fp1fPrpp3jwwQdx7NgxZGVlwc/PDxEREWbPREVF8Us0s7KyzMRL5f3Ke7bcFBQUoLS01Oqpp9OmTcOnn35a7frly5ddcj4KQRAEQRDOU1BQgIYNGyI0NNSmO0ECpnfv3vz/27Rpg86dO6Nx48ZYsWKFVWHhKiZMmIB3332X/10ZAWFhYSRgCIIgCEJl2Jv+4dROvBEREbjnnntw7tw5REdHQ6fTVTsVNzs7G9HR0QCA6OjoaquSKn/bcxMWFmZTJPn7+/NihUQLQRAEQbg3TgmYoqIinD9/HnXr1kXHjh3h6+uLLVu28PdPnz6NzMxMxMXFAQDi4uKQnp6OnJwc3s2mTZsQFhbGz1eJi4sz86PSTaUfBEEQBEEQggTM+++/j+3bt+PixYvYs2cPnnnmGXh7e+Pll19GeHg4hgwZgnfffRfbtm1DamoqXn/9dcTFxaFLly4AgJ49e6Jly5YYMGAAjhw5gg0bNuCjjz7CiBEj4O/vDwAYNmwYLly4gHHjxuHUqVOYN28eVqxYYXYqK0EQBEEQno2gOTBXrlzByy+/jFu3bqFOnTro1q0b9u7dizp16gAAZs+eDS8vL/Tr1w9arRYJCQmYN28e/7y3tzfWrFmD4cOHIy4uDsHBwRg0aBA+++wz3k1MTAzWrl2LMWPGYM6cOWjQoAEWLlzo1FH1tjAYDNDr9bL4TRCs4OvrC29vb6XNIAiCkAy3PQupoKAA4eHhyM/PtzgfhuM4ZGVlVZuzQxDuSkREBKKjo2lfJIIgmMZe+12Jx56FVCleIiMjERQURJU64bZwHIeSkhJ+7lndunUVtoggCMJ5PFLAGAwGXrzUqlVLaXMIQnYqV/Dl5OQgMjKShpMIglA9Tq1CUiuVc16CgoIUtoQgXEdlfqc5XwRBuAMeKWAqoWEjwpOg/E4QhDvh0QKGIAiCIAh1QgKGYJoBAwbg888/l83/EydOoEGDBiguLpbtHQRBEIT0kIBREa+99hr69u0r+vmkpKRqh22yzJEjR/Dvv//inXfeke0dLVu2RJcuXTBr1izZ3kEQBEFIDwkYglm+/fZbPP/88wgJCZH1Pa+//jrmz5+P8vJyWd/jDnAch1KdQWkzCIIgSMC4E7NmzULr1q0RHByMhg0b4q233kJRUREAIDk5Ga+//jry8/Oh0Wig0WgwadIkAIBWq8X777+P+vXrIzg4GJ07d0ZycjLvb2XPzYYNGxAbG4uQkBD06tUL169fN3v/zz//jFatWsHf3x9169bFyJEjAQCDBw/GE088YeZWr9cjMjISP/30k8WwGAwG/P7773jyySfNrjdp0gRTpkzBwIEDERISgsaNG+Pvv//GjRs38PTTTyMkJARt2rTBwYMH+WcuXbqEJ598EjVq1EBwcDBatWqFf//9l7//2GOPITc3F9u3bxcW4R7IeyuOIHbiepzLKVTaFIIgPBwSMPhvoy9duSJ/Um6E7OXlhW+++QbHjx/HokWLsHXrVowbNw4A8MADD+Drr79GWFgYrl+/juvXr+P9998HAIwcORIpKSlYtmwZjh49iueffx69evXC2bNneb9LSkrw1VdfYfHixdixYwcyMzP55wFg/vz5GDFiBN58802kp6fj77//RrNmzQAAQ4cOxfr1680Ez5o1a1BSUoIXX3zRYliOHj2K/Px8dOrUqdq92bNno2vXrjh8+DD69OmDAQMGYODAgXj11Vdx6NAhNG3aFAMHDuTjdsSIEdBqtdixYwfS09PxxRdfmPXq+Pn5oV27dti5c6fYqPcYVh2+CgBYuDNDYUsIgvB0PHIju6qU6g1oOXGDIu8+8VkCgvykSYbRo0fz/6/sqRg2bBjmzZsHPz8/hIeHQ6PRIDo6mneXmZmJxMREZGZmol69egAqDu1cv349EhMT+Qm0er0eCxYsQNOmTQFUiB7TM6ymTJmC9957D6NGjeKv3XfffQAqxFPz5s2xePFiXlAlJibaHB66dOkSvL29ERkZWe3e448/jv/9738AgIkTJ2L+/Pm477778PzzzwMAxo8fj7i4OGRnZyM6OhqZmZno168fWrduDQC46667qvlZr149XLp0yVb0EgRBEAxBPTBuxObNm9GjRw/Ur18foaGhGDBgAG7duoWSkhKrz6Snp8NgMOCee+5BSEgI/7d9+3acP3+edxcUFMSLF6BiO/rKrelzcnJw7do19OjRw+p7hg4disTERABAdnY21q1bh8GDB1t1X1paCn9/f4t7l7Rp04b/f1RUFADw4sT0WqV977zzDqZMmYKuXbvik08+wdGjR6v5GRgYaDOeCHNoSxmCIJSGemAABPp648Rn8px27ci7peDixYt44oknMHz4cEydOhU1a9bErl27MGTIEOh0Oqu7DhcVFcHb2xupqanVtpc37R3x9fU1u6fRaPghmspt6m0xcOBAfPDBB0hJScGePXsQExODBx980Kr72rVro6SkBDqdDn5+fmb3TG2pFDiWrhmNRgAV4ikhIQFr167Fxo0bMW3aNMycORNvv/02/0xubq6ZQCMIgiDYhgQMKho8qYZxlCI1NRVGoxEzZ86El1dFx9qKFSvM3Pj5+cFgMF9B0r59exgMBuTk5NgUFLYIDQ1FkyZNsGXLFjzyyCMW3dSqVQt9+/ZFYmIiUlJS8Prrr9v0s127dgAq9mmp/L8zNGzYEMOGDcOwYcMwYcIE/Pjjj2YC5tixY3juueecfg9BEAThGtTdansg+fn5SEtLM7tWq1YtNGvWDHq9Ht9++y2efPJJ7N69GwsWLDBz16RJExQVFWHLli1o27YtgoKCcM8996B///4YOHAgZs6cifbt2+PGjRvYsmUL2rRpgz59+jhk16RJkzBs2DBERkaid+/eKCwsxO7du81EwtChQ/HEE0/AYDBg0KBBNv2rU6cOOnTogF27djktYEaPHo3evXvjnnvuwe3bt7Ft2zbExsby9y9evIirV68iPj7eqfcQBEEQroPmwKiM5ORktG/f3uzv008/Rdu2bTFr1ix88cUXuPfee7FkyRJMmzbN7NkHHngAw4YNw4svvog6depgxowZACom1A4cOBDvvfcemjdvjr59++LAgQNo1KiRw3YNGjQIX3/9NebNm4dWrVrhiSeeMFvFBADx8fGoW7cuEhIS+AnDthg6dCiWLFnisA3WMBgMGDFiBGJjY9GrVy/cc889mDdvHn//t99+Q8+ePdG4cWOn3+U50CQYgiCURcNJuY6XIQoKChAeHo78/HyEhYWZ3SsrK0NGRgZiYmIQEBCgkIWeR1FREerXr4/ExEQ8++yzdt2XlpaiefPmWL58OeLi4mSxSafT4e6778bSpUvRtWtXWd7BClLk+yYfrAUAvHx/I0x7trUd1wRBEMKx1X6bQkNIhOwYjUbcvHkTM2fOREREBJ566imHngsMDMQvv/yCmzdvymZbZmYm/u///s/txYvU0CokgiCUhgQMITuZmZmIiYlBgwYNkJSUBB8fx7Nd9+7d5TMMQLNmzfgN9wiCIAj1QAKGkJ0mTZpIuuMwQRAEQdAkXoIgBEMjSARBKA0JGIIgCIIgVAcJGIIgCIIgVAcJGIIgCIIgVAcJGIIgBEPLqAmCUBoSMARBEARBqA4SMAQhMx9//DHefPNN2fxfv3492rVrx5++TRCEeP48fAVjlqdBV07liXVIwKiI1157DRqNBsOGDat2b8SIEdBoNHjttddcb5iMNGnSBF9//XW165MmTZLklGq5ycrKwpw5c/Dhhx9Wuz5q1Cg0a9YMAQEBiIqKQteuXTF//nyUlJTw7po0aQKNRgONRgNvb2/Uq1cPQ4YMwe3bt3k3vXr1gq+vryTnRjmKhhZSE27KmOVH8Ofhq1h+IFNpUwg7kIBRGQ0bNsSyZctQWlrKXysrK8PSpUsFHb7oSjiOQ3l5udJmKMLChQvxwAMPmB0UeeHCBbRv3x4bN27E559/jsOHDyMlJQXjxo3DmjVrsHnzZjM/PvvsM1y/fh2ZmZlYsmQJduzYgXfeecfMzWuvvYZvvvnGJWEiCE/gdoleaRMIO5CAURkdOnRAw4YNsWrVKv7aqlWr0KhRI7Rv397MrdFoxLRp0xATE4PAwEC0bdsWv//+O38/OTkZGo0GGzZsQPv27REYGIhHH30UOTk5WLduHWJjYxEWFoZXXnnFrFdAq9XinXfeQWRkJAICAtCtWzccOHCgmr/r1q1Dx44d4e/vj19//RVeXl44ePCgmY1ff/01Gjdu7PTwx++//47WrVsjMDAQtWrVQnx8PIqLiwEABw4cwGOPPYbatWsjPDwcDz/8MA4dOmT2/KlTp9CtWzcEBASgZcuW2Lx5MzQaDVavXs27uXz5Ml544QVERESgZs2aePrpp3Hx4kWbdi1btgxPPvmk2bW33noLPj4+OHjwIF544QXExsbirrvuwtNPP421a9dWcx8aGoro6GjUr18fjzzyCAYNGlTN/ieffBIHDx7E+fPnBcacOGgSL0EQSkMCBgA4DtAVK/MnYov9wYMHIzExkf/9888/4/XXX6/mbtq0afjll1+wYMECHD9+HGPGjMGrr76K7du3m7mbNGkSvvvuO+zZs4dvpL/++mssXboUa9euxcaNG/Htt9/y7seNG4c//vgDixYtwqFDh9CsWTMkJCQgNzfXzN8PPvgA06dPx8mTJ/HUU08hPj7ezG4ASExMxGuvvQYvL/FZ8fr163j55ZcxePBgnDx5EsnJyXj22Wf54wsKCwsxaNAg7Nq1C3v37sXdd9+Nxx9/HIWFhQAAg8GAvn37IigoCPv27cMPP/xQbchHr9cjISEBoaGh2LlzJ3bv3o2QkBD06tULOp3Ool25ubk4ceIEOnXqxF+7desWNm7ciBEjRiA4ONjicxob6uDq1av4559/0LlzZ7PrjRo1QlRUFHbu3Gk/wgiCINwAOgsJAPQlwOf1lHn3/10D/Cw3ZNZ49dVXMWHCBFy6dAkAsHv3bixbtgzJycm8G61Wi88//xybN29GXFwcAOCuu+7Crl278P333+Phhx/m3U6ZMoU/jXnIkCGYMGECzp8/j7vuugsA8Nxzz2Hbtm0YP348iouLMX/+fCQlJaF3794AgB9//BGbNm3CTz/9hLFjx/L+fvbZZ3jsscf430OHDsWwYcMwa9Ys+Pv749ChQ0hPT8dff/0lKPxVuX79OsrLy/Hss8/yQzWtW7fm7z/66KNm7n/44QdERERg+/bteOKJJ7Bp0yacP38eycnJiI6OBgBMnTrVzPbly5fDaDRi4cKFvMBITExEREQEkpOT0bNnz2p2ZWZmguM41Kt3J2+dO3cOHMehefPmZm5r166NsrIyABXzmb744gv+3vjx4/HRRx/BYDCgrKwMnTt3xqxZs6q9r169enyeIAiCcHeoB0aF1KlTB3369EFSUhISExPRp08f1K5d28zNuXPnUFJSgsceewwhISH83y+//FJtmKFNmzb8/6OiohAUFMSLl8prOTk5AIDz589Dr9fzggcAfH19cf/99+PkyZNm/pr2PABA37594e3tjT///BMAkJSUhEceeQRNmjQRHxkA2rZtix49eqB169Z4/vnn8eOPP5pNcs3OzsYbb7yBu+++G+Hh4QgLC0NRUREyMysm6Z0+fRoNGzbkxQsA3H///WbvOHLkCM6dO4fQ0FA+LmvWrImysjKrwzaV85QCAgLshmH//v1IS0tDq1atoNVqze6NHTsWaWlpOHr0KLZs2QIA6NOnDwwGg5m7wMBAs6E+T6ZMb6ADRAnCzaEeGADwDaroCVHq3SIYPHgwRo4cCQCYO3dutftFRUUAgLVr16J+/fpm9/z9/c1N8PXl/6/RaMx+V14TM0el6hCJn58fBg4ciMTERDz77LNYunQp5syZY9OPsLAw5OfnV7uel5eH8PBwAIC3tzc2bdqEPXv28MNdH374Ifbt24eYmBgMGjQIt27dwpw5c9C4cWP4+/sjLi7O6tCPJYqKitCxY0eLK33q1Klj8ZlKUXn79m3eTbNmzaDRaHD69Gkzt5WCMTAw0KI/zZo1AwDcfffd+PrrrxEXF4dt27YhPj6ed5ebm2vVFqlheQrMjUIt7pu6GQ80rYWlb3RR2hyCIGSCemCAihmJfsHK/ImcDVk596JybkZVWrZsCX9/f2RmZqJZs2Zmfw0bNhQdVU2bNoWfnx92797NX9Pr9Thw4ABatmxp9/mhQ4di8+bNmDdvHj/sY4vmzZsjNTW12vVDhw7hnnvu4X9rNBp07doVn376KQ4fPgw/Pz++p2f37t1455138Pjjj6NVq1bw9/fHzZs3zd5x+fJlZGdn89dMJyUDFZOnz549i8jIyGrxWSmkqtK0aVOEhYXhxIkT/LVatWrhsccew3fffcdPMhaKt7c3AFRbiXb+/PlqE7mVhuM4TF17Ar+kXHTZO9cerfgY2XP+lsveSRCE6yEBo1K8vb1x8uRJnDhxgm/QTAkNDcX777+PMWPGYNGiRTh//jwOHTqEb7/9FosWLRL93uDgYAwfPhxjx47F+vXrceLECbzxxhsoKSnBkCFD7D4fGxuLLl26YPz48Xj55Zct9jiYMmbMGKxduxZTp07FyZMncezYMXz44YdISUnBqFGjAAD79u3D559/joMHDyIzMxOrVq3CjRs3EBsbC6Ci12Lx4sU4efIk9u3bh/79+5u997HHHkPTpk0xaNAgHD16FLt378ZHH30E4M6E2v79+6N27dp4+umnsXPnTmRkZCA5ORnvvPMOrly5YtF2Ly8vxMfHY9euXWbXK8Vbp06dsHz5cpw8eRKnT5/Gr7/+ilOnTlVLz8LCQmRlZeH69evYv38/xo4dizp16uCBBx7g3ezdu5fvWWKJo1fy8ePODEz867jSphAE82TcLMbbvx3GqawCpU1RBSRgVExYWBjCwsKs3p88eTI+/vhjTJs2DbGxsejVqxfWrl2LmJgYp947ffp09OvXDwMGDECHDh1w7tw5bNiwATVq1HDo+SFDhkCn02Hw4MF23T7wwANYt24d1q1bh65du6J79+7Ys2cPtmzZgnvvvRdARTzs2LEDjz/+OO655x589NFHmDlzJj/J+KeffsLt27fRoUMHDBgwgF8CXom3tzdWr16NoqIi3HfffRg6dCi/Cqly/kpQUBB27NiBRo0a4dlnn0VsbCyGDBmCsrIym2kwdOhQLFu2zGwIrmnTpjh8+DDi4+MxYcIEtG3bFp06dcK3336L999/H5MnTzbzY+LEiahbty7q1auHJ554AsHBwdi4cSNq1arFu/ntt9/Qv39/BAWJG5IUiq2VUqYUaV2z/8+1vFLk074dhMRk5Zdh+rpTuHJbvrll528UYdSywziTXYhBP+/HP0euoe/c3fYfJKDh3HSmW0FBAcLDw5Gfn1+tgSkrK0NGRgZiYmIcmmBJSMvkyZOxcuVKHD16VGlTrLJ7925069YN586dQ9OmTUX7w3EcOnfujDFjxuDll1+W0MI73Lx5E82bN8fBgwdtilMp8n2TD9YCAF57oAkmPdXKrvvd526i/8J9AICL0/uIeqc9bhVp0XHKZv4dSbszMOmfE7K+k7DO/oxcrDl6DWMTmiM0wNf+A4xRmcfffewebDyRhWNXC9CkVhCSxz4iy/semrENmbklCPX3QaGJ4PfkvGur/TaFJvG6AUaOw81CLUICfBDkx26SFhUV4eLFi/juu+8wZcoUpc0x488//0RISAjuvvtunDt3DqNGjULXrl2dEi9ARU/FDz/8gPT0dIksrc7Fixcxb948p3vW1MqprEKlTSBMeOH7FACAt5cGnzxpX+SyzLGrFUM5F2/J1wOTmVvhd6GLeivdCXZbO8Jhcot0yCooAwqANg0ilDbHKiNHjsRvv/2Gvn37OjR85EoKCwsxfvx4ZGZmonbt2oiPj8fMmTMl8btdu3ayntvUqVOnakvWPRlHh7cIebl4U9wkdYJwFBIwbkBZufl+ICW6chg5IMTf9cl7Na8UunIjmtQKqtaQJCUlISkpyeU2OcLAgQMxcOBApc0gROCeg+CEp6DRWM7DBWV6XL1diti61odQPB2axOuGnMspwoUbRdAbXH8c/K0iLQrL9Cim7lBCJrLyy/DEtzvptGDCLfD1ttwM95i5Hb3n7MT+jFyL9wkPFzCszV/mOA6Zt0qQU1AmiX9KCJg772YrbgnL+Z3jOPyw4zz2XhC2Z4qSozSf/3sSx64WYPwf6YrbQrgvQsuENQxGDuuPZSGn0HK97utlOQPfKKzYkXvrqRxJ7HBHPFLAVO40y9q260XacuSV/jefRQKU1GflRhIwrFGZ3013Wt5wPAuf/3sKL/2wVymzBFO1d09sPj94MRe/7r3E3IcMwQa2NkI8eiUPj8/Zid3nblp1U8mSfZcw7NdUPDZrh8X7PlZ6YCoJ8qu+zxdRgUfOgfH29kZERAR/vk9QUPX5GlKhKzfCYDQi0IHVQWVlenDluv/+77iIKddpwZXr+efu+FEKb861yxgr311aBpT5UsOgNIWlOuiNHPxRjpycHERERJhtlHf+hnomWuoNRjw8Yxuu5dsuG44W5ecWVKyWaVQzCA/d45ojGAj34NWF+1BQVo7+C/fZXe685WRFO5NfanmfIl9v2xmWBIx1PFLAAOAP7qsUMXLAcRyu5lVUtlFh/mZjnYVlemjLjagV7MeLp1KdAbeKKwSAX6ntHWpNuV2iQ7HWwD+Xc7tii3ljgR8CfOXP/KU6Awq15agZ5Iucgopuz9IAHxQHqm8PCHfjyu1SABwiQwMQWbum2YGVAFAucqhPo8BpSMmnb9gVL4DwHplzOUUkYDyQMr0B4/84ikdbROLpdvXtPwDgdFYhIkP9UVDm+Bw/e9nR28oQUiWBJGCs4rECRqPRoG7duoiMjIReL88OnjM3nsa/6RUCaULvWMS3jOLv9ZiZDAAY36sFeraqaFR2nrmBSdsqtlzf8l53h9+zyuQ9W97rjqGrKvye+sy9iI2pbeNJaagMS+sG4Ui/UnHw4gv3NcT/HpJnX5KcwjJEBPrCz8d9CnZeiQ6lOgPqRlQXrp/8dRzacgOmPdva4Z5CXbkRNwu1mLRqH4wcMPvljqhbt3ojXe7gIZ2FZXqHusuFois3okRXjoggP7tuy/QGu25KdQb8uveSIBuUnCvmSoxGDl52GksW0JYb8Nz8FLRpEI6pz7SW7T1L9mXir7Rr+CvtmkMC5lRWAXp9vRN+doZ8hOLjZdu/wP8+Qq/cLkH9iEDaJsAEjxUwlXh7e1s8S0gKftxz54ycco2P2e6nVwsrKuMSgxd/3eDly18XslNqoV5j9lzl/7Wcj11/9mfk4lDmbbz54F2iK7fK94Xe1lcLF8dxyCvRo0ZwRQO1Lv066oT6o1OTmmZ+5Bbr4OfjZXfp98nrBeg9ZzdaRIdi/eiHRNkrNUm7M1AnNAB92tQV7UeXSVsAAKkfxaNWyJ3Twkt1Biw/nAUAGPmYAc0iQxzy74XvduHoFZNTvDWWK0lH5iqtPnwVo5enOfReocTP2o7M3JJq4baEJbG14sBl5JbcOVV8xoZTOJtTJMgGTxAwk/4+jrXp17Fh9EOoGWxfLMpJmd5gs2d404lspF/NR/rVfFkFTG6xVpD77advAAB0EucXe8I8yM8bv+69hI9WH8PgrjGY+KT9Q3M9BY+cxKs009ad5P9vbwKXMzgyPPDC9ymYvu4U/j12XRYbxixPQ/vJm3DwYi6yC8owfMkhPLcgBdpyA/Zn5KJEV47CMj06TN6Eez/ZYNe/v49UnDTMyu6r53IKMemfExix9JBoP0wnkVadk8KZdEC/lrjfIf8u3iw2Fy82MNgRMNkFZbKJF+DOLqQHLtpfKmppZdu4P45i+rpT/O8dZ24ItkHnASvmkvZcxI1CreDeKalZl34dLT5eb/N08hKd/Z62SvJL9Ch3kQBdfuCyIPfacgMen7PTLE8eyrzN/79YW477p27mpw1Yw9/Hm8/jP+/OEGSDnDjSIyo3JGBcTOatEny//QL/28dKr8fNImFfB5aw9MWqNxjxwoIUTPrb/HRgKXbNtNQYrk6rEBxzt51DqUnF9P7Ko3jh+xQM/Gk/ztn4YjYaORy7ms/sV3JusfPDj6ZfdIE2vkyv/De3yRaHMm+j+1fJDr/bXryWWmlMxPRiZ1bZjl1rsgFjeKD9XgGx83Xs+8tm3qrEYOTw3Pw9GOmESHaEy7kl2Hg8S9ZVWZVC39bp5Npyx9Ljyu0StP1sI56euxtJuzPwwoIUFJbJd6DnBYF15J7zt3Diuvmp0s/O28P//5utZ5FT6Fg9HxbA1mDJ99vPo8XH67FdxAeDlJCAcTFVd8319tKA4zisOXoNGSYFpNOUzdCZFOSCMj1WHrxsdSa7JSx9sW4/fQP7L+Yiac9FvPnLQUG268qNNlW3ra77Im05vExavX/+60k5eOm2tUcAAN9tO4cnvt2F91ceEWSrXExbdxKvJ+6323MhhDKdiYDxc6xIJp/OwcKdF6o1Nv8eFdaTZi8cUk4gfOjLbWa/TfNyqAMVtJQi9nr+HTHIqjiu5MS1Ahy8dBtrBKatUB6csQ1vLk7FhuPZkvi37fQN9JiZjGNX7/QG+pvMW5uw6ijySqr3Pmgd/LJfl14xtHr8WgEm/XMC+y/m4uddF50zWkKC7CyguJzr+DYejh6KyXEc/kq7ijPZ8vZQT/uvR2iswvUyCRiF8fHSYOOJbIxcehhfrD9lds90B8Z3l6dh7O9H8daSVIf9tlQxm8552HhCWEXV+fPNiJ243rGuwyrtYpFWXHfjvORzAIC//uvJkXP6GsdxGLHkEMb9br1Qfr/9AradvoE956Wb0Fqiv7OiwdvOhL6Nx7Mw4Kd9eC3xAKasPYkDF20LQHuY5ofP/jvB2RRr8W16Pf1KPv7vz3TBvYZ5JXcEjCM9OlIJjY3HsxA3bauJv64ZQirTG7D9zA2znidHMLp4nxopd349f6MY/1t8p84K8L2Tv3/bfxmframe5xydY8JZWN9jWpaURsqTuB0R+EDFpnejlqWh52zzPWfWHr1udWO+3eduiu6BV3oYiQSMi5iy9gQKLHRv+nh7Ie1ynsVnTCv1zf/tJbD7nOO7Q0rdNX67RA+OuzNvQQjF2nKLFY49vKq0bKY/l+yzPp5/o1CLp+fuxrL9jm83fzm3FGvTr2PFwSt2GxlHGtO0y3noMTMZ207bXqpvOkzz8g97bXbhv7k4FTvP3hFPt5wcajSYNN4/785AdpVNFB1JsSe/24Wl+zLx4Z/CTty+lmd/SMwUqYTG9zsumP2ubDBTL93GzSItyg1GTPr7ODYez8LhzNt4dt5us7kLYnlvxREM+nk/PrExfOIoJbpybDuVI1gMKUGpSSPn52Pe5GRYaDi1erZ7xBzF3vJoIYQ4KGDSTXq7Kns4M24WY8TSQxY3qzx2NR/9F+4TNOxsSpnCaUUCxkXcLNJh9qYz1a5bmwMjBWIq/Cu3S1Cik+ArRlOxYqiSIgfPRtIbjDh/485QVFUBY8qHfx6z+gXw6T/HceRyHj5YdadRNRg5vLfiCH6zImoMAr50DUbgcOZtm2c+DfhpH87fKMbriQds+mU6MS+roAy3TXom7JkU4OAQT6neYLHLWl9lnpQzvRxns4Wt/rl0S5gQlmuuir7ciJTzt9Bv/h7cP3UzVhy8UjHEujgVzy1IwaHMPLO5C2JZm14xBLRM4GTQPAvDxiOWHMLrSQcw2UIPBsv429j6YPOJbJy8XoAVB4XFjxSYTgRnDQ6czaXbxdpyvPRDChKrTPD9ePUxAObDpVVxdjGE1CuyhEICRiL2Z+RicNIBnMspwp7zNy02rJYaEClVelUsTeK1te/H2exCdPtiGx6asa3aPUu9R/b4xGSisKOHOw5OOoAeM7fzIsNe7FhbBnziWkG1ax/8cRR/HLqCCavsD3foyo2Yn3we53IsF/A3fjmIZ+btwetJ1sWJo2H+da+5oBKSIwIc3Avn9cQDeHDGNpytMjYudi6PrSGfwjI9vt9+3u4Yf9XVF2V6A95feQQbjmdZdK+X6XgKvcGIHWcrJiMaOZgd5SHlXCdHyC/R41SWed4d9HP11Wfb/lvSW5l3SnTlmL3pjMV8b4rSpyaYDiEBd+w5djUfQ385iN5zduK6A5sVSs2C7eel+XD7j+TTOfjz8FXJ/LM13J+05yL2XsjFp1WGgB3Zt8nWogFruLpM2IKtqc0q5oXvK7Ylrzx46+l29aq5sbTSwlYPzMyNp9G1mfiN6Cz1wIxcetii23nJ5/mdgm8WmTcsKedv4eUfBZ6Vw5mPpTua5SuHRyasSsd9TWqaNZTF2nKHd4A1nd3/3dazmJ98HsUmQzVjlqdh8ZDOVp/feioHX6w/hS/Wn0LGtMeh0Wgkq+DyS/RYvPcinm5XHw1rBqFXq2ist9Jo25sbUtkgrDh4Gd9tPYemdYJtut96Kgd3R4Xyv+3tAyOmwfvk7+NYdegqZm48U+0LLb9Uj3ALOzRzHPDTrgz8nnoFv6dewczn22J/Ri4+6N2C30NIrsm2LB082u2LrSjUlmP1iK5o1zDC4ecWbL+Ab7acxZwtZ+1uba8k1npgbK1EtIbUYkzKdvk1O72uQrh40/aHgLV6qfIj2tZelaaLBjiOg8HI2d3ag6VeP+qBkYnKSaemWKq4bWWWQ5l5gpcFms6fMP2/ttyAfTZOV7W190L/hdIc9GetwjGdzGnK4pSLZj1UD83YhtMOzq43nfj41cYzZuIFqBBKn/5jfS6Cae9BZSM8duVRu+89m12IFQcvV5vHsstk3sr4P47iq41n8My8PUg5f6uaeDF90l4lXSk6x/1+FJm5JfyXuTWqeifHsEzKf4fgWepe7jEz2eo7TeffvLfyCJYfvIxRy9OQV6LD8gOZFles2OPnXRl4eu5u5FvJY0B1YWSvF8OUYm055ieftziXo5KbRVqHJ0kW/tdrt1Pg8lTTYVdbKL2Jq3+VOTBK26MGSqv05s/ddg7dvthabb4aAKw6dKfXp1RvwIUbRXj1p31W/Tbtwf1o9TG0nrTRbs9p0p6LDlouPyRgXEhYYPUOL3v72Dm407tFTLecnvBHOl60c+KwtbZS1JeJxuZPM37aZX1zJtMw3CrWYVOVrtRDl25bHHJw5OsscfdFpJy/ZfHL3tKXYuUcBls8NnsHxv1+lN9wr5KBP9+pRCpXMN0s0uJ/i6svZX/dwQ3rgIoGoOrqNSFY6w6en3weCbN34LYI0WCLm0V3zu0yTaT0q/kW95zZceYG3lycivF/pOO3/cLnRny25gSOXM7D9zvOW3VTNQ42n3R8dd6An/bhi/Wn8IiNSZCdpmwWPElS6PL1GkHWV7wYjaYfNfb9OptTKGovGG25AcMWV18lmVusQ5MP1qLn7O3V7lW+RswEf0tcvFmM+cnnVTG52VGqpsWXG07jyu1SdP58S7V7pgssjBzw6MzqcW6Kv8mQ3pJ9mSjVGzAv2XpZYQ0SMC7E14JaEXOuxcS/jmHbKWGHUK6ScDxWDLaqJ1tjqvamCA38eT/+tzgVzy/Yg6NX8kze51iF+PKPezH+j6NmlbyQ561x5HK+xbS9Uag1OwjO0qFwR0x20V240/bOmxxXITbEsGjPRX51mynL9mfii/WncDq7EN9vt+y3U+exWHh0wqp0rEy9Uv0GpFnWa2u1xJZTOaJ6opbuy8ShzDybbsT2cAX4eqNMb8DLVT46qm5AWUkNK2dJGY0cPv7rmKB37zx7E/OtpLstVhy8YnUoFADOZBfZ3fdJCJZK6Ibj2fhi/Sn8baEH3OxZjsPcbeppqK1hqfwKwdI5TFXrQlOE7EPmCpwSMNOnT4dGo8Ho0aP5a2VlZRgxYgRq1aqFkJAQ9OvXD9nZ5l80mZmZ6NOnD4KCghAZGYmxY8eivNy8Ik9OTkaHDh3g7++PZs2aISkpyRlTZePAxVyrlYpc/JJyyebk0UqEfkVJ2ptr4dXWrLElFhxtKA9cvI2nvtt9x08BQV916KrNJdlSjbWfvF6A+6ZuFvTM7M3VV65JxScW8m1eid5s5ZalFTAA8MOOC4rvASEUo5GzOgz1ox2haIn/c2DZeLHI/Y8CfL2x7th1pFQZ9rXWfW+6T4ipaHr/9yNYss/xrQQqmbXRer4rNxgxde0Js+0BJv19nF/1IoTK4i316eZX/1uiby2P7r0g3V43SmJrhZE1Csr0+G1/Jm5bOcLAwFWUk6rTF/Zn5KLtpxtF2SkXogXMgQMH8P3336NNmzZm18eMGYN//vkHK1euxPbt23Ht2jU8++yz/H2DwYA+ffpAp9Nhz549WLRoEZKSkjBx4kTeTUZGBvr06YNHHnkEaWlpGD16NIYOHYoNG+yfleNqnl+QwtSYoNKrDKwhtnpy1eG5Vb/+pa5QAeAPKz0MrsQ0f1g6OPPF/yajV5JsY07NarG9egrl0UGJ+9Hus01my/vlplDr+BeraWMb6OttdQuB9ceq93JYc2s6JwJwfM6JrcndKw5ewY87M/jtAYq15aLrQGfqK1vPBvl543DmbbSetMHiEulcO+cPsYSUdfrCnRfw3oojmLAqHUN/OWixKJboytHus01oPWmj2Ufw3G3npDNEIkQJmKKiIvTv3x8//vgjatSowV/Pz8/HTz/9hFmzZuHRRx9Fx44dkZiYiD179mDv3oqu0I0bN+LEiRP49ddf0a5dO/Tu3RuTJ0/G3LlzodNVZKoFCxYgJiYGM2fORGxsLEaOHInnnnsOs2fPliDIbCGn4BDazS+pKQJebUss2NoHxhbOhkWqMXmWsXQqcdXJzrZwdlWQq2O4coWbkMMCncXR/Y82nchGi4/X878D/bwQZmUn12G/Or4bd1WcrW8Wp1w06xXML9Hj4S+TnfNUJlamXoHewGGBiOEwtSC0dpyy9iQ/jzDVynBe1Xk0APBLykXFzz2yhCgBM2LECPTp0wfx8fFm11NTU6HX682ut2jRAo0aNUJKSsWXXUpKClq3bo2oqCjeTUJCAgoKCnD8+HHeTVW/ExISeD8sodVqUVBQYPZHyINDlaAErdOilEt2l/lK+X7TyoDVnixCXTiyF5DByOGNKueS+Xl7O7z7alXkyro3CrX4+K/juGGyRcG85HOSHDwrB9FhAfz/pT6g0p1XT1X9Lkm7nGfz8E0lESxgli1bhkOHDmHatGnV7mVlZcHPzw8RERFm16OiopCVlcW7MRUvlfcr79lyU1BQgNJSy2N+06ZNQ3h4OP/XsGFDoUFzOxSdA2MBMZXIDQdPa5UCU+ukqKDcuI5THa7uUTMaOey9cAvZBZbz76NfJWPrqYov4X+OWJ5wKucml9fzSwUtFwcs7zdS9VgGobhKCHSaslnS88vc5QPHUp1c9dp1gUd+uBJBAuby5csYNWoUlixZgoCAAPsPuJAJEyYgPz+f/7t82fXbUbOAM+XK0Wfd+etDSlis46RsyMX65AnDc0v3Z+KlH/birSWHLN6/cLMYg5Mqel2uStBACG1Q46ZtxePf7BR0IrIcSL2M2hq3inUY8JPj2xOwhKtLC0s77dpDkIBJTU1FTk4OOnToAB8fH/j4+GD79u345ptv4OPjg6ioKOh0OuTl5Zk9l52djejoaABAdHR0tVVJlb/tuQkLC0NgYKBF2/z9/REWFmb2pw7kyyxOLXV1+uXKvRoQVyGqUZep+UvQlbbLMSnbFn+lKbttgaMcF9gLoxYs5S0pG2ZXVq2uLuNCzoRTGkECpkePHkhPT0daWhr/16lTJ/Tv35//v6+vL7Zs2cI/c/r0aWRmZiIuLg4AEBcXh/T0dOTk3FmCt2nTJoSFhaFly5a8G1M/Kt1U+kE4BmvLqF2JmDJodZm3yLCoURAJQUX1nOp6fThwguJ36r8n7zyrrqACECcwhaYp9Rw7hq19YFhD0Cyx0NBQ3HvvvWbXgoODUatWLf76kCFD8O6776JmzZoICwvD22+/jbi4OHTp0gUA0LNnT7Rs2RIDBgzAjBkzkJWVhY8++ggjRoyAv78/AGDYsGH47rvvMG7cOAwePBhbt27FihUrsHbtWinCTFhB7myrnmKhzkZALVDcElWRWmC6k1iRMyyWYt20B0bqyc9SI/lhjrNnz4aXlxf69esHrVaLhIQEzJs3j7/v7e2NNWvWYPjw4YiLi0NwcDAGDRqEzz77jHcTExODtWvXYsyYMZgzZw4aNGiAhQsXIiEhQWpzFUfq/MFShtNo7oTP1RWKBKcfuCUMZQ/V9YoQhBLYLLMyVKzOHF/japwWMMnJyWa/AwICMHfuXMydO9fqM40bN8a///5r09/u3bvj8OHDzprn0cg1B8ahRlCjbGPJkpCrRA6LlBQBLMaxnHAc5+J5ZZ4Tv6KGkBSMHg3cN3WMKirXdBaSG6NoAyPg1VW3S3fx6+0+Tz0FBODcBnJCcWqHWsqvFpFSerpzDJOAIdwaRz9Cq7pTUbmQBU8YonIEtWaDDccdP6WaEIYcoktofeOJ5dPyai2T+64zRRQkYBRG6gzi6gy3+5yVzaGUXkYtcUR8brLKg5AGTxe0hHxIuQrREq4dSHTxJowqKpgkYNwAaz0irpgDs+2Uc8e5s4ppGHOLdfhBxI6jrpguoZJRQkIEYtNWRe0Pjxz79FQtf4rui6Uiyp0848yVkIBxA6xVWOzNgVFPzWr61aM3qMduwj3wtBwnxzJqTxhCct7m6pGkom1gSMAQwhHzIePqnVCdRW32OgoLK4cqbVDeEoK4A+XHCkyHkBioLmxCAkZhpM4gBaV6aT20g1UxU3UCr8qqB1kmFUruo3N+OtujznrlRrCPXFnInfKmy48SUFEXDAkYN8C0Ieo4ZbNyhpiinjLAY61BZ6HXQg6cDZZ7xgobcJw8kl9tHxK2UPKoFFfOp3F1iqmpuiMB4wawmuE0Zv/XMGtnJXLb556DUsJhPBuojh9FTDBnBWfKhK3y6k7zdeUMisVl1KZDSIyXVhIwBI+YBtwTZvazXYSFYZrGSicd64JWLUwVsMTfXed2SYEae1rlKMM0hEQ4jJyFRumjBNSMFMniisbCmfyjwvrao/CEZdSVppKwso6KktPlkIBxYwSPEdNe24SboqZGHfC84uOKoQpX9TiqsSdHrZCAUZisgjIYVdRlB4ivCNQVSvfH+VVI4lKU6neiEqfmwEhmhediLw5ZL6tOn0ZNOMeoZWlKm8DDemZVOywso2YpiV05QVDp+T6EZVhaRs1S2TCDKmarUA8MIRiziaDWHKmwwTBt5DyhG9gDgsijurByTvRw2WiKWV9VIiVVwyrlPBtbPkmd1zwnxYRDAobg8fQ5MNaPZHD0+SoOq57FItwkt0TORlSF2c6jkatMuFOPm62gOCvKVCfsq0AChuBxNDM7WjmYroISczaJ6uFs/lQcxSt51iLEA2BttY8zWcBefSI0rB5XP7kBJGAIwYgaX1ZB5aB4gy4CRc/rVEGaEuZ4yhASx7lPWN0jFPJAAoZxHCmEUjW8ovxRYaMvHJFVCA0hESLhnGh+SVhawVXLqKX2j9LTKiRg3ACpMrjUQ0hVn1HTF5FpXDg+B0YeW6SECRvltIGF8BEuwVZ9osbeVGvIGRZ7E8WZqC9sQAKGIFwEC3WBlCLSWb9YiA9Pg7U5MEzBaIZkXUQoCQkYQjCeVqDY/przsMTwIEQfJWDznufmF1cVY0/YgoEVSMAwjiNlge0GVp2YRrvY+qhqsghNpuyCMnEvJlQPtYFsw0qdy4odSkECxg1wdWVnWmhsdUlXvePulXLV4Nn7bYtbRVp0/nyLkxaxCU2BIaTA3esTV2D3KAHGSxQJGEI22M76bHM4M88hd0IrcSkrfWf9oq52Qi7EHSUgTX50xJeMm8UwGh31j8qJNegsJCe4nFsi+zs4AMXacsv3OM5sszhX4VDloPKuTSnaVmeGkIp1ltOcJahalQ9n9jEhYVhBtY2xGamTftufiQmr0pU2wy2gHhgneHDGNpe8Z8rakzbvs1IwzeDMG2wWTbSFaeMhRXMgVGgWWRGt7gC1r4TcMFkn/sd3W89J5pezwbRXFlkvqyRgVMDOszds3ldyDowQWC8MpohZbmrry5fjhH1PW+t1cyfkyA/U++A5sLphHGVB10ECRgWw9jVhdho1Y7ZJhdLjzkVag0PuqK50YyhxJcXeR8nwJanQljtW7lwJCSLrkIBhHNtf9RX/uquIcDXWKjipKhAhyVSigh4Yqljlw6lDDiWzwrPYeyEXKw9eccgtbQjIBiRgCOI/rPW6ONobU9VVVWEppGExytQKsdC4ydm7xUL4COGIEsMyKOjCMvY/HKTEXllkvTyRgFEB9tS+ZGO3DmZX831gHHMnxH8WMBsmU+BryxN61dSTGxyjRFeOudvO4VxOkdKmeDyKntJuJ2cLLdvuVk6khASMCrCW4ZXK2O50uKEpcgsVIb6rIY7VJEhdwYz1p/HlhtOIn7VdEv884jTq/4x1hWD3hI8CT4MEDOMYHKiNpCqYnj6u6+wQkn3/CXfmwMVcyfyi1VTO4SliRe5wsp4PScAwzuNzdjEnKxwtNGZDTRqNqr4MlR6Sl61iYiARKk2QZxm19H46ilzzltwaJzK6otFtw2x7eZApcaXyPEsChnFuFmntunH1HBhTHC2MrCt5gK0eKMeHkMTHq9ybYHkaasjjzCFTnFXsZGyOpKXbhUlN2co6dJSAC5m16QwMEn6mVVSYjB4loHZMAqnEMmpX4AnJ6EokLdtwIt9RwhIeAvXAuBCDkcOsTWcEP2dvG3o1zIFR4swmpakan0LaFXeOLndtX40qUPasmijq8EUbz2g0Mn8wOOG5lPWs3L3GjGYXHhIwKsBVbZmYZdTEHaRclSNXQ2PqrdLJKMcqJiVXRrEqDjwVd0kO2jfJOiRg1ICdZdSurjg9raJmeQiJdmxlB0dWDApB9GnUNp5j9ePDNcuoGQ08IRoSMASPmO5Ilia+so67DSGxOGlVyfwo7fw2ybxSBXKEl9UipIayrRZoEq+KOZdThOTTN1Cmt30AmaMNjdRdlRWNiXpqYtPwK221bENISgcM8gofGkJSGU4to2ZzG3w15QN7trIeFhIwKsBaEe89Z6fdZ5fuyxQ1cdimPSLqnCJtOQ5fzpPUDtZQorA7806lPwTdbx8Yxmt7FpFpJ15LSaF0fhcNZSurkIBxc/7vz3TJ/RRbT3+8+pi0hrgxruhmdrZepPbaHCmHkADx8UvpwjaqFVIMQnNgVABNPnM9po2A2GGJqjsRi30/4ThWzw1zQYRKq1/k2+CNRaS2S+4q05b3Lh3GlLtpYDS/VEICRgWwJl/MG2bl7JAaKUSLVf8YaTlMw6VU0skZFdb8XpRySb6X8u9mI409BU+JbjmDqfYDWUnAEILxlIqjEraXUYs3zh2TUcm8Kf0yaulh9YPDJXYxEnbqUZcOEjBujCsqc3ctilLEXdV6SoiXavgyYt9C12JkZBm1rUdZ/fhg1S4xuFNYXvpxL7TlRqXNsAoJGBVAgt01cFb+765QtpIWd2q41IDd6PaABJG7DJ+8XoDVaVdlfot4SMC4MXIJH4f9dZMW0vGToeW1QwpYstHd5oxIPYTkSajtI80Ze4U+Kuu+SQ54rdVTDwzhBKztdusJ9bTSYXT4/QraqXQcsYbU+8CIbbhsPceqUFDbcLcte6lYuA4SMCpAbKVDDYx4zHfldT4iaeKe+yPlMmq5ii7VCYQ7QQKGcA43apil7qo1jRl3Gy5xBneNCikn8RL2UTIfOVXtMVRlqj3HkoBxY4QWMkcrBDfSLNYx28PF0UdsO2ThNGopKyxne6bkqDyVFIp0lAA7WDxKQKUVF2Ur65CAIQTjCQVKlsZVBj8JdpC6A0asd2rKZ/ZsVWPPpRptViskYAgelX6gyI4U1ZFav/7UhrvEM7WBzuEm2QCA7bC4S34XCwkYFcBaJnXUHLasto/ZPjAmLYgUX1Tu+FUm+rDB/2JantOo3S+eWUGOqHWmjlDDZo+WEL6MWhYz/vNbnXFYCQkYFeAqISBmvxPbh5p5FlLWBY5WLELfqfL6yi5uFTw6jdp2WASG01XfgVJHvxslp+SQgHFj3KkiczXmBzs6D2u9aFJA2Us+1Nq7IBTPCCUhFyRgCB6p21i1NdnWRItYIWgqWtTeVSslckaF2vIcIR9yljhnNheV8mPG0/O7IAEzf/58tGnTBmFhYQgLC0NcXBzWrVvH3y8rK8OIESNQq1YthISEoF+/fsjOzjbzIzMzE3369EFQUBAiIyMxduxYlJeXm7lJTk5Ghw4d4O/vj2bNmiEpKUl8CN0AsfmdllEL470VaU49L+0SZUfdCXsrS1/2ctjCTujcD1lOx5bBT8DKMmqZ3uXIu1nyz8xv+bx2CYIETIMGDTB9+nSkpqbi4MGDePTRR/H000/j+PHjAIAxY8bgn3/+wcqVK7F9+3Zcu3YNzz77LP+8wWBAnz59oNPpsGfPHixatAhJSUmYOHEi7yYjIwN9+vTBI488grS0NIwePRpDhw7Fhg0bJAoy4Szu2plw5Eq+0iaoC3fNCIwgVuCxJFLtYXcZtRN+u+l3FmGCjxDHTz75pNnvqVOnYv78+di7dy8aNGiAn376CUuXLsWjjz4KAEhMTERsbCz27t2LLl26YOPGjThx4gQ2b96MqKgotGvXDpMnT8b48eMxadIk+Pn5YcGCBYiJicHMmTMBALGxsdi1axdmz56NhIQEiYKtLlx1lICY97hrb4x53NFRAoRrIW3oPKxGIdUE0iF6DozBYMCyZctQXFyMuLg4pKamQq/XIz4+nnfTokULNGrUCCkpKQCAlJQUtG7dGlFRUbybhIQEFBQU8L04KSkpZn5Uuqn0wxparRYFBQVmf+6Cqw5zpErzDmZnIUkQLzQH5g5yxgRFs3zIkYedW0bNKMwaZgE12WoBwQImPT0dISEh8Pf3x7Bhw/Dnn3+iZcuWyMrKgp+fHyIiIszcR0VFISsrCwCQlZVlJl4q71fes+WmoKAApaWlVu2aNm0awsPD+b+GDRsKDZrb4UzezC/RY3HKRdwq0lrw947PtsSVp3U6KCFSBL+SoQpLln1gWAogYRfZDq20cE3SybMurNsoT1tH0BASADRv3hxpaWnIz8/H77//jkGDBmH79u1y2CaICRMm4N133+V/FxQUkIhxgtHLD2Pb6RtYdfiq0qYogtSNq9DKUw09CSowUdWIzgN2nlNTb2CFrex9CTkThZ72YScnggWMn58fmjVrBgDo2LEjDhw4gDlz5uDFF1+ETqdDXl6eWS9MdnY2oqOjAQDR0dHYv3+/mX+Vq5RM3VRduZSdnY2wsDAEBgZatcvf3x/+/v5CgyOKy7klLs2ErpoDY8q20zcAAIcz88S9W/yrmUNsWKiecj2uGm6VG3cqP4RzUC+3dZzeB8ZoNEKr1aJjx47w9fXFli1b+HunT59GZmYm4uLiAABxcXFIT09HTk4O72bTpk0ICwtDy5YteTemflS6qfRDacr0Bjw4Yxu6fbFNaVPsQqfjSoOYaPx5dwYKtXe2BxD61evO3cbT/j0JQK4DM9mIN1aFlDOxw0bM3kFomZIyRWwJB6nzoJx5mpXyIhZBPTATJkxA79690ahRIxQWFmLp0qVITk7Ghg0bEB4ejiFDhuDdd99FzZo1ERYWhrfffhtxcXHo0qULAKBnz55o2bIlBgwYgBkzZiArKwsfffQRRowYwfeeDBs2DN999x3GjRuHwYMHY+vWrVixYgXWrl0rfehFcLtEp7QJDiPb+LIDHrNZfcuLElWBklNgxOrjjSey7TsSCSuaXYqGQfxZU/L4qwQ2TxKwcZPlMAoVtyyHRWkECZicnBwMHDgQ169fR3h4ONq0aYMNGzbgscceAwDMnj0bXl5e6NevH7RaLRISEjBv3jz+eW9vb6xZswbDhw9HXFwcgoODMWjQIHz22We8m5iYGKxduxZjxozBnDlz0KBBAyxcuJCZJdTlBhXlJhfkfHftwpT6MEdaRk0IQa55KpQLCXdCkID56aefbN4PCAjA3LlzMXfuXKtuGjdujH///demP927d8fhw4eFmOYyDEb1CBglLVVPLFmGs/J/0f7RZ5RLIJ0oH5SFHcOV8eTp+Z3OQhJIuZoEjEBTd5y5IY8hhGAcPxlcfH50tlfI2WESWZZRq6d4EjJiKW+y0tizYgeg/vJCAkYgSvTAuCq/78vIdcidJ8yBEVOw1VYZUK8Q24heRW1vGbVIf5XA5jwXJ551FrXXb+4CCRiBqGsIST22sobUO/G64xwYT9Y/p7LcZ6dvFnCnvORGQWEeEjACUZOAcYWptpplVpeSCkWaFSVs5BtTO5QXVWzEiRheXbjf5n1W876Tg34SWaEUd9JETR93rNQdLEICRiDlRqPSJjiMXPleTYVfLKLiTsJocdQrp/b1kCmDKCnyXfXmmxaO2DC3QznRa+/dLDaIUmtpjrMdDxP+SJf2hTJCZ4dZhwSMQJTYHI7lPKb4R7xMcFZ/iEP53g7p4QBsqrKni9HIIafQduNOKIfacqFcH0umm0xKDYsC0V0hASMQvYr2gXFFT4knlFV3XUYthah645eDZr9vFatno0cWOH+jCF9uOF3tuny9p+4Dg0XKIaQ8F43VoUpXQQJGIGpahaTWAs4CSsedw+8XaKfc4RIimmkZNfD2Ujb3u7KE2uK2qr1u2AnqNCpL0mqQgBGImibxyvXVb+qtu1QKUmxLrra5Qc7mD7U1aCxi62gSWkZtbxm1ciFx5ZCw2uoVV0ICRiDqEjBKW6BmTJZRS1CBuOMcGItQnpMEuRot1nJhZR0lS2+c9F5K8l7W0kDNkIARiKoEjAveYa1hVk8sWUbqCpWVOTCmVniMqFIAVucmOLVqTTIrCCHYnAPDZjZzGSRgBKLEMmq5upLlRk2Fi71uWsfsYc9ux5Enf7IRH2pOF1dSWUdIv4zawlEC0r7CZci7jNq+7yx/tJOAEYiazkJSErVWFpZw/Fwi6/fcsbfDUiNNpUM6xAo8JbfYF4ozQ0h25/ooFE5771VbVbD/omNHzCgBCRiBqGoVEjUnojGthCgWrUARIxssiQxCWSgvWIcEjEBY7k6ripIZXz2xVEHVuJJa/AmdAyPbPiByL6MW4D8JbHUh6+GIMvRKVM1fauv5cJQz2UVKm6AYPkoboDbUJWDkX0btCbAyAdcSDJumCO4VHyKPErATCZ4gHPNL9bhyu1Q2/21pIXvxK1xI2fav6m7Y0vnMPiRgBEJzYBxDbR871XpgxIzJ27jnjnNgCEIqrJU3sYL0x50Zop77Yv0pcS8kFIGGkASiJgGjHkvZg7Pyf9H+uVfXgFU84eveEZxdRs1idqG0VQYW8wIrkIARiFFFAkaJgycr4aC+XhirKBCNUqx8suhewsA465M7V8xKNvb95u/B0St5ir1fCHIto7b4LkZqJFbsANRfBknACERVPTByTQR1eI8S9VDVVnHLOq0/JPgAN1XF3h3UXiGyhNi4zC7Q4qUf9krurxzYW0Ztqxww26sp+SaY0vrnTpCAEYhBgY3sxOKKfM9sJeIkphWnEmKCpa80a7hr2rOAszFbojNIYgdhBfaLp0dAAkYg1APj6Ls5KuMmCF5GrdIeGKWhWJMPtS2jlvUdTsSFUDts1QWevjiABIxADAbXV5Hi3yj/Mmp3+QivJjBUEkahpsm+D4xMbh32k+XEEoj7hMQ+bpRsKks3dVlbFRIwAjEoUNJY1tjqzv7SosQyaiUbbHdqdAg2sZXHFM1+LqyUqZxZhwSMQNS1kZ3SFqgXs2XUDK9CUhJLJrpTD4gzOL+Mmr14ZNAkVSI0Z1C0W4cEjEDUNAfGFcuopd6AihVMGxCWg6Kkbc5+hKo9j9iC5jCxh6s6Tdw5X7MGCRiBUA+M442mp08wcwVUWZrjTtEhS9pynNvkGXcJh5KoPQ5JwAikXIFJvCzjLl+aNubwivbDGdQaq2qvEKVCDcvg1QBlJypTtiABIxAldrcV+0bX7APjgpcojBTzEVic00DIh7PCXrbc4kSvqBwfK678AHJVj7DdMEm4qaWny2QSMAIpV2AjO7GZVC6xNWXNCfyVdtWmG7U312ZLxZUzwwEE7i8jo5C6cLNYkHt36b1TGxTvRCVqzwkkYASipjkwcnHw0m2MWpYGQP0FoBK5K3XBRwnIdBaS+bPSh7lYVy65nyyhK3fdB4wseZKxnkB7Q2228qiSQszTez5YgQSMQPRqmgOjIlNZQ8wyak/7srW8jNrlZriUYb+mOuTO6Tkwbh6PlXDgwHEcTlwvUNoUlyFYBHtIXhADCRiBqKkHxiXzday8Q20NWbVJvGYXFNi8UKWfeILSXYUdDFtP5ThmB6utjjNzYGQK0p+Hr+KnXRnyeC4TtqLCXjydFCjWbG+QKcgrt4MEjEDUtA+M0pZ6eNmqhpAGwPFeH4E2OPEs4Vo8YRm1Bhr8tj9TaTM8FpbyghhIwAhkQJfGSpvAFDa/RFxmBSNIuozaMc/UXgG5K7SMWhqc6elwF2gFo3VIwAjk/piaqum2U3AESXVUDYYcRwkIyTdqbQCZHTpxMc4vo5YpHhmrvFy7jNo175E6RFSirEMCRgRsVQHWUWLPmko4cKqJJ4sovIza4R4Yxqo3QcNksrxf/vhYl35d9nfIBsdajlEpjESis9n9+x3npTFEIUjAuDEu2ciOlZLsbsi0jNpsfxs3TDpXBGn4kkMueIvzYTmVxf7KHmd6Gt0w+7qco1fylTbBKUjAqAHGun3dkapf7qbCTK4Jte6IO4oiMTjaMMtZskf/t1cTy9j7AGI2P9lIOKl7AZmNAwYgASMClx9SKDYH0xwY0ZjvxEtHCagFVqS+40OAVq5LkF0s7jdSpe5affgqjA6urFR7Flbrd6DKo11WSMC4Ma6YAyNnBexKWDs3ylGnztit9Nk2cog6lWU718NxZvG+/ngWHv5qm4IGieNqXimW7lNw+TUjGY0RMxSDBIwIVCrkCQFI3bbK1WsnVATIPWdJbcJVLqRYReaqXrvLuaUueY8orETBu8vTXGqGEKgMuA4SMGpAZONHy6jFI2oOjB13au3CFoKbZgfBMHsatROwNGH/4i1hB4cCKt6awF0rWQkgAePGKLmMWnXYiCpFllF7QNq5fwjto0ST6g7x7spDNZXGHdJLLkjAiMDlX9IiGzPFl1Gr84MHgHp6lpw7jVr6dysuvFSSboRzKC1gWMlmipc3hSEBQziHB5QfT68khEAxVYFahytYw9oHkt6gbE6j1GUDEjAiUEvlRO2u49g6SsBxP6w/5Y4iyFJPpDuGUwwsLKOWGpZs0hmE98Codg4aQ/HOGiRg3BhXNCbWK2AFjzGQ4N1ymC+LnwJrN7mTRY4Tt4VQqC2X3lM3gyUholZcGYUsTZ5mDRIwakDsKiSJzXBnbIkeKSp8ocuoHd4HRsE5MBb9lN5LVSLNMmoJDHGCG4VaZQ1QKUqnmydBAkYMau2KlAEWhwykMUm4J562jNryJF7X28Ei7vDVPOmf42a/lVmNp8BLGYPiwDokYNSA2FVIlPNFI/VRAlX9VAq5TaCl+8KwpmllFUAOen0tj+EN7kSg1u8HKlLWIQHjxjh4xIlTWCtcHNRVYai1jmCtcpPjuAR3xlYcuEMvjrtic8jZhenGWvl3NSRgRKCmhllubFfAyiD1e6WqJIQMIcl1Arb8k3g9vEb9D7WsVLQFC0nJgAkEw5CAUQFiCzEVfvGYxp1cYsIdoTiowOmjBBiMSLWLUynPI5PrbDNLUE+cdUjAiEAtkzFdsoyawY14xYS76iOynJQsy9Jstio3Ycuo2bLdU2CpQaQsYB+KI+uQgHFjlMz47lToFFl9IbE7i8/KIdIYahyVxNEhJFuuWChDJ64VoKBMr7QZzCH3tgsO2+Hh5Y0EjApgoSKzhrsWIDlCpZaeO0exlPYs51VXwmq5EGJV2uU8PP7NTjz61XbZ7LFXJqiXjoZlbUECRgRqmaDnikqUxfpFimMAzJZROxhIe+5YjCupoWXU0iBnLApNoptFFRvayWGTK7OLOmrt6tiKI08vbiRg3Bijgge2svoFKgZlNvCSa+awdKGxuJGdIpawh6MfOa6OAzU24qmXbuPizWKlzXAYlx4z4M6FyAEECZhp06bhvvvuQ2hoKCIjI9G3b1+cPn3azE1ZWRlGjBiBWrVqISQkBP369UN2draZm8zMTPTp0wdBQUGIjIzE2LFjUV5ufoZJcnIyOnToAH9/fzRr1gxJSUniQigDrh4KECsGlBYRrpyp7yyuqAjkiA5n0tjZICftuSi9p26CFGXPlUc9KDVUY69MXMotQb/5e9D9q2SX2KM2PL3HU5CA2b59O0aMGIG9e/di06ZN0Ov16NmzJ4qL76jjMWPG4J9//sHKlSuxfft2XLt2Dc8++yx/32AwoE+fPtDpdNizZw8WLVqEpKQkTJw4kXeTkZGBPn364JFHHkFaWhpGjx6NoUOHYsOGDRIEmXB3pCjTSneXy1Utyb4PDCkYQahH4su1is72/dNZhdK9TE2RbQaVKWv4CHG8fv16s99JSUmIjIxEamoqHnroIeTn5+Onn37C0qVL8eijjwIAEhMTERsbi71796JLly7YuHEjTpw4gc2bNyMqKgrt2rXD5MmTMX78eEyaNAl+fn5YsGABYmJiMHPmTABAbGwsdu3ahdmzZyMhIUGioLs/rhDn7jrJzixcjo7mqCAq5P5iU3LY0q2QMZ0s+ayGvKsWpK4Tbc6BkfRN6sOpOTD5+fkAgJo1awIAUlNTodfrER8fz7tp0aIFGjVqhJSUFABASkoKWrdujaioKN5NQkICCgoKcPz4cd6NqR+Vbir9sIRWq0VBQYHZn1y4WsiLLQ+euoxaTC+A7UpCgcDIJJrKTc6XUPw0ajeufSU5jVoCO6qi2k4IxmAm6zJjiDKIFjBGoxGjR49G165dce+99wIAsrKy4Ofnh4iICDO3UVFRyMrK4t2YipfK+5X3bLkpKChAaanlA8amTZuG8PBw/q9hw4Zig0YIwMPLjyAEHSXgYMwKjX+jzAdkuWuPnFBYHUqzOgfGpVY4DmUn22lDc2BEMmLECBw7dgzLli2T0h7RTJgwAfn5+fzf5cuXlTZJcTx1GbUUmC+jlt5PpTDIPgeGkAI245FNqxzFVb1PLl2F5MJ3sYigOTCVjBw5EmvWrMGOHTvQoEED/np0dDR0Oh3y8vLMemGys7MRHR3Nu9m/f7+Zf5WrlEzdVF25lJ2djbCwMAQGBlq0yd/fH/7+/mKCIxhXr64R2/C54jRqFpFaKCizjFpad5Ww1APDai+FFLC6V5QGltOIes7YhZVdf1lEUA8Mx3EYOXIk/vzzT2zduhUxMTFm9zt27AhfX19s2bKFv3b69GlkZmYiLi4OABAXF4f09HTk5OTwbjZt2oSwsDC0bNmSd2PqR6WbSj88DbF6ySVnIVlphDioe+dZORpXeU6jFmZnucksWznC6OkVaiXSLKOWIX0k99EzYSWfe/oQkqAemBEjRmDp0qX466+/EBoays9ZCQ8PR2BgIMLDwzFkyBC8++67qFmzJsLCwvD2228jLi4OXbp0AQD07NkTLVu2xIABAzBjxgxkZWXho48+wogRI/gelGHDhuG7777DuHHjMHjwYGzduhUrVqzA2rVrJQ6+OFTcLrsUNZUtpQ++dBVy98oxEERCBI6kmxL5V0qRraZ9qUyhMmUdQT0w8+fPR35+Prp37466devyf8uXL+fdzJ49G0888QT69euHhx56CNHR0Vi1ahV/39vbG2vWrIG3tzfi4uLw6quvYuDAgfjss894NzExMVi7di02bdqEtm3bYubMmVi4cCEtoRaIKzI+C42yHIiZA6OGuJB/CElW7z0GOeORksh5bGkh1x7m6NkI6oFx5Cs1ICAAc+fOxdy5c626ady4Mf7991+b/nTv3h2HDx8WYp7bwvIcGOs7e6p8CMlsGxjXVxMOr0JibBm1kC5tdxY7rC6jtvouN04LtWMzbTw84egsJDGopWH20MwtJtiuiCl55sAIQ+4xc8/McdXZfzFXaRMIGWGll8VTF2pUQgLGjaEhJPGY9oC41TJqhlYhEewIPkd6/FixVSyu++50XUx5enkjASMCl+/EK/Y5D83bUh+k56hvUg41OeyTwESWW8AIwVPzpynW4kDOhom1eLc13Maarazh6dFDAsaNcc3cDbXt7WkZ1k6jlmsIyVTAyBFkanAkxJXDFJRuzEL7wFiHBAzhFJ5egITAQlwZ5D7MkYVAqgg1beanzDJq6XDVogJW5sd4AiRgROD6nXjFZVNPbUvETeLlqvx20kOnkeedl26V4I/UKyg3yHNstKfmOamRMxrVJJoI23j6HBhRRwkQ6sAly6htvEMti7UsImYOjITxLdfeM7nFOry38ggKy/TCjXLEHkFuPbvyBUjwqTX8lHfZgHpg3Bgl1TnHUfemKxCbxikXbsnSeHj6F6GUuLKRdCTZrueXym+IGyB1qtlKG08fsiUBIwI1b9AmNSx+iYixqGo9IMcyaiHI/Uq5wuTh9anbkn4lH6OWpUnuryvrUlYP2LSHrTrW08sbCRiCsID5UQIO7oorky1qgkVByzK2drJ25UvtpdtfaVflMcVeOD29hbaDp8cOCRgRqEXHu6J70dYrlKp7WB3GELaMWl7RxDn1tHWEzLtiNJmYgeKHANisY1mBBIwKEJtJXZG5FfmClIHyKi2v0nukyD1xmIaQ2ICV+GLFDndA6ri0KWA8vA+GBIwb49lZWxi5xTqz36Y9IBSPjuPpFaoaYCmF7PVKqnEfGFfi6cKTllGLwNX7wIjFFUMp1l6x/OBl2d9tDVeuAjB3J+FRAjLtxCvFkzZ9laGXyRNxddwolRae3gATzkE9MCpA7Fct1Q3iUTruHB9CckLCyLGMWnov3RwbK0xcaAUhHVL3QtpeheTZuYQEjAhc3f/C8vI/FocMJB+DltY7JpBvDow7xpZ8sBJdSqWbSjqzq+HS4wJs7gPjOjtYhASMChAtElyRud20AIlZRq0m5AqRG0aVIsiV5ziOrTSyZwtLtrKIp8cPCRg35sLNYqVN8Cjs1SVyLKN2BnlWWjnuq6dXvvagQwHZxVZZdmW6rTt23XUvYxASMCJQa7enHDBZ8UlglNLLqOX2U7YvfFl8dV9cHV9Ud0kDK+LyVFahy+xgERIwKoC+VAmpkStLefqYPOtUnFFWPZFYrWOkFNpqWT1aDUbThgVIwBCi4TiOyfkhkkwsNt0HRoEgMhitDsFifmAZii/3g5LUdZCAEYVKlTzhMGLqICkrLkdFmOgl9hwLjafS71ceWzHg0hV+KkgK5fMrwRokYAjRVHRHuz8sLhWvRPQcGGnNuOMvu1FFwLPnwKg16CzXP0pDAkYErq4EWM6+LDZYUthkvozaef+ceb+acMUBooR4rC2jVkMjyVLWcqUpLIWbNUjAEKJx53IlrkJXz1ECLKxCoorZehxQ3FSgxmhQgxh0F0jAEE7hCUWV5TA609CpZam4pyJHXFrrPVZDurFkois74VkKN2uQgBGBq8dSafKaMNwhttT6FadWuz0Ftc1bk3JHbLXO/6H63zokYAjRsLqMWgqUngMjN/KdhSSPv56Gq4UgJZswtOVGq/eoDLgOEjAiUKuSlwMWy6oUospMwDi6pFnKZdQOz4ER/1J5dgYWcJSA9K9XHbbii+LHHIoPoiokYAjRUIVijpTx4ahf4pdRy3dYIOE4rESXu/akmqLWD0/3TxnxkIBRAUxnYKaNE4+bBotHtiEkebwlJMSSWFFDunmAxrKIp4bbEUjAiECj2i2RpIXVgiWFWaaVvMPDOUp0wTCGkDjwhK9+uzCyjJrVpGDULIIRSMAQTkGrTu4gZVw4Pu+GraMEaCM7CfGAuBS0b9B/rkn4EpWQgBGBy8dSGS2vHDgm61gWbWIN2ebAyOIrISVq3Ym3koE/7xf1nFp7ztWTMq6HBAxBWEDK/ScA4XsHqVWE0dexMKzFlstjUaFkE1IuKrPWzrM3ZbFFKiQvAlSmrEIChhCNtXNV3A0p5sAIjSbZVyHJlG5ChpA8IOvYhRXBp5QV9t7LSvwQbEICRgTq7Ij0HKToDhfjhyKHPjrxnBzmGqm9kQxXRqWiOsFFFSoto3Y/fJQ2gFA3aho7VxNq/fI0koIRhJyxZcnvbadzsOPsDQtu2U83tRQJqeNSLeFWAhIwKoDyr0AkiDDznXjdEAaGkAjryBWNx68VyOMxQSgADSGJQKPWvkgZcNf2yjRYrtjW39b7bbpzZideGdJOSAeMu+YdqXA2fsRMkGUZNfQSEa6FBAwhGrWdbCs3ajsLSQ6oB0YYrEQXI2bIiqs+O6VOU9bKOEvQEJIKUOt8CKVgMbZY67Oj06jZwFrjVKIrx9nsQtfZoYKEU4GJsuCp4XYEEjCEaFjdyE4KzI4SkGRVkzzuWYt/WkYtDd9uPae0CUxQdS5aTmGZYrYQ7EFDSARhAVFzYCQ9C8nBowTEeu/Es7agISR1ooZkyy3S4f6pW8R74KK5i7SPnesgASMCV8/hZTX/VhQs9qyjAm8fuYYMaBU1IRdHruQpbQLBGCRgVADLDTLLtjmFxMuoWZsDIxdqmEvBEqxEFyt2ENWhpLEOCRgRuL4Hhs0szKZVMsxZcXg4x7o7uebAiG15OMgjNgwCumBI7BD28gCrdR/BBiRgCKeg6uUOyiyjltd/odAQkjBYiS41CAVnLXTdMmqpd+JlP22UggSMCmA1/3pKwfKMUEoDTeIVCCPRpYZk85T6hnAcEjAi0Lh4RgOr5ZYDsPVUjtJmVEOK+DJbRi1zb4hlvxwcthK9E688sJpXCdtQshFqhAQMIZqU87eUNkE2lK7QZd+Jl5Nn0EDIHBiCsIcaBTEto3YdJGBUAKtdp1n5bG4qpVRssZpOroSGkITBytwTNeRdp8+G8pSlgB4ECRgR0D4wFaih0hOL+Q6gzodTaJaRO2ppCIkwhZKNUCMkYFQAq40Co2ZJgqlokWIODGtHCci3CknIMmp5bFATrMSBUnbYe6/ZjtgqqXHoMEfXQQJGBK7uiaQMLAyleobKDa5/rzNvlCOaaAiJEIIr6za1jiBRkbIOCRgVwOq8SHcuWFUPkXOEIYsOSPh+eSNXroaD1bzKKuxEFzuWWMOd6xtCHCRgVAAVXGFIs4xauH+FZeVW76n1608o7jwvyp1RKtlcvSWFa5B4IztJfXMvfJQ2QI1oXDyLl9VGgU2r2IS1SbMcJ08vjJAeGBoaZQelUiLjVjF02Uar98V8SLgbrNb/LEA9MCqA5hUoixINrdz7wLAwiZegxklXbl28VOW9lUeceperPzwJ+REsYHbs2IEnn3wS9erVg0ajwerVq83ucxyHiRMnom7duggMDER8fDzOnj1r5iY3Nxf9+/dHWFgYIiIiMGTIEBQVFZm5OXr0KB588EEEBASgYcOGmDFjhvDQyYTrJ/GyiadXvkLwlKqTNrJTJ1SUCTUiWMAUFxejbdu2mDt3rsX7M2bMwDfffIMFCxZg3759CA4ORkJCAsrK7mx61r9/fxw/fhybNm3CmjVrsGPHDrz55pv8/YKCAvTs2RONGzdGamoqvvzyS0yaNAk//PCDiCCqH6pcXA8nZhavLf8Eu5f3/AIWhrQoX7MDq8N5rNpVFb3BiBnrT2HP+ZsyLKMmrCF4Dkzv3r3Ru3dvi/c4jsPXX3+Njz76CE8//TQA4JdffkFUVBRWr16Nl156CSdPnsT69etx4MABdOrUCQDw7bff4vHHH8dXX32FevXqYcmSJdDpdPj555/h5+eHVq1aIS0tDbNmzTITOp4Cdcu7Hum3Axfmo/ynUXO0jJoBWIktVpNNSrvk7AVdsvcS5iWfx7zk8/h9WJy0njOaNiwg6RyYjIwMZGVlIT4+nr8WHh6Ozp07IyUlBQCQkpKCiIgIXrwAQHx8PLy8vLBv3z7ezUMPPQQ/Pz/eTUJCAk6fPo3bt29LabI4aAwJgGdUegCz0c8kJGCEwUp0sWKHWsm4Way0CR6JpKuQsrKyAABRUVFm16Oiovh7WVlZiIyMNDfCxwc1a9Y0cxMTE1PNj8p7NWrUqPZurVYLrVbL/y4oKHAyNOzAaqOglu5dwnXQFBh1wmpZVss8O87K/wl5cZtVSNOmTUN4eDj/17BhQ6VNkgxqFIQhRWVsvnxTAv9kci/Wtopl1NIjxB6VtE2Egqghj8g1HMv7L5/XqkdSARMdHQ0AyM7ONruenZ3N34uOjkZOTo7Z/fLycuTm5pq5seSH6TuqMmHCBOTn5/N/ly9fdj5AVqCjBCpQQ+UiFjFnIdn0T6AfjgoB1tKAViGpE9byUSVSmiXXKuoDF2/LWkerpRdKCSQVMDExMYiOjsaWLVv4awUFBdi3bx/i4iomNsXFxSEvLw+pqam8m61bt8JoNKJz5868mx07dkCv1/NuNm3ahObNm1scPgIAf39/hIWFmf25C5R/CamRq8I9lJkni78EwSovfJ9i1ktO9bXrECxgioqKkJaWhrS0NAAVE3fT0tKQmZkJjUaD0aNHY8qUKfj777+Rnp6OgQMHol69eujbty8AIDY2Fr169cIbb7yB/fv3Y/fu3Rg5ciReeukl1KtXDwDwyiuvwM/PD0OGDMHx48exfPlyzJkzB++++65kAXcG1+/E69LXqR7JjxJw3jvh75fbf8pThAms5gdW7aqKWux0NwRP4j148CAeeeQR/nelqBg0aBCSkpIwbtw4FBcX480330ReXh66deuG9evXIyAggH9myZIlGDlyJHr06AEvLy/069cP33zzDX8/PDwcGzduxIgRI9CxY0fUrl0bEydO9Mgl1ADDQ0hKGyAjZpPyJAqooHSUeRm18w87D6v5mmAHKfOIvOcuyTiEJJvP6kewgOnevbvNMTmNRoPPPvsMn332mVU3NWvWxNKlS22+p02bNti5c6dQ89wSVqcVsPrVwahZsiD6LCRpzSBUDqtiktU6piqyTuJVSRwogdusQnIlrp7ES8uoXY8cUS7kC9B9Y5ZgEUarGNUg9apFwjFIwKgAKg9KYLIKSZHDHB1chST6MEflMxUDJhD/wWpSsGpXVVj9yHR3SMCIgA41rYDVMit148xqOAHxtuWXlkNncPwkYIJQBJWcJSBnFeHOPd3OQgKGICygtGiR+/U3i7T2HREeAws9cpZg06rqyLlqkdGkYQISMITbIUV5pzqDUDtC8jCr+V0tjTf1kigDCRgRyLscj2ANqb5OhVRyaqm4ncEDgqgaWM1v0i6jlhFahaQIJGAIwgJKd6k7WnErbSfBLu7wmaWW7G06iVctNrsDJGAI0bDaeDJqFvXcEQzDaKFRCRR7ykACRgS0CqkCVoWCFHBW/u+cn9JHmBsnAeFCWC3LjJpVDVbjz90hAUMQdlCicnL0naquONVsu5thLymU+miTdBW1i5ZRS/2hwmpPNwuQgCFEw26xct4yowznNwjaiZfdyCU8EKXyo5RiQN7t/uksJCUgAUMQFqBKg/AkmBXMrNpVBWbjz80hAUO4HVJUJuYbU7l+GbWSfroKNdvubtjrQVAqpdSSQ+TMyySOrEMCRgQamsULwL0LltLLIh0+C8mN04BwHfaykWJDSBK+WNbt/uWY9c97R4XcGiRgCMICchzORsuoCUJ55JivQh8SykACRgTUDFXA6peBFFaZzuFdsi9TAh8F7sQryRsJwjFYbYBZtasqrNaF7g4JGEI0aqlcxKD00kWHl1HLa4asuHP+URv2GuCl+y+5yBJz5MgicuQ70w8eOszRdZCAIUTDarmSosDLsIpaliEkqtwIV1CmNyryXknzt0rPK6Iibh0SMCKgObzujxxzYIRAXdKES6Hs5iQUgUpAAoYQjxt//ssRNHlEifumAeE6WM1Fajl+w2zbBalXIbGaOAxAAkYE1APDNu7Qe+EJRwmo2HS3g9V8JKVdsu7VIpvP7KPkfEESMIRoPLnQikHQUQIy2kEQnow8y6hl3WVGRr/VDQkYwu1g9muSKiKCUVjNm0qvBnQU81VI6rDZHSABIwLakKwCldQtqsQjhpDUbLybwWpSMGpWNVy2yy+DKGkfCRhCNPSlIQxhwtfBowQoDQgJYDUXSToH5j+/5JnEy2oMujckYAi3g9W6hMQGQRBCoVrDOiRgRECrkCpgVSh4EpQGhBSw2oMgyzJqObZIMPFTjk0wCcuQgCFEQwVVPjziKAGlDSB4WE0LtdQxphtfSr0JJqvishIlrSMBIwLqgKmA1SERVu0SgvpDQKgKD8hwrppoa1SL6nIDSMAQomH8w4A5KL4IQhhq2RHbvAdGWr+p2rAOCRhCNKx2bTJqliAcjVs1h1XNtrsb7PZasmqXdQwSKxjWywntxKs2aBYvAPWMT6sRR6OW3YaHUBOsNpLSLqPmJPeT99vCewj5IQFDiEbpE5sJgpAGKsrOYSpaDB42iVdJSMCIgPpfKqByJR8Ut4QrYTW7qaUcmPZGSz2ERFiHBAwhGvoykA+H41bFSaBi090OVsuyWoZITUWL1FHJegzQMmpClbD6ocFoXSwLHhRUwgORdA6MdF5V99vE0HJWK0Y3hASMCGgObwVq+TpiBSFfuRSzhCthNb/Jc26R9H6aDyEZpfWc1cRhABIwhGjoQ0N5WO36J9QFq9mIVbuqYjqEZCD94jJIwBCiYbVycYueIfefAsNuBnITirTlAlyzmRZSlmU5s5vpikzJe2AYR8liTAJGBDSCVAF9/QtDSGxRzBLOklOoVdoEJpHjA8e8B4ZKr6sgAUOIhtV9YBg1C3oDo4YRHg+rZUYtSt4o4yRe+lC0DgkYEWhoFi8Ahis9RhHStewRRwkobQDBw2paqGUSr6mfkh8lIKlv0qPkkD0JGBGQIq6AekqFUS6gB8bxowQIwnlYrdKkrGvlDKKBllErAgkYEVD2rIDVybJsWkUVG0EIRS0lhubAKAMJGBEYKYMCYPerjVXKPWx1AqEeWP0YkQM5QmqUUcBQPWsdEjAiIP1SAQ2lCUPIJF6HTxJQcRqo2HS3g9W0kNQuGQNp4GQUMIyLS1pGrTJYXX3jalgVcqw26kIqNkcrLTZDSqgNVvORPJN4pffVtGjTULHrIAEjAhrjrIBFIVesLWc2ffQCtugs0zvmdu3R62LNUZRygxElOiEbrRFywqrol8Ou9Kv5kvt5w2TPHU/byE5JfJQ2QI0wWtZdDovx0OqTDUqbYJWdZ28qbQIzdP8qGVdulyptBvEfv6deUdoEi0hZxRy5ko8F289j+rpTEvpaHamPEiA9ZB3qgREBiz0PSqAtp5JFiIPEC1s4K67rhgdIZIk5Wgd7Ih1FbvECSN8Do5NaEbkRJGBEYCABAwAo1dMQAEFIwadPtVLaBKeY3q+NLP6W6Q2y+CsnNAfGdZCAEQHplwpKdOqrXAiCRSKCfNGoZpDSZojGS6bNydU4T4rVOXjuCAkYEVAGraCUBAxBSILajyfxksl+NX4keVoPDC2jVhk0B6aCUhV27xIEIT1y6S811jG00anrIAEjAtIvFajx64ggWETd/S/UA2PKDlpt6DJIwIiAhpAqMN37gCAIz0UuAaPGOuZmkfpsdoYDF3MVezcJGBF0bFLDrpuX728oW7cqIQ2vd22itAmEiqkXHoDaIX6S+FUjyA+9742WxC8lqBXih4Y1Ax1yG+rvAz9vanrcBSUFm4ZjdQtGJykoKEB4eDjy8/MRFhYmqd+lOgM2nczGzUItOAB+Pl64UVAGby8vFOvK0bROMF7o1BAajQYlunKcyylCeKAvfLy9cCarEGV6A2qH+qNOiD9KdAacyirAlduliGtaC/klevj6eOFmoRYxdYJRqjMg7XIe7qodjGv5ZfD38YKPlwZnc4pQrC1HobYcr9zfCI1qBuHcjSKE+PugUc0gZOaW4FaRFrVD/BEVFoDj1wrAcRzKjRy8NBq0rh+OYH9vnM4uhK+3F/T/7enSLCoEBaV6XLpVghB/H+gMRtQK9kd2YRka1giCt5cG1/NLERnqD44DQgJ8EBHoB125EV5ewMnrhWhSOwhhAb64dKsE0WEBKCjTo0hbDn8fLwT4esPHW4PDmXko0ZWD44Do8ADcHRmKy7dLUFRWjvBAX5y/UYT8Uj0CfL1RLyIQPl4atIgOrYiLOiE4nHkbxdpyRIYFoH5EILLyy5BTqMXtEh1KdOVoXCsYLaJDcbtEj7i7aiGvRIewQF8E+Hrjal4pbhfr+PuHM28jt1iHgjI94u6qjYggX/h4a1BUVo7cYh0igvzg662BkQMa1gzE/oxchAb4Iq9Eh2aRIci8VQJtuRG+3l4I9PNCYVk5yg0cinXlaBEdBr3BiLTLeagT6o+64QE4nVWIK7dLUao3oFmdEDSsGYT6EYEI8PPCmawiZBeUIeNmMRJaRSO/VI/GtYJwNa/C/a0iHXy9NQgN8EGAjzcu3ipBgK8X8kr08PXWIMDXG5m5JWgWGYJyA4cgP2/UCvHHtbxSRAT54tjVfNSvEQh9OYfcEh0evqcOagb74Y9DV9CuYQR2n7uJ0ABfFJTq4efjhbrhAagT6o8mtYJxo1CLAF9vZBeU4fLtUlzLK4W23IA29SMQ6OeNUr0B/j5e6NCoBvJL9cgv1cPX2wvRYQG4lFsMIwekX8lDo1rBKCjVo0xvQGzdMJQbOdws1KJJ7SBEhgbg5PUC3CjS4nxOMe6JCkF+qR7lRg4xtYMR5OeN+5rUhI+3hm8ES/UGBPn5wGjkUKQrR1iAL4CKntL8Uj2C/LxxNrsIa9Ovo3GtIBRry5FdUAaDEXjonto4nVWIAF9vABWrgdo1jEBBaTmu5pXi4q1iRIcFIDyw4nrqpdvw9tLAz8cLzSJDEBrgg683n8WZ7EI0qBGINg0icKNQi7PZhWgWGYLsAi2a/Gf3vgu3AAAdG9dAVkEZvDQa5JfqcV+Tmni8dV2UG41YceAyUi/dxtmcItSPCETbhhHQaIBDl/KQ0CoK2QVl8PH2QlZ+GYL8vGHgOJTpDIhrWhvNo0Pxyd/HAQChAT5oWjsYRg44ejUftYL9cNd/v8/kFKJ9wwhculWCa3mlaFgzCJFh/mhaJwR1wwOgLTfCaORw6VYJUi/dht5oRJ1Qf5zNLkKjmkFIu5yHuyND0DQyBGezi9CnTTQebRGFnIIyJO25iAea1kZIgA9qBfvh+LUC3C7RIdjfB0G+3riWX4quzWqjQY1AHL9WgONX81GiM6B+jUAE+noj0Ncbwf4+uJ5fimKtAUXacpzNKURusQ4vdGqI+hGBKNEZ4KXRIDo8AMXacvx77DoMBg6Nawcj9WIuYmoHQ2cwQlduRO/WdZF+JR9ncwrh4+UFDkBRWTlyCstgMHJ4vlMDtG0QgRPXC1A3PABhgb7QQIMaQb44eb0QZeUVeTTYzxsGIwcDx+FWkQ5X80oRGuCDYD8fRIb543JuCaLDA1GiK0eArzfOZheieXTFcwVl5fDz9oJGAwT4ekNvMKJMb4C3lwaBvt4oKC1HkL83LwTCA33hpdFAo6mYrnDpVglqBvshIsgXZ7OLcC2vFO0bReCfI9dwLb8MzeqEwMBx8P4vTry9NAj080a5gcOtIi1ul+ihLTcgwNe7op4r1YNDhQD39tKgdog/OjSuge1nbsDHS4Pk0zfQvlEEdOVG5JfqodEAdcMDwXEcTmUV4r4mNaEzGGEwcrh6uxSDu8WgZrA0Qr4SR9tvEjAEQRAEQTCDo+039eMRBEEQBKE6mBYwc+fORZMmTRAQEIDOnTtj//79SptEEARBEAQDMCtgli9fjnfffReffPIJDh06hLZt2yIhIQE5OTlKm0YQBEEQhMIwOwemc+fOuO+++/Ddd98BAIxGIxo2bIi3334bH3zwgd3nZZsDU3wT0JdI5x9BEARBqJWgWoBfsKReOtp++0j6VonQ6XRITU3FhAkT+GteXl6Ij49HSkqKxWe0Wi202jvLuQoKCuQxbt144Njv8vhNEARBEGqi309A6+cUeTWTAubmzZswGAyIiooyux4VFYVTpywfhz5t2jR8+umn8hvn7Qv4yHN0PEEQBEGoCi9vxV7NpIARw4QJE/Duu+/yvwsKCtCwYUPpX/TMgoo/giAIgiAUg0kBU7t2bXh7eyM7O9vsenZ2NqKjLe9W6e/vD39/f1eYRxAEQRCEwjC5CsnPzw8dO3bEli1b+GtGoxFbtmxBXFycgpYRBEEQBMECTPbAAMC7776LQYMGoVOnTrj//vvx9ddfo7i4GK+//rrSphEEQRAEoTDMCpgXX3wRN27cwMSJE5GVlYV27dph/fr11Sb2EgRBEATheTC7D4yz0FlIBEEQBKE+6CwkgiAIgiDcFhIwBEEQBEGoDhIwBEEQBEGoDhIwBEEQBEGoDhIwBEEQBEGoDhIwBEEQBEGoDhIwBEEQBEGoDhIwBEEQBEGoDhIwBEEQBEGoDmaPEnCWyg2GCwoKFLaEIAiCIAhHqWy37R0U4LYCprCwEADQsGFDhS0hCIIgCEIohYWFCA8Pt3rfbc9CMhqNuHbtGkJDQ6HRaCTzt6CgAA0bNsTly5c95owlTwszhde98bTwAp4XZgqvuuE4DoWFhahXrx68vKzPdHHbHhgvLy80aNBANv/DwsLcIqMIwdPCTOF1bzwtvIDnhZnCq15s9bxUQpN4CYIgCIJQHSRgCIIgCIJQHSRgBOLv749PPvkE/v7+SpviMjwtzBRe98bTwgt4XpgpvJ6B207iJQiCIAjCfaEeGIIgCIIgVAcJGIIgCIIgVAcJGIIgCIIgVAcJGIIgCIIgVAcJGIHMnTsXTZo0QUBAADp37oz9+/crbZIopk2bhvvuuw+hoaGIjIxE3759cfr0aTM3ZWVlGDFiBGrVqoWQkBD069cP2dnZZm4yMzPRp08fBAUFITIyEmPHjkV5ebkrgyKY6dOnQ6PRYPTo0fw1dwzr1atX8eqrr6JWrVoIDAxE69atcfDgQf4+x3GYOHEi6tati8DAQMTHx+Ps2bNmfuTm5qJ///4ICwtDREQEhgwZgqKiIlcHxS4GgwEff/wxYmJiEBgYiKZNm2Ly5MlmZ6moPbw7duzAk08+iXr16kGj0WD16tVm96UK39GjR/Hggw8iICAADRs2xIwZM+QOmkVshVev12P8+PFo3bo1goODUa9ePQwcOBDXrl0z88NdwluVYcOGQaPR4Ouvvza7rqbwSgJHOMyyZcs4Pz8/7ueff+aOHz/OvfHGG1xERASXnZ2ttGmCSUhI4BITE7ljx45xaWlp3OOPP841atSIKyoq4t0MGzaMa9iwIbdlyxbu4MGDXJcuXbgHHniAv19eXs7de++9XHx8PHf48GHu33//5WrXrs1NmDBBiSA5xP79+7kmTZpwbdq04UaNGsVfd7ew5ubmco0bN+Zee+01bt++fdyFCxe4DRs2cOfOnePdTJ8+nQsPD+dWr17NHTlyhHvqqae4mJgYrrS0lHfTq1cvrm3bttzevXu5nTt3cs2aNeNefvllJYJkk6lTp3K1atXi1qxZw2VkZHArV67kQkJCuDlz5vBu1B7ef//9l/vwww+5VatWcQC4P//80+y+FOHLz8/noqKiuP79+3PHjh3jfvvtNy4wMJD7/vvvXRVMHlvhzcvL4+Lj47nly5dzp06d4lJSUrj777+f69ixo5kf7hJeU1atWsW1bduWq1evHjd79myze2oKrxSQgBHA/fffz40YMYL/bTAYuHr16nHTpk1T0CppyMnJ4QBw27dv5ziuooLw9fXlVq5cybs5efIkB4BLSUnhOK6iwHl5eXFZWVm8m/nz53NhYWGcVqt1bQAcoLCwkLv77ru5TZs2cQ8//DAvYNwxrOPHj+e6detm9b7RaOSio6O5L7/8kr+Wl5fH+fv7c7/99hvHcRx34sQJDgB34MAB3s26des4jUbDXb16VT7jRdCnTx9u8ODBZteeffZZrn///hzHuV94qzZwUoVv3rx5XI0aNczy9Pjx47nmzZvLHCLb2GrQK9m/fz8HgLt06RLHce4Z3itXrnD169fnjh07xjVu3NhMwKg5vGKhISQH0el0SE1NRXx8PH/Ny8sL8fHxSElJUdAyacjPzwcA1KxZEwCQmpoKvV5vFt4WLVqgUaNGfHhTUlLQunVrREVF8W4SEhJQUFCA48ePu9B6xxgxYgT69OljFibAPcP6999/o1OnTnj++ecRGRmJ9u3b48cff+TvZ2RkICsryyzM4eHh6Ny5s1mYIyIi0KlTJ95NfHw8vLy8sG/fPtcFxgEeeOABbNmyBWfOnAEAHDlyBLt27ULv3r0BuF94qyJV+FJSUvDQQw/Bz8+Pd5OQkIDTp0/j9u3bLgqNOPLz86HRaBAREQHA/cJrNBoxYMAAjB07Fq1atap2393C6wgkYBzk5s2bMBgMZg0YAERFRSErK0shq6TBaDRi9OjR6Nq1K+69914AQFZWFvz8/PjKoBLT8GZlZVmMj8p7LLFs2TIcOnQI06ZNq3bP3cIKABcuXMD8+fNx9913Y8OGDRg+fDjeeecdLFq0CMAdm23l56ysLERGRprd9/HxQc2aNZkL8wcffICXXnoJLVq0gK+vL9q3b4/Ro0ejf//+ANwvvFWRKnxqy+eVlJWVYfz48Xj55Zf5wwzdLbxffPEFfHx88M4771i8727hdQS3PY2acJwRI0bg2LFj2LVrl9KmyMLly5cxatQobNq0CQEBAUqb4xKMRiM6deqEzz//HADQvn17HDt2DAsWLMCgQYMUtk56VqxYgSVLlmDp0qVo1aoV0tLSMHr0aNSrV88tw0vcQa/X44UXXgDHcZg/f77S5shCamoq5syZg0OHDkGj0ShtDjNQD4yD1K5dG97e3tVWpmRnZyM6Olohq5xn5MiRWLNmDbZt24YGDRrw16Ojo6HT6ZCXl2fm3jS80dHRFuOj8h4rpKamIicnBx06dICPjw98fHywfft2fPPNN/Dx8UFUVJTbhLWSunXromXLlmbXYmNjkZmZCeCOzbbyc3R0NHJycszul5eXIzc3l7kwjx07lu+Fad26NQYMGIAxY8bwPW7uFt6qSBU+teXzSvFy6dIlbNq0ie99AdwrvDt37kROTg4aNWrE12GXLl3Ce++9hyZNmgBwr/A6CgkYB/Hz80PHjh2xZcsW/prRaMSWLVsQFxenoGXi4DgOI0eOxJ9//omtW7ciJibG7H7Hjh3h6+trFt7Tp08jMzOTD29cXBzS09PNCk1lJVK18VSSHj16ID09HWlpafxfp06d0L9/f/7/7hLWSrp27VptWfyZM2fQuHFjAEBMTAyio6PNwlxQUIB9+/aZhTkvLw+pqam8m61bt8JoNKJz584uCIXjlJSUwMvLvDrz9vaG0WgE4H7hrYpU4YuLi8OOHTug1+t5N5s2bULz5s1Ro0YNF4XGMSrFy9mzZ7F582bUqlXL7L47hXfAgAE4evSoWR1Wr149jB07Fhs2bADgXuF1GKVnEauJZcuWcf7+/lxSUhJ34sQJ7s033+QiIiLMVqaoheHDh3Ph4eFccnIyd/36df6vpKSEdzNs2DCuUaNG3NatW7mDBw9ycXFxXFxcHH+/cmlxz549ubS0NG79+vVcnTp1mF1abIrpKiSOc7+w7t+/n/Px8eGmTp3KnT17lluyZAkXFBTE/frrr7yb6dOncxEREdxff/3FHT16lHv66actLrtt3749t2/fPm7Xrl3c3XffzcyyYlMGDRrE1a9fn19GvWrVKq527drcuHHjeDdqD29hYSF3+PBh7vDhwxwAbtasWdzhw4f5VTdShC8vL4+LioriBgwYwB07doxbtmwZFxQUpMgyW1vh1el03FNPPcU1aNCAS0tLM6vDTFfYuEt4LVF1FRLHqSu8UkACRiDffvst16hRI87Pz4+7//77ub179yptkigAWPxLTEzk3ZSWlnJvvfUWV6NGDS4oKIh75plnuOvXr5v5c/HiRa53795cYGAgV7t2be69997j9Hq9i0MjnKoCxh3D+s8//3D33nsv5+/vz7Vo0YL74YcfzO4bjUbu448/5qKiojh/f3+uR48e3OnTp83c3Lp1i3v55Ze5kJAQLiwsjHv99de5wsJCVwbDIQoKCrhRo0ZxjRo14gICAri77rqL+/DDD80aM7WHd9u2bRbL7KBBgziOky58R44c4bp168b5+/tz9evX56ZPn+6qIJphK7wZGRlW67Bt27bxfrhLeC1hScCoKbxSoOE4k60qCYIgCIIgVADNgSEIgiAIQnWQgCEIgiAIQnWQgCEIgiAIQnWQgCEIgiAIQnWQgCEIgiAIQnWQgCEIgiAIQnWQgCEIgiAIQnWQgCEIgmk0Gg1Wr16ttBmYNGkS2rVrp7QZBEH8BwkYgvBwbty4geHDh6NRo0bw9/dHdHQ0EhISsHv3bqVNk4SLFy9Co9EgLS1NaVMIgpAQH6UNIAhCWfr16wedTodFixbhrrvuQnZ2NrZs2YJbt24pbRpBEIRVqAeGIDyYvLw87Ny5E1988QUeeeQRNG7cGPfffz8mTJiAp556inc3a9YstG7dGsHBwWjYsCHeeustFBUV8feTkpIQERGBNWvWoHnz5ggKCsJzzz2HkpISLFq0CE2aNEGNGjXwzjvvwGAw8M81adIEkydPxssvv4zg4GDUr18fc+fOtWnz5cuX8cILLyAiIgI1a9bE008/jYsXLzoc5uTkZGg0GmzZsgWdOnVCUFAQHnjggWqnd0+fPh1RUVEIDQ3FkCFDUFZWVs2vhQsXIjY2FgEBAWjRogXmzZvH3xs8eDDatGkDrVYLANDpdGjfvj0GDhzosK0EQViHBAxBeDAhISEICQnB6tWr+YbWEl5eXvjmm29w/PhxLFq0CFu3bsW4cePM3JSUlOCbb77BsmXLsH79eiQnJ+OZZ57Bv//+i3///ReLFy/G999/j99//93suS+//BJt27bF4cOH8cEHH2DUqFHYtGmTRTv0ej0SEhIQGhqKnTt3Yvfu3QgJCUGvXr2g0+kEhf3DDz/EzJkzcfDgQfj4+GDw4MH8vRUrVmDSpEn4/PPPcfDgQdStW9dMnADAkiVLMHHiREydOhUnT57E559/jo8//hiLFi0CAHzzzTcoLi7GBx98wL8vLy8P3333nSA7CYKwgtKnSRIEoSy///47V6NGDS4gIIB74IEHuAkTJnBHjhyx+czKlSu5WrVq8b8TExM5ANy5c+f4a//73/+4oKAgs9NwExISuP/973/878aNG3O9evUy8/vFF1/kevfuzf8GwP35558cx3Hc4sWLuebNm3NGo5G/r9VqucDAQG7Dhg0Wba08ufjw4cMcx9059Xfz5s28m7Vr13IAuNLSUo7jOC4uLo576623zPzp3Lkz17ZtW/5306ZNuaVLl5q5mTx5MhcXF8f/3rNnD+fr68t9/PHHnI+PD7dz506LNhIEIRzqgSEID6dfv364du0a/v77b/Tq1QvJycno0KEDkpKSeDebN29Gjx49UL9+fYSGhmLAgAG4desWSkpKeDdBQUFo2rQp/zsqKgpNmjRBSEiI2bWcnByz98fFxVX7ffLkSYu2HjlyBOfOnUNoaCjfe1SzZk2UlZXh/PnzgsLdpk0b/v9169YFAN62kydPonPnzlbtLC4uxvnz5zFkyBDejpCQEEyZMsXMjri4OLz//vuYPHky3nvvPXTr1k2QjQRBWIcm8RIEgYCAADz22GN47LHH8PHHH2Po0KH45JNP8Nprr+HixYt44oknMHz4cEydOhU1a9bErl27MGTIEOh0OgQFBQEAfH19zfzUaDQWrxmNRtF2FhUVoWPHjliyZEm1e3Xq1BHkl6ltGo0GABy2rXL+z48//lhN6Hh7e/P/NxqN2L17N7y9vXHu3DlB9hEEYRvqgSEIohotW7ZEcXExACA1NRVGoxEzZ85Ely5dcM899+DatWuSvWvv3r3VfsfGxlp026FDB5w9exaRkZFo1qyZ2V94eLhkNsXGxmLfvn1W7YyKikK9evVw4cKFanbExMTw7r788kucOnUK27dvx/r165GYmCiZjQTh6ZCAIQgP5tatW3j00Ufx66+/4ujRo8jIyMDKlSsxY8YMPP300wCAZs2aQa/X49tvv8WFCxewePFiLFiwQDIbdu/ejRkzZuDMmTOYO3cuVq5ciVGjRll0279/f9SuXRtPP/00du7ciYyMDCQnJ+Odd97BlStXJLNp1KhR+Pnnn5GYmIgzZ87gk08+wfHjx83cfPrpp5g2bRq++eYbnDlzBunp6UhMTMSsWbMAAIcPH8bEiROxcOFCdO3aFbNmzcKoUaNw4cIFyewkCE+GBAxBeDAhISHo3LkzZs+ejYceegj33nsvPv74Y7zxxhv8apm2bdti1qxZ+OKLL3DvvfdiyZIlmDZtmmQ2vPfeezh48CDat2+PKVOmYNasWUhISLDoNigoCDt27ECjRo3w7LPPIjY2ll/iHBYWJplNL774Ij7++GOMGzcOHTt2xKVLlzB8+HAzN0OHDsXChQuRmJiI1q1b4+GHH0ZSUhJiYmJQVlaGV199Fa+99hqefPJJAMCbb76JRx55BAMGDDBbSk4QhDg0HMdxShtBEIRn0qRJE4wePRqjR49W2hSCIFQG9cAQBEEQBKE6SMAQBEEQBKE6aAiJIAiCIAjVQT0wBEEQBEGoDhIwBEEQBEGoDhIwBEEQBEGoDhIwBEEQBEGoDhIwBEEQBEGoDhIwBEEQBEGoDhIwBEEQBEGoDhIwBEEQBEGoDhIwBEEQBEGojv8HW+GaK2vSlbgAAAAASUVORK5CYII=", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "#BRAND NEW TESTING SCRIPT FOR ALL METHODS BASED OFF PREVIOUS TWO\n", "#ENHANCED VERSION OF MY PREVIOUS 2 TESTING SCRIPTS WITH EXTRAS\n", "#Testing script for Granite3.2-2B-Instruct using NF4 base + FP16 Adapters\n", "\n", "import os\n", "import torch\n", "import time\n", "import json\n", "import numpy as np\n", "import pandas as pd\n", "import evaluate\n", "import nltk\n", "import gc\n", "import math\n", "import re\n", "import matplotlib.pyplot as plt\n", "import mauve\n", "from transformers import AutoModelForCausalLM, AutoTokenizer, TrainingArguments, Trainer, EarlyStoppingCallback, TrainerCallback, BitsAndBytesConfig\n", "from transformers.trainer_utils import get_last_checkpoint\n", "from peft import LoraConfig, get_peft_model, prepare_model_for_kbit_training, PeftModel\n", "from datasets import Dataset\n", "from sklearn.metrics import accuracy_score, precision_recall_fscore_support\n", "from nltk.translate.bleu_score import sentence_bleu\n", "from rouge_score import rouge_scorer\n", "from torch.utils.data import DataLoader\n", "from sentence_transformers import SentenceTransformer, util\n", "\n", "nltk.download(\"punkt\")\n", "os.environ[\"TOKENIZERS_PARALLELISM\"] = \"false\"\n", "os.environ[\"PYTORCH_CUDA_ALLOC_CONF\"] = \"expandable_segments:True\"\n", "\n", "MODEL_NAME = \"ibm-granite/granite-3.2-2b-instruct\"\n", "ADAPTER_PATH = \"Granite3.2-2B-lora_adapters-FP16\"\n", "TEST_CSV_PATH = \"Testing Dataset RE.csv\"\n", "OUTPUT_JSON_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/NF4/FP16/Granite3.2-2B-NF4-lora-FP16-Evaluation_Results.json\"\n", "OUTPUT_INFER_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/NF4/FP16/Granite3.2-2B-NF4-lora-FP16-Inference_Curve.png\"\n", "OUTPUT_MEMORY_USAGE_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/NF4/FP16/Granite3.2-2B-NF4-lora-FP16-Memory_Usage_Curve.png\"\n", "OUTPUT_LATENCY_HIST_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/NF4/FP16/Granite3.2-2B-NF4-lora-FP16-Latency_Histogram.png\"\n", "OUTPUT_MEMORY_HIST_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/NF4/FP16/Granite3.2-2B-NF4-lora-FP16-Memory_Histogram.png\"\n", "SEMANTIC_MODEL = \"all-MiniLM-L6-v2\"\n", "\n", "tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)\n", "tokenizer.pad_token = tokenizer.eos_token\n", "tokenizer.padding_side = \"right\"\n", "\n", "test_df = pd.read_csv(TEST_CSV_PATH)\n", "\n", "def preprocess_function(examples):\n", " inputs = []\n", " labels = []\n", " \n", " for context, question, answer in zip(\n", " examples.get(\"Context\", [\"\"] * len(examples[\"Question\"])),\n", " examples[\"Question\"], \n", " examples[\"Answer\"]):\n", " \n", " context = context.strip() if context else \"\"\n", " question = question.strip()\n", " answer = answer.strip()\n", "\n", " if context:\n", " prompt = f\"Context: {context}\\nQuestion: {question}\\nAnswer:\"\n", " else:\n", " prompt = f\"Question: {question}\\nAnswer:\"\n", "\n", " full_text = prompt + \" \" + answer\n", " \n", " tokenized = tokenizer(full_text, padding=\"max_length\", truncation=True, max_length=512)\n", " prompt_ids = tokenizer(prompt, truncation=True, max_length=512, add_special_tokens=False)[\"input_ids\"]\n", "\n", " input_ids = tokenized[\"input_ids\"]\n", " attention_mask = tokenized[\"attention_mask\"]\n", " label_ids = input_ids.copy()\n", " label_ids[:len(prompt_ids)] = [-100] * len(prompt_ids)\n", " \n", " if all(id_ == -100 for id_ in label_ids):\n", " continue\n", "\n", " inputs.append({\"input_ids\": input_ids, \"attention_mask\": attention_mask, \"labels\": label_ids})\n", "\n", " return {\"input_ids\": [x[\"input_ids\"] for x in inputs], \"attention_mask\": [x[\"attention_mask\"] for x in inputs],\n", " \"labels\": [x[\"labels\"] for x in inputs]}\n", "\n", "test_dataset = Dataset.from_pandas(test_df).map(preprocess_function, batched=True, batch_size=32,\n", " remove_columns=test_df.columns.tolist())\n", "\n", "bnb_config = BitsAndBytesConfig(\n", " load_in_4bit=True, \n", " bnb_4bit_quant_type=\"nf4\", \n", " bnb_4bit_compute_dtype=torch.float16, \n", " bnb_4bit_use_double_quant=True)\n", "\n", "model = AutoModelForCausalLM.from_pretrained(MODEL_NAME, quantization_config=bnb_config, device_map=\"auto\")\n", "model = PeftModel.from_pretrained(model, ADAPTER_PATH).eval()\n", "model.config.pad_token_id = tokenizer.pad_token_id\n", "\n", "# Load semantic similarity model\n", "semantic_model = SentenceTransformer(SEMANTIC_MODEL)\n", "\n", "def compute_loss_and_perplexity():\n", " losses = []\n", " for sample in test_dataset:\n", " with torch.no_grad():\n", " input_ids = torch.tensor(sample[\"input_ids\"]).unsqueeze(0).to(\"cuda\")\n", " attention_mask = torch.tensor(sample[\"attention_mask\"]).unsqueeze(0).to(\"cuda\")\n", " labels = torch.tensor(sample[\"labels\"]).unsqueeze(0).to(\"cuda\")\n", " \n", " loss = model(input_ids=input_ids, attention_mask=attention_mask, labels=labels).loss.item()\n", " losses.append(loss)\n", " \n", " avg_loss = sum(losses) / len(losses)\n", " return avg_loss, math.exp(avg_loss)\n", "\n", "def extract_answer(text):\n", " return text.split(\"Answer:\")[-1].strip() if \"Answer:\" in text else text.strip()\n", "\n", "def normalize(text):\n", " return re.sub(r\"[^\\w\\s]\", \"\", text.strip().lower())\n", "\n", "def compute_metrics(preds, refs):\n", " decoded_preds = tokenizer.batch_decode(preds, skip_special_tokens=True)\n", " #decoded_refs = tokenizer.batch_decode(refs, skip_special_tokens=True)\n", "\n", " # Replace -100s in refs before decoding\n", " safe_refs = [[token if token != -100 else tokenizer.pad_token_id for token in ref] for ref in refs]\n", " decoded_refs = tokenizer.batch_decode(safe_refs, skip_special_tokens=True)\n", "\n", " preds_clean = [normalize(extract_answer(p)) for p in decoded_preds]\n", " refs_clean = [normalize(extract_answer(r)) for r in decoded_refs]\n", "\n", " sim_scores = util.cos_sim(semantic_model.encode(preds_clean, convert_to_tensor=True),\n", " semantic_model.encode(refs_clean, convert_to_tensor=True)).diagonal()\n", " semantic_threshold = 0.8\n", " matches = [1 if sim >= semantic_threshold else 0 for sim in sim_scores]\n", "\n", " accuracy = sum(matches) / len(matches)\n", " precision, recall, f1, _ = precision_recall_fscore_support(matches, matches, average=\"binary\", zero_division=0)\n", " avg_bleu = sum([sentence_bleu([r.split()], p.split()) for r, p in zip(refs_clean, preds_clean)]) / len(preds_clean)\n", "\n", " rouge = rouge_scorer.RougeScorer([\"rouge1\", \"rouge2\", \"rougeL\"], use_stemmer=True)\n", " rouge_scores = [rouge.score(ref, pred) for ref, pred in zip(refs_clean, preds_clean)]\n", " avg_rouge = {k: sum([s[k].fmeasure for s in rouge_scores]) / len(rouge_scores) for k in rouge_scores[0]}\n", "\n", " return {\"accuracy:\": accuracy, \"precision:\": precision, \"recall:\": recall, \"f1:\": f1,\n", " \"bleu:\": avg_bleu, \"rouge:\": avg_rouge, \"semantic_similarity_avg:\": sim_scores.mean().item()}, decoded_preds, decoded_refs\n", "\n", "def measure_inference_and_generate():\n", " preds, latencies, memory_used_per_sample, peak_memories = [], [], [], []\n", "\n", " #Measure model load memory (after full load + preparation)\n", " gc.collect()\n", " torch.cuda.empty_cache()\n", " torch.cuda.reset_peak_memory_stats()\n", " model_load_memory = torch.cuda.memory_allocated() / (1024 ** 3)\n", "\n", " for idx, sample in enumerate(test_dataset):\n", " gc.collect()\n", " torch.cuda.empty_cache()\n", " torch.cuda.reset_peak_memory_stats()\n", "\n", " input_ids = torch.tensor(sample[\"input_ids\"]).unsqueeze(0).to(\"cuda\")\n", " attention_mask = torch.tensor(sample[\"attention_mask\"]).unsqueeze(0).to(\"cuda\")\n", " \n", " # Measure base memory BEFORE\n", " base_memory = torch.cuda.memory_allocated()\n", "\n", " # Wait for everything to settle\n", " torch.cuda.synchronize()\n", " #mem_before = torch.cuda.memory_allocated() / (1024 ** 3)\n", " start_time = time.time()\n", "\n", " with torch.no_grad():\n", " output = model.generate(input_ids=input_ids, attention_mask=attention_mask, max_new_tokens=50,\n", " do_sample=True, top_p=0.9, top_k=50,\n", " temperature=0.7, repetition_penalty=1.1, length_penalty=0.8)\n", "\n", " torch.cuda.synchronize()\n", " end_time = time.time()\n", " #mem_after = torch.cuda.memory_allocated() / (1024 ** 3)\n", " peak_memory = torch.cuda.max_memory_allocated() \n", "\n", " inference_memory = (peak_memory - base_memory) / (1024 ** 3) # in GB\n", "\n", " preds.append(output[0].tolist())\n", " latencies.append((end_time - start_time) * 1000) # ms\n", " memory_used_per_sample.append(inference_memory) # Memory used by this inference\n", " peak_memories.append(peak_memory / (1024 ** 3)) # Peak memory usage during this sample\n", "\n", " # Calculate averages now\n", " avg_inference_memory = np.mean(memory_used_per_sample)\n", "\n", " return preds, latencies, memory_used_per_sample, peak_memories, model_load_memory, avg_inference_memory\n", "\n", "def compute_mauve(pred_texts, ref_texts):\n", " return mauve.compute_mauve(p_text=pred_texts, q_text=ref_texts,\n", " device_id=0, max_text_length=256).mauve\n", "\n", "print(\"Generating predictions...\")\n", "loss, perplexity = compute_loss_and_perplexity()\n", "generated_preds, latencies, memory_used_per_sample, peak_memories, model_load_memory, avg_inference_memory = measure_inference_and_generate()\n", "ref_labels = [sample[\"labels\"] for sample in test_dataset]\n", "metrics, decoded_preds, decoded_refs = compute_metrics(generated_preds, ref_labels)\n", "mauve_score = compute_mauve(decoded_preds, decoded_refs)\n", "\n", "# 1) Plot Inference_Performance curves for latency and memory usage\n", "plt.plot(latencies, label=\"Latency (ms)\")\n", "plt.plot(memory_used_per_sample, label=\"Memory Usage (GB)\")\n", "plt.title(\"Inference_Performance Curve\")\n", "plt.xlabel(\"Sample Index\")\n", "plt.legend()\n", "plt.savefig(OUTPUT_INFER_PATH)\n", "\n", "# 2a) Compute latency stats and then plot the latency histogram\n", "latencies_np = np.array(latencies)\n", "latency_stats = {\n", " \"min_latency_ms\": float(np.min(latencies_np)),\n", " \"max_latency_ms\": float(np.max(latencies_np)),\n", " \"lower_quartile_ms\": float(np.percentile(latencies_np, 25)),\n", " \"median_latency_ms\": float(np.median(latencies_np)),\n", " \"upper_quartile_ms\": float(np.percentile(latencies_np, 75)),\n", " \"avg_latency_ms\": float(np.mean(latencies_np))\n", "}\n", "\n", "# 2b) Plot the Histogram for Latency (ms)\n", "plt.figure(figsize=(8, 6))\n", "plt.hist(latencies, bins=20, color='skyblue', edgecolor='black')\n", "plt.axvline(latency_stats[\"min_latency_ms\"], color='green', linestyle='dashed', linewidth=1, label='Min')\n", "plt.axvline(latency_stats[\"lower_quartile_ms\"], color='orange', linestyle='dashed', linewidth=1, label='Q1')\n", "plt.axvline(latency_stats[\"median_latency_ms\"], color='red', linestyle='dashed', linewidth=1, label='Median')\n", "plt.axvline(latency_stats[\"upper_quartile_ms\"], color='purple', linestyle='dashed', linewidth=1, label='Q3')\n", "plt.axvline(latency_stats[\"max_latency_ms\"], color='black', linestyle='dashed', linewidth=1, label='Max')\n", "plt.title(\"Latency Histogram\")\n", "plt.xlabel(\"Latency (ms)\")\n", "plt.ylabel(\"Frequency\")\n", "plt.legend()\n", "plt.savefig(OUTPUT_LATENCY_HIST_PATH)\n", "\n", "# Line plot focusing on 0.1MB to 1MB\n", "plt.figure(figsize=(8, 6))\n", "plt.plot(memory_used_per_sample, label=\"Memory Usage (GB)\", color=\"teal\")\n", "plt.ylim(0.1, 0.5) # Zoom in to 0.1GB–0.5GB range\n", "plt.title(\"Memory Usage per Sample (Zoomed 100MB–500MB)\")\n", "plt.xlabel(\"Sample Index\")\n", "plt.ylabel(\"Memory (GB)\")\n", "plt.legend()\n", "plt.savefig(OUTPUT_MEMORY_USAGE_PATH)\n", "\n", "# 4) Compute memory stats and Plot the Histogram for memory usage\n", "memory_used_per_sample_np = np.array(memory_used_per_sample)\n", "memory_stats = {\n", " \"min_memory_gb\": float(np.min(memory_used_per_sample_np)),\n", " \"max_memory_gb\": float(np.max(memory_used_per_sample_np)),\n", " \"lower_quartile_gb\": float(np.percentile(memory_used_per_sample_np, 25)),\n", " \"median_memory_gb\": float(np.median(memory_used_per_sample_np)),\n", " \"upper_quartile_gb\": float(np.percentile(memory_used_per_sample_np, 75)),\n", " \"avg_memory_gb\": float(np.mean(memory_used_per_sample_np)),\n", " \"model_load_memory_gb\": model_load_memory,\n", " \"avg_inference_memory_gb\": avg_inference_memory\n", "}\n", "\n", "# Plot the Histogram for memory usage\n", "plt.figure(figsize=(8, 6))\n", "plt.hist(memory_used_per_sample, bins=20, color='lightcoral', edgecolor='black')\n", "plt.axvline(memory_stats[\"min_memory_gb\"], color='green', linestyle='dashed', linewidth=1, label='Min')\n", "plt.axvline(memory_stats[\"lower_quartile_gb\"], color='orange', linestyle='dashed', linewidth=1, label='Q1')\n", "plt.axvline(memory_stats[\"median_memory_gb\"], color='red', linestyle='dashed', linewidth=1, label='Median')\n", "plt.axvline(memory_stats[\"upper_quartile_gb\"], color='purple', linestyle='dashed', linewidth=1, label='Q3')\n", "plt.axvline(memory_stats[\"max_memory_gb\"], color='black', linestyle='dashed', linewidth=1, label='Max')\n", "plt.title(\"Memory Usage Histogram\")\n", "plt.xlabel(\"Memory Usage (GB)\")\n", "plt.ylabel(\"Frequency\")\n", "plt.legend()\n", "plt.savefig(OUTPUT_MEMORY_HIST_PATH)\n", "\n", "# Save all results\n", "results = {\"eval_loss:\": loss, \"perplexity:\": perplexity, \"performance_metrics:\": metrics, \"mauve:\": mauve_score,\n", " \"inference_performance:\": {**latency_stats, **memory_stats}}\n", "\n", "with open(OUTPUT_JSON_PATH, \"w\") as f:\n", " json.dump(results, f, indent=4)\n", "\n", "print(f\"Evaluation Complete. Results saved to {OUTPUT_JSON_PATH}\")\n" ] }, { "cell_type": "code", "execution_count": 2, "id": "f9bc3665-63a7-4a58-a465-bf2554334a20", "metadata": {}, "outputs": [], "source": [ "#2)##########################################################################################################################\n", "#STARTED ABOVE TESTING AT 2:39PM ON 29/04/25\n", "#FEATURISING STARTED (MIN AFTER STARTING)\n", "#ENDED ABOVE TESTING AT 3:25PM (46 MIN AFTER STARTING)" ] }, { "cell_type": "code", "execution_count": 3, "id": "81f02dab-26c2-4de6-8d6a-79ff861e98d1", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "[nltk_data] Downloading package punkt to /home/jovyan/nltk_data...\n", "[nltk_data] Package punkt is already up-to-date!\n", "Map: 100%|██████████| 1500/1500 [00:02<00:00, 603.36 examples/s]\n", "Loading checkpoint shards: 100%|██████████| 2/2 [00:26<00:00, 13.44s/it]\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Generating predictions...\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "/home/jovyan/Falcon1B/lib/python3.11/site-packages/transformers/generation/configuration_utils.py:695: UserWarning: `num_beams` is set to 1. However, `length_penalty` is set to `0.8` -- this flag is only used in beam-based generation modes. You should set `num_beams>1` or unset `length_penalty`.\n", " warnings.warn(\n", "Featurizing p: 100%|██████████| 1498/1498 [00:42<00:00, 34.87it/s]\n", "Featurizing q: 100%|██████████| 1498/1498 [00:42<00:00, 35.31it/s]\n", "WARNING clustering 2996 points to 150 centroids: please provide at least 5850 training points\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Evaluation Complete. Results saved to Complete_Evaluation_Results/Granite3.2-2B/INT8/FP16/Granite3.2-2B-INT8-lora-FP16-Evaluation_Results.json\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAArcAAAIjCAYAAAAZajMiAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjEsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvc2/+5QAAAAlwSFlzAAAPYQAAD2EBqD+naQAAgJlJREFUeJzt3XlcVNX/P/DXzMDAsAzDvm8KKG4oWoZZVppoamqpLbhblku5tGnlUpamfTTLXFpcWjRLszI/ZZm5lGvhhoKKooDIvq8DzNzfH/6c72catAGBM4yv5+Mxj+Sec+e8LvfzwTfHM+fKJEmSQERERERkBeSiAxARERERNRYWt0RERERkNVjcEhEREZHVYHFLRERERFaDxS0RERERWQ0Wt0RERERkNVjcEhEREZHVYHFLRERERFaDxS0RERERWQ0Wt0REtxmZTIb58+eLjkFE1CRY3BKRxduwYQNkMhn+/vvvW36viooKzJ8/H3v37r31YBZg/vz5kMlkyMvLq7M9JCQEAwcOvOVxNm3ahOXLl9/y+xARNTUb0QGIiJpTRUUF3njjDQDAfffdJzaMIJWVlbCxqd+P/02bNuH06dOYPn1604QiImoknLklIrrN2Nvb17u4Fa28vFx0BCJqIVjcEpFVqK6uxty5c9G1a1e4uLjA0dER99xzD/bs2WPoc/nyZXh6egIA3njjDchkMpP1p2fPnsWwYcPg5uYGe3t7dOvWDdu3bzca6/oyiQMHDmDmzJnw9PSEo6Mjhg4ditzcXJNsP//8M3r16gVnZ2eo1Wrccccd2LRpEwBg3rx5sLW1rfO8iRMnQqPRoKqqqjG+RQb/vObS0lJMnz4dISEhsLOzg5eXFx588EEcO3YMwLUZ7v/+979ITU01fM9CQkIM5+fk5GDChAnw9vaGvb09oqKi8Nlnn5mMm5+fj1GjRkGtVkOj0WDMmDE4efIkZDIZNmzYYOg3duxYODk54eLFi3jooYfg7OyMuLg4AMAff/yB4cOHIygoCHZ2dggMDMSMGTNQWVlpNNb190hLS8PAgQPh5OQEf39/rFy5EgCQkJCABx54AI6OjggODjbcDyJq+VrWr+5ERDdQUlKCTz/9FE888QSefvpplJaWYu3atYiNjcXRo0fRuXNneHp6YvXq1Zg0aRKGDh2KRx55BADQqVMnAMCZM2dw9913w9/fH7NmzYKjoyO++eYbDBkyBN9++y2GDh1qNOZzzz0HV1dXzJs3D5cvX8by5csxdepUfP3114Y+GzZswPjx49G+fXvMnj0bGo0Gx48fx86dO/Hkk09i1KhRePPNN/H1119j6tSphvOqq6uxdetWPProo7C3t//X6y8oKKjzuF6v/9dzn332WWzduhVTp05Fu3btkJ+fjz///BNJSUmIjo7Ga6+9huLiYly5cgXvvfceAMDJyQnAtSUO9913Hy5cuICpU6ciNDQUW7ZswdixY1FUVIRp06YZcgwaNAhHjx7FpEmT0LZtW/zwww8YM2ZMnZlqa2sRGxuLnj174j//+Q8cHBwAAFu2bEFFRQUmTZoEd3d3HD16FCtWrMCVK1ewZcsWo/fQ6XTo378/7r33XixZsgQbN27E1KlT4ejoiNdeew1xcXF45JFHsGbNGowePRoxMTEIDQ391+8XEVk4iYjIwq1fv14CIP3111837FNbWytptVqjY4WFhZK3t7c0fvx4w7Hc3FwJgDRv3jyT9+jdu7fUsWNHqaqqynBMr9dLPXr0kMLDw03y9OnTR9Lr9YbjM2bMkBQKhVRUVCRJkiQVFRVJzs7OUvfu3aXKykqjsf73vJiYGKl79+5G7du2bZMASHv27LnhNUuSJM2bN08CcNPXgAEDjM755/W7uLhIU6ZMuek4AwYMkIKDg02OL1++XAIgffnll4Zj1dXVUkxMjOTk5CSVlJRIkiRJ3377rQRAWr58uaGfTqeTHnjgAQmAtH79esPxMWPGSACkWbNmmYxXUVFhcmzRokWSTCaTUlNTTd5j4cKFhmOFhYWSSqWSZDKZtHnzZsPxs2fP3vB/E0TU8nBZAhFZBYVCAaVSCeDaLGFBQQFqa2vRrVs3wz+v30xBQQF+//13jBgxAqWlpcjLy0NeXh7y8/MRGxuL5ORkZGRkGJ0zceJEyGQyw9f33HMPdDodUlNTAQC7du1CaWkpZs2aZTL7+r/njR49GkeOHMHFixcNxzZu3IjAwED06tXLrOv/9ttvsWvXLpOXt7f3v56r0Whw5MgRXL161ayx/tdPP/0EHx8fPPHEE4Zjtra2eP7551FWVoZ9+/YBAHbu3AlbW1s8/fTThn5yuRxTpky54XtPmjTJ5JhKpTL8uby8HHl5eejRowckScLx48dN+j/11FOGP2s0GrRp0waOjo4YMWKE4XibNm2g0WiQkpJi5lUTkSXjsgQishqfffYZli5dirNnz6KmpsZw3Jx/ar5w4QIkScKcOXMwZ86cOvvk5OTA39/f8HVQUJBRu6urKwCgsLAQAAzFaocOHW469mOPPYbp06dj48aNmDt3LoqLi7Fjxw7MmDHDqAi+mXvvvRceHh4mx81Z0rBkyRKMGTMGgYGB6Nq1Kx566CGMHj0arVq1+tdzU1NTER4eDrnceK4kMjLS0H79v76+voblBdeFhYXV+b42NjYICAgwOZ6Wloa5c+di+/bthu/zdcXFxUZf29vbG9ZYX+fi4oKAgACT76uLi4vJ+xFRy8TiloiswpdffomxY8diyJAheOmll+Dl5QWFQoFFixYZzYjeyPW1qS+++CJiY2Pr7PPPQkyhUNTZT5KkemV3dXXFwIEDDcXt1q1bodVqMXLkyHq9T0ONGDEC99xzD7777jv8+uuvePfdd7F48WJs27YN/fv3b5YM/2RnZ2dSMOt0Ojz44IMoKCjAK6+8grZt28LR0REZGRkYO3asyfriG92fxrpvRGSZWNwSkVXYunUrWrVqhW3bthnNys2bN8+o341mQq/PUtra2qJPnz6Nkql169YAgNOnT99whvK60aNHY/Dgwfjrr7+wceNGdOnSBe3bt2+UHObw9fXF5MmTMXnyZOTk5CA6Ohpvv/22obi90fctODgYp06dgl6vNypGz549a2i//t89e/agoqLCaPb2woULZmdMSEjA+fPn8dlnn2H06NGG47t27TL/QonI6nHNLRFZheuzcf87+3bkyBEcOnTIqN/1wqqoqMjouJeXF+677z589NFHyMzMNHn/urbq+jd9+/aFs7MzFi1aZLKd1z9nCfv37w8PDw8sXrwY+/bta7ZZW51OZ/LP+V5eXvDz84NWqzUcc3R0NOkHAA899BCysrKMdoiora3FihUr4OTkZFgzHBsbi5qaGnzyySeGfnq93rA1lznquseSJOH99983+z2IyPpx5paIWox169Zh586dJsenTZuGgQMHYtu2bRg6dCgGDBiAS5cuYc2aNWjXrh3KysoMfVUqFdq1a4evv/4aERERcHNzQ4cOHdChQwesXLkSPXv2RMeOHfH000+jVatWyM7OxqFDh3DlyhWcPHmyXnnVajXee+89PPXUU7jjjjvw5JNPwtXVFSdPnkRFRYXRXrC2trZ4/PHH8eGHH0KhUBh9QKsplZaWIiAgAMOGDUNUVBScnJzw22+/4a+//sLSpUsN/bp27Yqvv/4aM2fOxB133AEnJycMGjQIEydOxEcffYSxY8ciPj4eISEh2Lp1Kw4cOIDly5fD2dkZADBkyBDceeedeOGFF3DhwgW0bdsW27dvN2xhZs7a4rZt26J169Z48cUXkZGRAbVajW+//ZZrZYnImLiNGoiIzHN9660bvdLT0yW9Xi8tXLhQCg4Oluzs7KQuXbpIO3bskMaMGWOyhdXBgwelrl27Skql0mQLqIsXL0qjR4+WfHx8JFtbW8nf318aOHCgtHXrVpM8/9yabM+ePXVu37V9+3apR48ekkqlktRqtXTnnXdKX331lcl1Hj16VAIg9e3b1+zvzfWtwHJzc+tsDw4OvulWYFqtVnrppZekqKgoydnZWXJ0dJSioqKkVatWGZ1TVlYmPfnkk5JGo5EAGH1Ps7OzpXHjxkkeHh6SUqmUOnbsaLS113W5ubnSk08+KTk7O0suLi7S2LFjpQMHDkgAjLbmGjNmjOTo6Fjn9SQmJkp9+vSRnJycJA8PD+npp5+WTp48Wed2YnW9R69evaT27dub9X0iopZJJklcQU9EZAlOnjyJzp074/PPP8eoUaNEx2kW33//PYYOHYo///wTd999t+g4RGQFuOaWiMhCfPLJJ3BycjI8Oc3a/PMRuTqdDitWrIBarUZ0dLSgVERkbbjmlohIsB9//BGJiYn4+OOPDY+HtUbPPfccKisrERMTA61Wi23btuHgwYNYuHCh0cMZiIhuBZclEBEJFhISguzsbMTGxuKLL74wfAjL2mzatAlLly7FhQsXUFVVhbCwMEyaNAlTp04VHY2IrAiLWyIiIiKyGlxzS0RERERWg8UtEREREVkNfqAM156Sc/XqVTg7O5u1kTgRERERNS9JklBaWgo/Pz+jx33/E4tbAFevXkVgYKDoGERERET0L9LT0xEQEHDDdha3gOGTyenp6VCr1U0+3omsE3huUy/8EQDA52Ng0ERg3z6gc+cmH5uIyNJkncjC+l7rMW7fOPh09hEdh4jMcOLECfTq1Qv79u1D52aqX0pKShAYGPivO8qwuMX/PdNcrVY3S3HrVO4EhT2gdgDg6PD/DzoBzTA2EZGlKXcqhz3s4ezk3Cw/g4no1jk5ORn+29z/v/23JaT8QBkRERERWQ0Wt0RERERkNVjcEhEREZHV4BPKcG2BsouLC4qLi5tl3Ui1rho5pVfgpQCUNl5AfhHg5QUolU0+NhGRpdFV61CeUw5HL0colArRceg2JEkSamtrodPpREdpMaqrq1FQUAA3NzcoG6l+USgUsLGxueGaWnPrNX6gTAClQokATav/OxDgJC4MEZFgCqUC6gB+kIzEqK6uRmZmJioqKkRHaZEyMjIa9f0cHBzg6+t7SwUzi1sBUgpTsHTXFCz2kMHJ82XgzZXA4sVAq1b/fjIRkZUpTCnEb6/8hj6L+8C1lavoOHQb0ev1uHTpEhQKBfz8/KBUKvkwJzNVV1cjKysLPj4+jTJzK0kSqqurkZubi0uXLiE8PPymD2q4GRa3AhRVFeFQyk441QKQxQFbtwKzZ4uORUQkRFVRFRK3JqLn7J6io9Btprq6Gnq9HoGBgXBwcBAdp0XR6XQoKSmBv78/7O3tG+U9VSoVbG1tkZqaiurq6ga/Lz9QRkRERLe1hs4QUuNrjHvBu0lEREREVoPFLRERERFZDa65FcDP2Q9j75qNUhfAWdMRWLgQ8PMTHYuISAhnP2c8sPABOPvd/HnxRNQw9913Hzp37ozly5c32nsqlUr4+/s32jZgjYkztwL4OPng+fsWwrnLQiC007UPk/n4iI5FRCSEk48T7pl9D5x8uC0ikbnGjh0LmUyGZ5991qRtypQpkMlkGDt2LABg27ZtWLBgQaOOb2trC19fX9ja2jbq+zYGFrcCFFUV4aczm1CWsgnISQW2bweKikTHIiISoqqoCue2n0NVUZXoKEQtSmBgIDZv3ozKykrDsaqqKmzatAlBQUGGY25ubnB2btx/GamtrUVRURFqa2sb9X0bA4tbAVIKU/D6j3FwOhwHnPkTGDwYSEkRHYuISIjClEJsHrwZhSmFoqMQtSjR0dEIDAzEtm3bDMe2bduGoKAgdOnSxXDsvvvuw/Tp0w1fh4SEYOHChRg/fjycnZ0RFBSEjz/+uF5ja7VaXLhwAVqt9pavo7FxzS0RERHRP2SWZiKzLNPomKu9K0JdQ1FVW4XE3ESTc6J9owEA5/LOobym3KgtRBMCN5UbcstzkV6SbtTm6+QLX2ffBuUcP3481q9fj7i4OADAunXrMG7cOOzdu/em5y1duhQLFizAq6++iq1bt2LSpEno1asX2rRp06AcloTFLREREdE/fBT/Ed7Y94bRsbiOcfjykS9xpeQKun7c1eQcaZ4EABj7w1gcvnLYqO2LoV9gZKeR+ObMN5j681Sjtnm95mH+ffMblHPkyJGYPXs2UlNTAQAHDhzA5s2b/7W4feihhzB58mQAwCuvvIL33nsPe/bsYXFLREREZI2e6foMHm7zsNExV/trj4cOUAcgfmL8Dc/dMHhDnTO3ADCi/QjEBMYYtfk6NWzWFgA8PT0xYMAAbNiwAZIkYcCAAfDw8PjX8zp16mT4s0wmg4+PD3Jychqcw5KwuBXA3sYeAZrWqHIE7B2cgXbtgEZ6dB0RUUtjY28Dz3aesLHnX0lkOXydb7xUwN7G3rAEoS5tPG48++np6AlPR89bzve/xo8fj6lTr80Gr1y50qxz/rnLgUwmg16vN3tMuVwOe3t7i3y6G3+SCNDOsx0+fOh3JOblXTvwxRdAVRVw7FiTjenh4WH0yUkiIkvh2c4Tk89MFh2DqMXq168fqqurIZPJEBsb2yxjqlQqdOjQoVnGqi8WtwKkpaWhbWQkKisqmm1MlYMDziYlscAlIiKyMgqFAklJSYY/3+5Y3Apw4OIBRLxSgQOtlPg67SU8seA9vD9nGTJCwppkvJxLyfjm9UnIy8tjcUtEFifrRBbW37se4/aPg09nPtCGqCHUanWzjldRUYGzZ8+ibdu2cHBwaNax/w2LWwH0kh5yJeAor4aHbwBUVRXwDgkDIqNERyMianaSXkJ1aTUkvSQ6ClGLsWHDhpu2f//994Y//3PnhMuXL5v0P3HiRL3GlyQJer0ekmR5/7+1vFXAREREREQNxOKWiIiIiKwGi1siIiIishosbgUIcQrB2dXA2/rlSIm4H+s3/ob8kHDRsYiIhPBo64GJ8RPh0fbfN54nIstgb2+PyMhI2FvgPv38QJkAKhsVKtOALHkH+Du4oyrSXXQkIiJhbB1s4Rvd8Cc0EVHzUygUcHR0FB2jTpy5FSCzIhOBw4Ahunfgl/4XHlz0MtSZV0THIiISojitGP+d8l8UpxWLjkJEZtJqtUhNTYVWqxUdxQSLWwGKqovg0RXorfgNmvxUdN2yHqqifNGxiIiEqMirwN+r/kZFXvM92IaIbk1tbS1yc3NRW1srOooJFrdEREREZDVY3BIRERGR1WBxS0RERERWg8WtAG52bsg5CvyiewgFHq1xNO5ZVLh5io5FRCSEo5cj7ppxFxy9LPOT10SWKj09HePHj4efnx+USiWCg4Mxbdo05Of/3+d4tm3bhr59+8Ld3R0ymazej9m9EVtbW3h7e8PW1rZR3q8xsbgVwFvljYzvgB2KmcgK6ILfX1iAUm8/0bGIiIRQB6gRuywW6gC16ChELUZKSgq6deuG5ORkfPXVV7hw4QLWrFmD3bt3IyYmBgUFBQCA8vJy9OzZE4sXL27U8ZVKJQIDA6FUKhv1fRsD97kVoKK2Ao6tgAD9cajKfKC5mIbc8EjUODiJjkZE1Oyqy6qRnZAN747eUDpZ3l+URJZoypQpUCqV+PXXX6FSqQAAQUFB6NKlC1q3bo3XXnsNq1evxqhRowAAly9fbtTxdTodKisroVKpoFAoGvW9bxVnbgVILUtFxNPAK/KXEHLhD4we9xDcUi+KjkVEJET++Xys67EO+ee5JSJZkMpMoOCY8avs0rU2XZVpW8Gx/zu35Jxpm/baTCqqck3bKjPrFa2goAC//PILJk+ebChsr/Px8UFcXBy+/vprSJJ0K9+Bm6qqqsLZs2dRVVXVZGM0FGduiYiIiP4p+SPg9BvGx0LigB5fAhVXgJ1dTc958v8Xk4fGAvmHjdtivgBCRwJp3wB/TzVu6zAP6DTf/GjJyZAkCZGRkXW2R0ZGorCwELm5ufDy8jL7fa0Fi1siIiKifwp/Bgh42PiY0vXafx0CgH7xNz43ZgNQW258zDHk2n+DRgAeMcZtqoY9fvrfZmYtcT1sc2BxS0RERPRPKt8bF50Ke8At+sbnqtvcuM3e89rrFoSFhUEmkyEpKQlDhw41aU9KSoKnpyc0Gs0tjdNScc2tADZyG9RWACWSE2ptlKjQuEOv4O8ZRHR7ktvI4eDhALkN/0oiMoe7uzsefPBBrFq1CpWVlUZtWVlZ2LhxI8aOHdukGWQyGWxsbCCTyZp0nIbgTxIBwtXhSFgAvCb7Huc7PIwPfj+L3Ij2omMREQnh3ckbL+W+BO9O3qKjELUYH374IbRaLWJjY7F//36kp6dj586dePDBBxEREYG5c+cCuPbhsxMnTiAxMREAcO7cOZw4cQJZWVm3NL6DgwM6d+4MBweHW76WxsbiloiIiKiFCQ8Px19//YVWrVphxIgRCA4ORv/+/REREYEDBw7Ayena9qLbt29Hly5dMGDAAADA448/ji5dumDNmjUi4zcpFrcCXCy9iHavAK9JTyAs6Wc88/Ad8Lh4VnQsIiIhcs7k4IOwD5BzJkd0FKIWJSQkBBs2bEBWVhb0ej3mzp2LX3/9FadOnTL0GTt2LCRJMnnNnz//lsaurKxEQkKCybIIS8CFngJU66phpwF8ZLlQasvheuUyFNVa0bGIiITQaXUovFgInVYnOgpRi/bGG28gJCQEhw8fxp133gm5vOnmMPV6PbRaLfR6fZON0VAsbomIiIisxLhx40RHEI7LEoiIiIjIarC4JSIiIiKrweJWgEDHQFz4AnhfNw9poT3w9YdfozCwlehYRERCuIW5IW5nHNzC3ERHISIz2dvbIzw8HPb29qKjmOCaWwGcbJ1QmghcUNwDfxc/lPXwEx2JiEgYO7UdwmLDRMcgonpQKBRwcXERHaNOnLkVILcqFz79gT76NfDMTEDPNUvgmHtrmykTEbVUpZml2Dt/L0ozS0VHISIzVVdX4+rVq6iurhYdxQSLWwHyqvLgey8wWL4Vntnn0PPjd+GUly06FhGREGWZZdj3xj6UZZaJjkJEZqqpqcHVq1dRU1MjOooJFrdEREREZDVY3BIRERGRkb1790Imk6GoqAgAsGHDBmg0GqGZzMXiloiIiKiFGTt2LGQyGZ599lmTtilTpkAmk2Hs2LGNNt5jjz2G8+fPN9r7NSUWtwKolWoUngEO6nugROOH0/2HoUqtER2LiEgIe1d7dIzrCHtXy9tSiMiSBQYGYvPmzaisrDQcq6qqwqZNmxAUFNSoY6lUKnh5eRm+trGxgZubG2xsLG/jLRa3Avg7+OPyl8BX8jdxJaQHdry9GsX+waJjEREJ4Rrqike+fASuoa6ioxC1KNHR0QgMDMS2bdsMx7Zt24agoCB06dLFcEyv12PRokUIDQ2FSqVCVFQUtm7davReP/30EyIiIqBSqXD//ffj8uXLRu3/XJZw5coVzJgxA0FBQXBycsIdd9yB3377zeickJAQLFy4EOPHj4ezszOCgoLw8ccfN9434AZY3Aqg1Wlh5wW46i9DWVkETVoKFNoq0bGIiISorapFwYUC1FbVio5C9H8yM4Fjx4xfly5da6uqMm07duz/zj13zrStoOBaW26uaVtmZoNjjh8/HuvXrzd8vW7dOowbN86oz6JFi/D5559jzZo1OHPmDGbMmIGRI0di3759AID09HQ88sgjGDRoEE6cOIGnnnoKs2bNuum4JSUlePDBB7Fr1y4cP34c/fr1w6BBg5CWlmbUb+nSpejWrRuOHz+OyZMnY9KkSTh37lyDr9ccLG4FSClNQbsZwJvypxB27jc8O6Q7PFKa9kYTEVmq3MRcrAhfgdzEXNFRiP7PRx8BXbsav+bMudZ25YppW9eu/3fu2LGmbT/9dK3tm29M2z76qMExR44ciT///BOpqalITU3FgQMHMHLkSEO7VqvFwoULsW7dOsTGxqJVq1YYO3YsRo4ciY/+/7irV69G69atsXTpUrRp0wZxcXH/ul43IiICd911F1q3bo3w8HAsWLAArVu3xvbt2436PfTQQ5g8eTLCwsLwyiuvwMPDA3v27Gnw9ZrD8hZKEBEREYn2zDPAww8bH3P9/0tnAgKA+Pgbn7thA1BebnwsJOTaf0eMAGJijNt8fRsc09PTEwMGDMCGDRsgSRIGDBgADw8PQ/uFCxdQUVGBBx980Oi86upqw9KFpKQkdO/e3ag95p8Z/6GsrAzLly/HX3/9hezsbNTW1qKystJk5rZTp06GP8tkMvj4+CAnJ6dB12ouFrdERERE/+Tre+Oi094eiI6+8blt2ty4zdPz2qsRjR8/HlOnTgUArFy50qitrOzaw1H++9//wt/f36jNzs6uwWO++uqr2Lt3L95991106NABKpUKw4YNM3lima2trdHXMpkMer2+weOag8UtERERUQvWr18/VFdXQyaTITY21qitXbt2sLOzQ1paGnr16lXn+ZGRkSbLCQ4fPnzTMQ8fPoyBAwfi4YcfhqOjI8rKykw+hCYKi1siIiKiFkyhUCApKcnw5//l7OyMF198ETNmzIBer0fPnj1RXFyMAwcOQK1WY8yYMXj22WexdOlSvPTSS3jqqacQHx+PDRs23HTM1q1bY8+ePTh16hQcHBwwZ86cJp+RNReLWwEiNZE4Pht4buNv8O8chcRjw0RHIiISxjfaF/OkeaJjELVoarX6hm0LFiyAp6cnFi1ahJSUFGg0GkRHR+PVV18FAAQFBeHbb7/FjBkzsGLFCtx5552GLbxu5IMPPsD48ePRu3dveHh44JVXXkFJSUmjX1dDyCRJkkSHEK2kpAQuLi4oLi6+6f84GsuxY8fQtWtXTN34G/wjo5p8vIykk/gwrg/i4+MRfbM1QkRERLeRqqoqXLp0CaGhobC350NELMHN7om59Rq3AhPgctllRDwHPK9/BqHnd2PUmP5wu3xBdCwiIiHyzuVhbcxa5J3LEx2FiMxUVVWFpKQkVFVZ3j79LG4FqKythKMfEC6/CFVFIfwT/oZtZfm/n0hEZIVqymtw5fAV1JTXiI5CRGbS6XQoLy+HTqcTHcUEi1siIiIishosbomIiIjIarC4JSIiIiKrweJWAD8HP1z+DvhUPw0ZQV3x44JVKPYLEh2LiEgITYgGQ78YCk2IRnQUIjKTnZ0dQkNDb+kpZ02F+9wK4KJ0QeFR4KR8EPzdQlE8IFR0JCIiYVRuKnQa2enfOxKRxbCxsYG7u7voGHXizK0AhdpCeNwL3KX7Cm655xH99VqoCrkFDhHdnspzy3F05VGU53LXGKKWoqamBjk5OaipsbxdTljcCpBVmYXA/kCcYi18Mk6h7+JZUGdliI5FRCRESXoJfp76M0rSLePpRkT076qrq5GWlobq6mrRUUywuCUiIiIiq8HiloiIiKgFSk9Px/jx4+Hn5welUong4GBMmzYN+fn5hj7z589H27Zt4ejoCFdXV/Tp0wdHjhwRmLrpsbglIiIiamFSUlLQrVs3JCcn46uvvsKFCxewZs0a7N69GzExMSgoKAAARERE4MMPP0RCQgL+/PNPhISEoG/fvsjNzRV8BU2HuyUI4GjjiNJLwKmADih3ckfKXfeh2tFJdCwiIiGUzkq07tsaSmel6ChELcaUKVOgVCrx66+/QqVSAQCCgoLQpUsXtG7dGq+99hpWr16NJ5980ui8ZcuWYe3atTh16hR69+7d4PEVCgXUajUUCsUtXUdTYHErQJBTEC58DHzSazn8w6KQuup+0ZGIiIRxD3fHyF9Gio5BZKQ0sxRlmWVGx+xd7eEa6oraqlrkJprOfPpG+wIA8s7loabceBcBTYgGKjcVynPLTT486eTrBGdfZ7OzFRQU4JdffsHbb79tKGyv8/HxQVxcHL7++musWrUKMpnM0FZdXY2PP/4YLi4uiIqKMnu8utjb2yMiIuKW3qOpsLgVQCfpILcHlFIJ5LXVsKnSokblAMkCf/shImpqep0eNeU1sHW0hVzB1XJkGeI/ise+N/YZHesY1xGPfPkISq6U4OOuH5ucM0+aBwD4YewPuHL4ilHb0C+GotPITjjzzRn8PPVno7Ze83rhvvn3mZ0tOTkZkiQhMjKyzvbIyEgUFhYiNzcXXl5e2LFjBx5//HFUVFTA19cXu3btgoeHh9nj1UWSJOj1esjlcqMC2hKwuBXgfPF5RM0DlsoewfbTq/Hw+ElYv/E3ZEfe2m9RREQtUfbJbHzc9WNMjJ9omPkiEq3rM13R5uE2RsfsXe0BAOoANSbGT7zhuYM3DK5z5hYA2o9oj8CYQKM2J9+GLU2UJOmm7UrltaU+999/P06cOIG8vDx88sknGDFiBI4cOQIvL68GjQsAFRUVSEpKQmRkJBwdHRv8Pk2BxS0RERHRPzj7Ot9wqYCNvc1NfxHzaHPjWVFHT0c4et5aMRgWFgaZTIakpCQMHTrUpD0pKQmenp7QaDTXxnR0RFhYGMLCwnDXXXchPDwca9euxezZs28ph6Xiv/8QERERtSDu7u548MEHsWrVKlRWVhq1ZWVlYePGjRg7duwNz9fr9dBqtU2cUhwWt0REREQtzIcffgitVovY2Fjs378f6enp2LlzJx588EFERERg7ty5KC8vx6uvvorDhw8jNTUV8fHxGD9+PDIyMjB8+HDRl9BkWNwSERERtTDh4eH466+/0KpVK4wYMQLBwcHo378/IiIicODAATg5OUGhUODs2bN49NFHERERgUGDBiE/Px9//PEH2rdvL/oSmozQ4lan02HOnDkIDQ2FSqVC69atsWDBAqMF0pIkYe7cufD19YVKpUKfPn2QnJxs9D4FBQWIi4uDWq2GRqPBhAkTUFZW9s/hLEaYOgwJC4GXpc043+4hvL87Cblh7UTHIiISwqujF17MeRFeHRv+4Rai21FISAg2bNiArKws6PV6zJ07F7/++itOnToF4Np2Xdu2bUNGRga0Wi2uXr2KH374AXfcccctj61SqRAVFWWyFZklEFrcLl68GKtXr8aHH36IpKQkLF68GEuWLMGKFSsMfZYsWYIPPvgAa9aswZEjR+Do6IjY2FhUVVUZ+sTFxeHMmTPYtWsXduzYgf3792PixBt/ilE0W7ktakuBSpkHapUOqHT1gN7WVnQsIiIhFLYKOHo6QmHL7RCJbsUbb7yBDz74AIcPH4Zer2/SseRyOWxtbSGXW94iAKGJDh48iMGDB2PAgAEICQnBsGHD0LdvXxw9ehTAtVnb5cuX4/XXX8fgwYPRqVMnfP7557h69Sq+//57ANc+Ebhz5058+umn6N69O3r27IkVK1Zg8+bNuHr1ap3jarValJSUGL2aU3p5OlpNAMbpX0JQyh94dPpIaNIvNWsGIiJLUXCxAF89/BUKLhaIjkLU4o0bNw7Tp09v8qKzqqoKycnJRpONlkJocdujRw/s3r0b58+fBwCcPHkSf/75J/r37w8AuHTpErKystCnTx/DOS4uLujevTsOHToEADh06BA0Gg26detm6NOnTx/I5XIcOXKkznEXLVoEFxcXwyswMLDOfk2lrKYMLmFAtPw4nEqyEb7/F9iVNW+BTURkKbTFWpz/8Ty0xdb76W0ia6PT6VBcXAydTic6igmh+9zOmjULJSUlaNu2LRQKBXQ6Hd5++23ExcUBuLadBQB4e3sbneft7W1oy8rKMtmE2MbGBm5uboY+/zR79mzMnDnT8HVJSUmzF7hERERE1PiEFrfffPMNNm7ciE2bNqF9+/Y4ceIEpk+fDj8/P4wZM6bJxrWzs4OdnV2TvT8RERERiSG0uH3ppZcwa9YsPP744wCAjh07IjU1FYsWLcKYMWPg4+MDAMjOzoav7/89CSQ7OxudO3cGAPj4+CAnJ8fofWtra1FQUGA4n4iIiIhuD0LX3FZUVJgseFYoFIZP+IWGhsLHxwe7d+82tJeUlODIkSOIiYkBAMTExKCoqAjx8fGGPr///jv0ej26d+/eDFdRf14qL2TsAr7WjUa2bzvsnvkmSr34PHUiuj05+zuj79K+cPav+1GnRGR5lEolAgICoFQqRUcxIXTmdtCgQXj77bcRFBSE9u3b4/jx41i2bBnGjx8PAJDJZJg+fTreeusthIeHIzQ0FHPmzIGfnx+GDBkCAIiMjES/fv3w9NNPY82aNaipqcHUqVPx+OOPw8/PT+DV3Zi7nTtyfgf+nDAa/t7tkD+Se9wS0e3LydsJMTNjRMcgonqwtbW12H8hF1rcrlixAnPmzMHkyZORk5MDPz8/PPPMM5g7d66hz8svv4zy8nJMnDgRRUVF6NmzJ3bu3Al7e3tDn40bN2Lq1Kno3bs35HI5Hn30UXzwwQciLsksJdUl0EQD7XU7oS/UwPfvE7jcvRe0ao3oaEREza6ysBIpv6WgVZ9WULla3obwRGSqtrYWJSUlUKvVsLERWk6aEJrG2dkZy5cvx/Lly2/YRyaT4c0338Sbb755wz5ubm7YtGlTEyRsGhkVGQgdDjyr+A+2p7bGw69MwvqNvyGbxS0R3YaKLhVh64itmBg/kcUtUQuh1WqRkpKCyMhIiytuLe+xEkRERER0U2PHjoVMJsOzzz5r0jZlyhTIZDKMHTu2+YNZABa3RERERC1QYGAgNm/ejMrKSsOxqqoqbNq0CUFBQQKTicXiloiIiKgFio6ORmBgILZt22Y4tm3bNgQFBaFLly6GYzt37kTPnj2h0Wjg7u6OgQMH4uLFi4b2zz//HE5OTkhOTjYcmzx5Mtq2bYuKiormuZhGZFmLJG4T9gp7VGYDlwKCoFWpkdW2I2rtuM6MiG5PNiob+HTxgY2KfyWR5cjMzERmZqbRMVdXV4SGhqKqqgqJiYkm50RHRwMAzp07h/LycqO2kJAQuLm5ITc3F+np6UZtvr6+Rvv518f48eOxfv16w9Nd161bh3HjxmHv3r2GPuXl5Zg5cyY6deqEsrIyzJ07F0OHDsWJEycgl8sxevRo7NixA3FxcTh48CB++eUXfPrppzh06BAcHBzqHFcul8PBwcFkS1dLwJ8kAoQ6h+LscmDZxnXwbxOFi5v6io5ERCSMZ6Qnnjn2jOgYREY++ugjvPHGG0bH4uLi8OWXX+LKlSvo2rWryTmSJAG4th728OHDRm1ffPEFRo4ciW+++QZTp041aps3bx7mz5/foJwjR47E7NmzkZqaCgA4cOAANm/ebFTcPvroo0bnrFu3Dp6enkhMTESHDh0M19upUyc8//zz2LZtG+bPn1/nNV6nUqnQrp1lbmXK4paIiIjoH5555hk8/PDDRsdcXV0BAAEBAUYPj/qnDRs21DlzCwAjRowwPIjquobO2gKAp6cnBgwYgA0bNkCSJAwYMAAeHh5GfZKTkzF37lwcOXIEeXl5hodlpaWlGYpbV1dXrF27FrGxsejRowdmzZrV4EyisbgV4GzxWXReACyT+uGnkx9iwMTn8cVnPyO7bSfR0YiIml3m8UysvWstJhyeAN8ufFojWYabLRWwt7c3LEGoS5s2bW7Y5unpCU9Pz1vO97/Gjx9vmA1euXKlSfugQYMQHByMTz75BH5+ftDr9ejQoQOqq6uN+u3fvx8KhQKZmZkoLy+Hs/ONnxpYUVGBpKQkREZG3nDpgiiWt1DiNiBJEmQ2gK2sFjJJgk1NNfD//ymDiOi2IwG6ah3AH4NEDdKvXz9UV1ejpqYGsbGxRm35+fk4d+4cXn/9dfTu3RuRkZEoLCw0eY+DBw9i8eLF+PHHH+Hk5GSydOKfJEkyvCwNZ26JiIiIWjCFQoGkpCTDn/+Xq6sr3N3d8fHHH8PX1xdpaWkmSw5KS0sxatQoPP/88+jfvz8CAgJwxx13YNCgQRg2bFizXUdj4cwtERERUQunVquhVqtNjsvlcmzevBnx8fHo0KEDZsyYgXfffdeoz7Rp0+Do6IiFCxcCADp27IiFCxfimWeeQUZGRrPkb0ycuSUiIiJqYTZs2HDT9u+//97w5z59+phsXfa/ywnWrVtncv7MmTMxc+bMW8ooCmduBQh1DkXSCuBN/SqkRPTGp1v+QH5ohOhYRERCeER6YNLpSfCI9Pj3zkRkEVQqFdq3bw+VyvL26efMrQD2CntUXQVy5RFQOriiqrWr6EhERMLYqmzh1d5LdAwiqge5XG6RhS3AmVshMisyEfQ4MEz/FvzSjqD/m9Ohvpr+7ycSEVmhotQibH9qO4pSi0RHISIzabVaXL58GVqtVnQUEyxuBSiqLoJ7FNBLvheagnREfb8RquIC0bGIiISozK/E8bXHUZlfKToKEZmptrYWeXl5qK2tFR3FBItbIiIiuq1Z4l6tt6vGuBcsbomIiOi2ZGtrC+Da07bIMly/F9fvTUPwA2VERER0W1IoFNBoNMjJyQEAODg4QCaTCU7VMlxfa6vVak0eHNEQkiShoqICOTk50Gg0t/SeLG4FcLdzR/ZB4L8Bg5HvGYZD455HuTs/KUxEtydHb0fcPetuOHo7io5CtyEfHx8AMBS4ZJ7a2lpUV1cjIyMDNjaNV05qNBrDPWkoFrcCeKm8cPVHYOfjz8HfPwrZz3UWHYmISBi1vxp9FvURHYNuUzKZDL6+vvDy8kJNTY3oOLc1W1vbRpkFZnErQHltOZwigBD9EShLvOF+/iKyIqNQ7egkOhoRUbPTlmqRGZ8J366+sHO2Ex2HblMKhaJRCqvbRWlpKeLj49G1a1c4OzuLjmOEHygTIK0sDeHjgBfkryE45U88OXEIXNMuio5FRCREQXIBPrv/MxQkc0tEopYiOTkZ999/P5KTk0VHMcHiloiIiIisBotbIiIiIrIaLG6JiIiIyGqwuBXARm6DmlIgX9Kg1tYOJV6+0Ns0fLNiIqKWTG4rh7O/M+S2/CuJqKWwtbWFv7//LT1soalwtwQBwtXhOL0QmL9xK/zbReH8zkGiIxERCePd0Rszr8wUHYOI6qFjx464cuWK6Bh14q/JRERERGQ1WNwKkFySjA6vAvOlYYg48yMm9+sEz+RE0bGIiITITsjGsoBlyE7IFh2FiMyUkJCAgIAAJCQkiI5igsWtALX6Wtg6A+6yItjUaKHOyYS8lk9FIaLbk75Gj9KMUuhr9KKjEJGZampqkJGRYZFPdWNxS0RERERWg8UtEREREVkNFrdEREREZDVY3AoQ5BSE5PXAUv3bSG3VE5s+/h6FQa1FxyIiEsIt3A1j9oyBW7ib6ChEZKbw8HDs2bMH4eHhoqOY4D63AjjaOKLsPHBZ3h3+ah+Ud/MRHYmISBg7ZzuE3BciOgYR1YOzszPuu+8+0THqxJlbAXIqc+A3COinXwHvjBPotWIBnHIyRcciIhKiJKMEv83+DSUZJaKjEJGZMjIyMHv2bGRkZIiOYoLFrQD52nx49wAGyH+Ae+4FxKz/AI75OaJjEREJUZ5djgPvHEB5drnoKERkpuzsbLzzzjvIzra8/alZ3BIRERGR1WBxS0RERERWg8UtEREREVkNFrcCaJQa5J8E9unvQ5FbIE4OiUOlC7fAIaLbk8pdhS4TukDlrhIdhYjM5O7ujgkTJsDd3V10FBPcCkwAXwdfpG0Gtg56Hf5BUbg6t7voSEREwmiCNXj404dFxyCieggODsann34qOkadOHMrQJWuCvZ+gKf+POwrCuFx8SxsqipFxyIiEqKmsgY5Z3JQU1kjOgoRmamyshJnzpxBZaXl1S8sbgW4VHoJkc8Bc+WT0er8bjw1/B64XzovOhYRkRB5SXlY3WE18pLyREchIjMlJSWhQ4cOSEpKEh3FBItbIiIiIrIaLG6JiIiIyGqwuCUiIiIiq8HiVgCZTAapFqiRbCDJZKi1VQIymehYRERiyACFUgHwxyBRiyGTyaBUKiGzwPqFW4EJ0NalLU7MAWZu3An/qCgkHXlUdCQiImF8u/jide3romMQUT106dIFWq1WdIw6ceaWiIiIiKwGi1sBLpVeQtvpwEz9eLQ+9yvGPvkA3FO4FRgR3Z5yk3LxUfRHyE3KFR2FiMyUlJSE6OhobgVG11TpqqDyBkLlabCrLIHP2QTYaC1vE2QiouZQW1mLrONZqK2sFR2FiMxUWVmJ48eP8yEORERERERNicUtEREREVkNFrdEREREZDVY3Arg7+CPS1uANboXcSX4Dny3+FMU+QeLjkVEJIQmVINh3wyDJlQjOgoRmSk0NBTffPMNQkNDRUcxwX1uBVAr1Sg6BpxR9IO/azBKHmRhS0S3L5WrCu2Htxcdg4jqwdXVFcOHDxcdo06cuRUgX5sPrweAnrrP4Z6diDu+XA2H/BzRsYiIhCjLLsOhZYdQll0mOgoRmSk7OxvLli1Ddna26CgmWNwKkFOZA/8HgccUn8M7MxG9l82Fc06m6FhEREKUZpTi1xd+RWlGqegoRGSmjIwMvPDCC8jIyBAdxQSLWyIiIiKyGixuiYiIiMhqsLglIiIiIqvB4lYAJ1snFF8Ajum7oEztjeR7Y6F1UouORUQkhJ2LHSIGRcDOxU50FCIyk4uLCwYNGgQXFxfRUUxwKzABAh0DkbIWWP/Au/BvFYW05feIjkREJIxbazc8sf0J0TGIqB5at26N7du3i45RJ87cClCjr4GNM6CS8mBTXQFVYR7kNTWiYxERCaGr0aE8txy6Gp3oKERkppqaGuTm5qLGAusXFrcCXCi5gI6vAktkjyMi8SdM6x0JzwuJomMREQmRk5CD/3j9BzkJ3O+bqKVISEiAl5cXEhISREcxweKWiIiIiKwGi1siIiIishosbomIiIjIarC4JSIiIiKrwa3ABIhwicDJN4AXPt0G7w7dcWF/LGpUDqJjEREJ4R3ljVnFs2DraCs6ChGZKSoqCsXFxXB0dBQdxQSLWwEUMgX0VUC1TA29jRLVTkrRkYiIhJEr5LBT8wEORC2JQqGAWm2ZD6DisgQB0srSEDYReFo/HcEX9mDE5OFwTbsoOhYRkRD5yfn4MvZL5Cfni45CRGZKTk5GbGwskpOTRUcxweJWgPLacjiHAp3kp+FYlo9Wh/dCWV4mOhYRkRDVpdW4+OtFVJdWi45CRGYqLS3Fr7/+itLSUtFRTLC4JSIiIiKrweKWiIiIiKwGi1siIiIishosbgXwUfkg/Wdgo24Csvw74ddX3kGJj7/oWEREQqgD1ej/YX+oAy3zk9dEZCowMBAffvghAgMDRUcxwa3ABHC1c0XefuDwM0/A3zMCBY9FiI5ERCSMo6cj7pxyp+gYRFQPnp6emDJliugYdeLMrQDF1cVwvROI0v8Il4JLaP/fLbAvLhQdi4hIiMqCSpz68hQqCypFRyEiMxUUFODLL79EQUGB6CgmWNwKcLXiKkKGAk/J34d/WjwGzZkMl6tpomMREQlRdLkI3436DkWXi0RHISIzXb58GaNGjcLly5dFRzEhvLjNyMjAyJEj4e7uDpVKhY4dO+Lvv/82tEuShLlz58LX1xcqlQp9+vQx2TC4oKAAcXFxUKvV0Gg0mDBhAsrKuG8sERER0e1GaHFbWFiIu+++G7a2tvj555+RmJiIpUuXwtXV1dBnyZIl+OCDD7BmzRocOXIEjo6OiI2NRVVVlaFPXFwczpw5g127dmHHjh3Yv38/Jk6cKOKSiIiIiEggoR8oW7x4MQIDA7F+/XrDsdDQUMOfJUnC8uXL8frrr2Pw4MEAgM8//xze3t74/vvv8fjjjyMpKQk7d+7EX3/9hW7dugEAVqxYgYceegj/+c9/4Ofn17wXRURERETCCJ253b59O7p164bhw4fDy8sLXbp0wSeffGJov3TpErKystCnTx/DMRcXF3Tv3h2HDh0CABw6dAgajcZQ2AJAnz59IJfLceTIkTrH1Wq1KCkpMXo1J5WNCuVXgWR9a1Q6uCKjYzfUqBybNQMRkaWwdbRFwF0BsHW0FR2FiMzk6OiIu+66C46Olle/CC1uU1JSsHr1aoSHh+OXX37BpEmT8Pzzz+Ozzz4DAGRlZQEAvL29jc7z9vY2tGVlZcHLy8uo3cbGBm5uboY+/7Ro0SK4uLgYXs29R1uIUwjOrwA+kH+ESxG98cVnP6MgJKxZMxARWQqPNh6YcGgCPNp4iI5CRGZq06YNDh06hDZt2oiOYkJocavX6xEdHY2FCxeiS5cumDhxIp5++mmsWbOmScedPXs2iouLDa/09PQmHY+IiIiImofQ4tbX1xft2rUzOhYZGYm0tGvbYvn4+AAAsrOzjfpkZ2cb2nx8fJCTk2PUXltbi4KCAkOff7Kzs4NarTZ6NaekoiR0WQSsQB+0O7EVs6I94Z10slkzEBFZisxjmXhD9gYyj2WKjkJEZjp27BhkMhmOHTsmOooJocXt3XffjXPnzhkdO3/+PIKDgwFc+3CZj48Pdu/ebWgvKSnBkSNHEBMTAwCIiYlBUVER4uPjDX1+//136PV6dO/evRmugoiIiIgshdDdEmbMmIEePXpg4cKFGDFiBI4ePYqPP/4YH3/8MQBAJpNh+vTpeOuttxAeHo7Q0FDMmTMHfn5+GDJkCIBrM739+vUzLGeoqanB1KlT8fjjj3OnBCIiIqLbjNDi9o477sB3332H2bNn480330RoaCiWL1+OuLg4Q5+XX34Z5eXlmDhxIoqKitCzZ0/s3LkT9vb2hj4bN27E1KlT0bt3b8jlcjz66KP44IMPRFwSEREREQkktLgFgIEDB2LgwIE3bJfJZHjzzTfx5ptv3rCPm5sbNm3a1BTxiIiIiKgFEV7c3o5aObdC4nvA3KWfQtOmF9Z8fwSl3lxCQUS3J892nngu+TmoA5r3w71E1HDt2rVDcnIyAgICREcxweJWADuFHbQ5QKE8BA4qDaqDNKIjEREJY2NvA7cwN9ExiKge7O3tERZmmXv0C90t4XaVUZGBkJHAE/q5CLh8EANfmwSXjFTRsYiIhCi8VIhtI7eh8FKh6ChEZKZLly5h5MiRuHTpkugoJljcClBSXQLX9kAP+UGoi66iw89bYV9SJDoWEZEQVYVVSNiYgKrCKtFRiMhMhYWF2LhxIwoLLe+XUha3RERERGQ1WNwSERERkdVgcUtEREREVoPFrQAe9h7I3A/8oB+GXO82+HPiSyjz8BYdi4hICCdfJ/Sa1wtOvk6ioxCRmXx9fTFv3jz4+vqKjmKiQVuBpaSkoFWrVo2d5bbhae+JrJ+B30Y+C3/fjsh9tqPoSEREwjj7OuO++feJjkFE9eDr64v58+eLjlGnBs3choWF4f7778eXX36Jqip+urW+ymrK4NwOCNP9Aafiqwg9+DuUZaWiYxERCaEt0eLCLxegLdGKjkJEZiopKcEvv/yCkpIS0VFMNKi4PXbsGDp16oSZM2fCx8cHzzzzDI4ePdrY2axWenk6wkYB0xRvIOjSQTw29TG4pqeIjkVEJETBhQJs7LcRBRcKREchIjNduHAB/fr1w4ULF0RHMdGg4rZz5854//33cfXqVaxbtw6ZmZno2bMnOnTogGXLliE3N7excxIRERER/atb+kCZjY0NHnnkEWzZsgWLFy/GhQsX8OKLLyIwMBCjR49GZmZmY+UkIiIiIvpXt1Tc/v3335g8eTJ8fX2xbNkyvPjii7h48SJ27dqFq1evYvDgwY2Vk4iIiIjoXzVot4Rly5Zh/fr1OHfuHB566CF8/vnneOihhyCXX6uVQ0NDsWHDBoSEhDRmVquhVCihLQKyAj1RbeeIwoAQ6JR2omMREQmhsFPAtbUrFHYK0VGIyEx2dnZo3bo17Owsr35pUHG7evVqjB8/HmPHjr3h/mZeXl5Yu3btLYWzVq2dWyNxMfD2xq/gHxmFC9v7i45ERCSMV3svPH/hedExiKge2rdvb5EfJgMaWNwmJyf/ax+lUokxY8Y05O2JiIiIiBqkQWtu169fjy1btpgc37JlCz777LNbDmXtkkuS0XEO8LY0BBGnt+P5B9rC8/wZ0bGIiITIPpWNdz3fRfapbNFRiMhMp06dgqenJ06dOiU6iokGFbeLFi2Ch4eHyXEvLy8sXLjwlkNZu1p9LWwcALWsDDa11XAoyodcVys6FhGREPpaPSryKqCv1YuOQkRmqq2tRV5eHmprLa9+aVBxm5aWhtDQUJPjwcHBSEtLu+VQREREREQN0aDi1svLq85p6JMnT8Ld3f2WQxERERERNUSDitsnnngCzz//PPbs2QOdTgedTofff/8d06ZNw+OPP97YGYmIiIiIzNKg3RIWLFiAy5cvo3fv3rCxufYWer0eo0eP5ppbMwQ7BeP8J8DiN96FKuwefL7+JxQEtxYdi4hICPcId4w/OB7uEfyXP6KWIiIiAgcPHkRERIToKCYaVNwqlUp8/fXXWLBgAU6ePAmVSoWOHTsiODi4sfNZJQcbB5SnAFfkXeDv5I2KKG/RkYiIhFE6KREYEyg6BhHVg5OTE2JiYkTHqNMtPX43IiICw4cPx8CBA1nY1kN2ZTb8hwIDdcvgc+U4Hlg6B87ZV0XHIiISouRKCX6Z+QtKrpSIjkJEZrpy5QpmzpyJK1euiI5iokHFrU6nw9q1a/Hkk0+iT58+eOCBB4xedHMF2gJ43QnEKn6CW95F3LlxDRwKckXHIiISojynHIffO4zynHLRUYjITDk5OXjvvfeQk5MjOoqJBi1LmDZtGjZs2IABAwagQ4cOkMlkjZ2LiIiIiKjeGlTcbt68Gd988w0eeuihxs5DRERERNRgDVqWoFQqERYW1thZiIiIiIhuSYOK2xdeeAHvv/8+JElq7Dy3BY1Sg7x4YLeuD4rcgxE/fBwqNdwCh4huTw4eDug2uRscPBxERyEiM3l4eGDy5Mnw8PAQHcVEg5Yl/Pnnn9izZw9+/vlntG/fHra2tkbt27Zta5Rw1srXwRfpW4Hvh86Cf2AUrs6+Q3QkIiJhXIJcMGDlANExiKgegoKCsHLlStEx6tSg4laj0WDo0KGNneW2UVlbCVUQ4KM/DfuKALikXkF+SDhqVZy1IKLbT01FDfLO5sGjrQdsHWz//QQiEq6iogJnz55F27Zt4eBgWfVLg4rb9evXN3aO28rlsstoOwl4TT4d28+vxsPjJ2H9xt+QHRklOhoRUbPLO5uHj7t+jInxE+Eb7Ss6DhGZ4ezZs+jatSvi4+MRHR0tOo6RBj/Eoba2Fr/99hs++ugjlJaWAgCuXr2KsrKyRgtHRERERFQfDZq5TU1NRb9+/ZCWlgatVosHH3wQzs7OWLx4MbRaLdasWdPYOYmIiIiI/lWDZm6nTZuGbt26obCwECqVynB86NCh2L17d6OFIyIiIiKqjwbN3P7xxx84ePAglEql0fGQkBBkZGQ0SjBrJpfJoa8GKiR76OUKaB2dIMkbvEKEiKhFk8llUDorIZPzaZdELYVcLoezszPkFli/NKi41ev10Ol0JsevXLkCZ2fnWw5l7dq4tMHJecArG3fAv1MUzv7BnSeI6Pbl09kHs0tmi45BRPXQuXNnlJSUiI5RpwaV23379sXy5csNX8tkMpSVlWHevHl8JC8RERERCdOg4nbp0qU4cOAA2rVrh6qqKjz55JOGJQmLFy9u7IxWJ6U0BZEvAK/oR6P12Z2YMKwn3FPOiY5FRCREbmIuVrVfhdzEXNFRiMhMiYmJaN++PRITE0VHMdGgZQkBAQE4efIkNm/ejFOnTqGsrAwTJkxAXFyc0QfMqG5anRb2HkCA/CrsqsrgmXIONtoq0bGIiISorapFbmIuaqtqRUchIjNVVVUhMTERVVWWV780qLgFABsbG4wcObIxsxARERER3ZIGFbeff/75TdtHjx7doDBERERERLeiQcXttGnTjL6uqalBRUUFlEolHBwcWNwSERERkRAN+kBZYWGh0ausrAznzp1Dz5498dVXXzV2RqsT4BiAlK+AlbpXcSWkO7a+9wWKAkJExyIiEsK1lSse/+FxuLZyFR2FiMzUqlUr/PDDD2jVqpXoKCYavOb2n8LDw/HOO+9g5MiROHv2bGO9rVVytnVG8SngrOIB+GsCUdIrUHQkIiJh7DX2aPNwG9ExiKgeNBoNHn74YdEx6tSoj5WwsbHB1atXG/MtrVJeVR68HwR66dbCM+sM7lq3HI552aJjEREJUZZVhj8W/YGyrDLRUYjITFlZWVi0aBGysrJERzHRoJnb7du3G30tSRIyMzPx4Ycf4u67726UYNYstyoXfg8AwxRfYXvWvbjvw7dxKeZ+lHt4i45GRNTsSq+W4vdXf0dYbBicfJxExyEiM1y9ehWvvvoqYmNj4ePjIzqOkQYVt0OGDDH6WiaTwdPTEw888ACWLl3aGLmIiIiIiOqtQcWtXq9v7BxERERERLesUdfcEhERERGJ1KCZ25kzZ5rdd9myZQ0Zwqo52zqj6CxwNOAOlLj44GyfQahydhEdi4hICHuNPdoNawd7jb3oKERkJo1Gg2HDhkGj0YiOYqJBxe3x48dx/Phx1NTUoE2ba9u3nD9/HgqFAtHR0YZ+MpmscVJamQDHAFz6DPii7yL4h0bhypKeoiMREQnj2soVw7cMFx2DiOqhVatW2LJli+gYdWpQcTto0CA4Ozvjs88+g6vrtU23CwsLMW7cONxzzz144YUXGjWktanR18DWFXCWrsKmujVUhSUod/OA3lYpOhoRUbPTVetQnlMORy9HKJQK0XGIyAzV1dXIycmBl5cXlErLql8atOZ26dKlWLRokaGwBQBXV1e89dZb3C3BDBdKLqDDy8BC2WhEJO7ElP5R8LyQJDoWEZEQOadz8F7ge8g5nSM6ChGZ6fTp0wgMDMTp06dFRzHRoOK2pKQEubm5Jsdzc3NRWlp6y6GIiIiIiBqiQcXt0KFDMW7cOGzbtg1XrlzBlStX8O2332LChAl45JFHGjsjEREREZFZGrTmds2aNXjxxRfx5JNPoqam5tob2dhgwoQJePfddxs1IBERERGRuRpU3Do4OGDVqlV49913cfHiRQBA69at4ejo2KjhiIiIiIjqo0HF7XWZmZnIzMzEvffeC5VKBUmSuP2XGdq4tMGJOcD0DTvg07Erzh0eBJ2NrehYRERC+HT2wWtVr0Fhy50SiFqKzp07o6qqCra2lle/NGjNbX5+Pnr37o2IiAg89NBDyMzMBABMmDCB24CZQS6TQ6oFdDJ7QGEDndIOkPNhcUR0e5LJZbCxs4FMzskRopZCLpfDzs4OcgusXxqUaMaMGbC1tUVaWhocHBwMxx977DHs3Lmz0cJZq9SyVIRPBibrpyAkeQ+efHowXFMvio5FRCRE/vl8bLhvA/LP54uOQkRmOn/+PO677z6cP39edBQTDVqW8Ouvv+KXX35BQECA0fHw8HCkpqY2SjBrVlFbAadAIFJ+Dsnl+QiKPwhlRZnoWEREQlSXVSN1Xyqqy6pFRyEiM5WVlWHfvn0oK7O8+qVBM7fl5eVGM7bXFRQUwM7O7pZDERERERE1RIOK23vuuQeff/654WuZTAa9Xo8lS5bg/vvvb7RwRERERET10aBlCUuWLEHv3r3x999/o7q6Gi+//DLOnDmDgoICHDhwoLEzEhERERGZpUEztx06dMD58+fRs2dPDB48GOXl5XjkkUdw/PhxtG7durEzWh0flQ/SdgCf6SchM6AzfpqzDCU+Af9+IhGRFXIJcsGgTwbBJchFdBQiMlNQUBA++eQTBAUFiY5iot4ztzU1NejXrx/WrFmD1157rSkyWT1XO1fkHwD+nvwo/D3CUDg0THQkIiJhHDwcEP1UtOgYRFQPHh4eeOqpp0THqFO9Z25tbW1x6tSppshy2yjUFsL9bqCb/lu45l1Ap+++gKqQW+AQ0e2pIq8Cxz49hoq8CtFRiMhMeXl5+PTTT5GXlyc6iokGLUsYOXIk1q5d29hZbhtZlVkIGgiMka+G75UTeGjBTKizroiORUQkRHFaMX58+kcUpxWLjkJEZkpLS8PTTz+NtLQ00VFMNOgDZbW1tVi3bh1+++03dO3aFY6Ojkbty5Yta5RwRERERET1Ua/iNiUlBSEhITh9+jSio6+tj/rnkylkMj4+kYiIiIjEqFdxGx4ejszMTOzZswfAtcftfvDBB/D29m6ScERERERE9VGvNbeSJBl9/fPPP6O8vLxRA90OHGwcUJYOJOnboMLRHWlde6DawUl0LCIiIZROSgT3CobSSSk6ChGZycnJCb169YKTk+XVLw1ac3vdP4tdMk+wUzCSVwGr7l4J//AoXP6ET3UjotuXe4Q7xu4dKzoGEdVDREQE9u7dKzpGneo1cyuTyUzW1HKNbf3pJT1kNoBCqgJ0tVBUawG9XnQsIiIhJL2EWm0tJD0nTIhaCr1eD61WC70F1i/1mrmVJAljx46FnZ0dAKCqqgrPPvusyW4J27Zta7yEVuhc8Tl0XgAslw3E9oTVeHj8JKzf+BuyI6NERyMianZZJ7LwcdePMTF+InyjfUXHISIznDhxAl27dkV8fLxhkwFLUa/idsyYMUZfjxw5slHDEBERERHdinoVt+vXr2+qHEREREREt6xBTygjIiIiIrJELG6JiIiIyGrc0lZg1DBh6jCcXgK8uuJzuLW7Byt/PolyNw/RsYiIhPDq4IUZ6TPg6OX4752JyCJ06NAB6enp8PLyEh3FhMXM3L7zzjuQyWSYPn264VhVVRWmTJkCd3d3ODk54dFHH0V2drbReWlpaRgwYAAcHBzg5eWFl156CbW1tc2cvn5s5baoKQRKZX6oVTqh1NsPeltuXk5EtyeFUgF1gBoKpUJ0FCIyk1KpREBAAJRKy6tfLKK4/euvv/DRRx+hU6dORsdnzJiBH3/8EVu2bMG+fftw9epVPPLII4Z2nU6HAQMGoLq6GgcPHsRnn32GDRs2YO7cuc19CfVypfwKQscAo/SzEXDpTwx5eTxcrlwWHYuISIjClEJsGb4FhSmFoqMQkZlSUlIwfPhwpKSkiI5iQnhxW1ZWhri4OHzyySdwdXU1HC8uLsbatWuxbNkyPPDAA+jatSvWr1+PgwcP4vDhwwCAX3/9FYmJifjyyy/RuXNn9O/fHwsWLMDKlStRXV0t6pL+VWlNKTRtgTvlf0FdnIW2v/0I+9Ji0bGIiISoKqpC4tZEVBVViY5CRGYqKirC1q1bUVRUJDqKCeHF7ZQpUzBgwAD06dPH6Hh8fDxqamqMjrdt2xZBQUE4dOgQAODQoUPo2LEjvL29DX1iY2NRUlKCM2fO3HBMrVaLkpISoxcRERERtXxCP1C2efNmHDt2DH/99ZdJW1ZWFpRKJTQajdFxb29vZGVlGfr8b2F7vf16240sWrQIb7zxxi2mJyIiIiJLI2zmNj09HdOmTcPGjRthb2/frGPPnj0bxcXFhld6enqzjk9ERERETUNYcRsfH4+cnBxER0fDxsYGNjY22LdvHz744APY2NjA29sb1dXVJms5srOz4ePjAwDw8fEx2T3h+tfX+9TFzs4OarXa6NWcPO09cfV3YKvuCeT6RGLv1NdQ5nnjvERE1szZzxkPLHwAzn7OoqMQkZn8/PywcOFC+Pn5iY5iQlhx27t3byQkJODEiROGV7du3RAXF2f4s62tLXbv3m0459y5c0hLS0NMTAwAICYmBgkJCcjJyTH02bVrF9RqNdq1a9fs12QuD3sPZO8C9ikmINenPQ6Pn45yD+9/P5GIyAo5+Tjhntn3wMnHSXQUIjKTj48PZs+efdPJRFGErbl1dnZGhw4djI45OjrC3d3dcHzChAmYOXMm3NzcoFar8dxzzyEmJgZ33XUXAKBv375o164dRo0ahSVLliArKwuvv/46pkyZAjs7u2a/JnOV1pTCpRPQVvc7ZEVu8Dp5BunRMdA6u4iORkTU7KqKqpC6PxXB9wbDXtO8y9SIqGGKioqwf/9+3HvvvSafjxJN+G4JN/Pee+9h4MCBePTRR3HvvffCx8cH27ZtM7QrFArs2LEDCoUCMTExGDlyJEaPHo0333xTYOp/d6X8Clo9AUxRLETA5SMYNmMUNNznlohuU4Uphdg8eDP3uSVqQVJSUjB48GCL3OfWoh6/u3fvXqOv7e3tsXLlSqxcufKG5wQHB+Onn35q4mRERERE1BJY9MwtEREREVF9sLglIiIiIqvB4lYAO4UdqvKAK3o/aO2dkNuqDWrt+CEKIro92djbwLOdJ2zsLWqlHBHdhL29Pdq1a9fszyowB3+SCNDKuRWSlgKLN34O/7ZRuLi1n+hIRETCeLbzxOQzk0XHIKJ6aNeuHc6cOSM6Rp04c0tEREREVoPFrQDnis8h6g1gsTQQbU99hxn3hMLrXILoWEREQmSdyMIi9SJkncgSHYWIzHTixAmo1WqcOHFCdBQTLG4F0Et6yJWAg6wKcr0OduVlkOn1omMREQkh6SVUl1ZD0kuioxCRmfR6PUpLS6G3wPqFxS0RERERWQ0Wt0RERERkNVjcEhEREZHVYHErQIhTCM6uBt7WL0dKxP1Yv/E35IeEi45FRCSER1sPTIyfCI+2HqKjEJGZ2rZti/j4eLRt21Z0FBPc51YAlY0KlWlAlrwD/B3cURXpLjoSEZEwtg628I32FR2DiOrBwcEB0dHRomPUiTO3AmRWZCJwGDBE9w780v/Cg4tehjrziuhYRERCFKcV479T/ovitGLRUYjITGlpaZgyZQrS0tJERzHB4laAouoieHQFeit+gyY/FV23rIeqKF90LCIiISryKvD3qr9RkVchOgoRmSkvLw+rVq1CXl6e6CgmWNwSERERkdVgcUtEREREVoPFLRERERFZDRa3ArjZuSHnKPCL7iEUeLTG0bhnUeHmKToWEZEQjl6OuGvGXXD0chQdhYjM5OXlhRkzZsDLy0t0FBPcCkwAb5U3Mr4DdgybCf+AKGS90EV0JCIiYdQBasQuixUdg4jqISAgAMuWLRMdo06cuRWgorYCjq2AAP1xOJRlw+/kX7CtKBMdi4hIiOqyaqQfSkd1WbXoKERkprKyMhw6dAhlZZZXv7C4FSC1LBURTwOvyF9CyIU/MHrcQ3BLvSg6FhGREPnn87Guxzrkn+eWiEQtxfnz59GjRw+cP39edBQTLG6JiIiIyGqwuCUiIiIiq8HiloiIiIisBotbAWzkNqitAEokJ9TaKFGhcYdewY0riOj2JLeRw8HDAXIb/pVE1FLY2NjAw8MDNjaWV79YXqLbQLg6HAkLgNc2fg//DlE4//vDoiMREQnj3ckbL+W+JDoGEdVDp06dkJubKzpGnfhrMhERERFZDRa3AlwsvYh2rwCvSU8gLOlnPPPwHfC4eFZ0LCIiIXLO5OCDsA+QcyZHdBQiMtOZM2cQFhaGM2fOiI5igsWtANW6athpAB9ZLpTacrheuQxFtVZ0LCIiIXRaHQovFkKn1YmOQkRm0mq1uHjxIrRay6tfWNwSERERkdVgcUtEREREVoPFLRERERFZDRa3AgQ6BuLCF8D7unlIC+2Brz/8GoWBrUTHIiISwi3MDXE74+AW5iY6ChGZKSwsDDt37kRYWJjoKCa4z60ATrZOKE0ELijugb+LH8p6+ImOREQkjJ3aDmGxlvcXJBHdmFqtRmxsrOgYdeLMrQC5Vbnw6Q/00a+BZ2YCeq5ZAsfcLNGxiIiEKM0sxd75e1GaWSo6ChGZKTMzE/Pnz0dmZqboKCZY3AqQV5UH33uBwfKt8Mw+h54fvwunvGzRsYiIhCjLLMO+N/ahLLNMdBQiMlNmZibeeOMNFrdERERERE2JxS0RERERWQ0Wt0RERERkNVjcCqBWqlF4Bjio74ESjR9O9x+GKrVGdCwiIiHsXe3RMa4j7F3tRUchIjO5uroiLi4Orq6uoqOY4FZgAvg7+OPyl8BX/d+Ef0gUrrzdQ3QkIiJhXENd8ciXj4iOQUT1EBoaii+//FJ0jDpx5lYArU4LOy/AVX8ZysoiaNJSoNBWiY5FRCREbVUtCi4UoLaqVnQUIjJTVVUVLly4gKoqy6tfWNwKkFKagnYzgDflTyHs3G94dkh3eKScEx2LiEiI3MRcrAhfgdzEXNFRiMhMiYmJCA8PR2JiougoJljcEhEREZHVYHFLRERERFaDxS0RERERWQ0Wt0RERERkNbgVmACRmkgcnw08t/E3+HeOQuKxYaIjEREJ4xvti3nSPNExiKgeoqOjIUmS6Bh14swtEREREVkNFrcCXC67jIjngOf1zyD0/G6MGtMfbpcviI5FRCRE3rk8rI1Zi7xzeaKjEJGZzp07h5iYGJw7Z3lbmbK4FaCythKOfkC4/CJUFYXwT/gbtpXlomMREQlRU16DK4evoKa8RnQUIjJTeXk5Dh8+jPJyy6tfWNwSERERkdVgcUtEREREVoPFLRERERFZDRa3Avg5+OHyd8Cn+mnICOqKHxesQrFfkOhYRERCaEI0GPrFUGhCNKKjEJGZQkJC8MUXXyAkJER0FBPc51YAF6ULCo8CJ+WD4O8WiuIBoaIjEREJo3JTodPITqJjEFE9uLm5YeTIkaJj1IkztwIUagvhcS9wl+4ruOWeR/TXa6Eq5BY4RHR7Ks8tx9GVR1Gea3mfuiaiuuXm5mLlypXIzc0VHcUEi1sBsiqzENgfiFOshU/GKfRdPAvqrAzRsYiIhChJL8HPU39GSXqJ6ChEZKb09HRMnToV6enpoqOYYHFLRERERFaDxS0RERERWQ0Wt0RERERkNVjcCuBo44jSS8ApfQeUO7kj5a77UO3oJDoWEZEQSmclWvdtDaWzUnQUIjKTs7Mz+vbtC2dnZ9FRTHArMAGCnIJw4WPgk17L4R8WhdRV94uOREQkjHu4O0b+YplbChFR3cLDw/HLL7+IjlEnztwKoJN0kNsDSqkE8tpqKMtKIdPpRMciIhJCr9NDW6KFXqcXHYWIzKTT6VBSUgKdBdYvLG4FOF98HlHzgKWyR9D29HbMvLcVvM6fFh2LiEiI7JPZeMflHWSfzBYdhYjMdPLkSbi4uODkyZOio5hgcUtEREREVoPFLRERERFZDRa3RERERGQ1WNwSERERkdXgVmAChKnDkLAQeHn1Zni1i8H7u5OgdXIRHYuISAivjl54MedF2GvsRUchIjN17NgROTk50Gg0oqOYYHErgK3cFrWlQKXMA7VKB9QqHURHIiISRmGrgKOno+gYRFQPtra28PT0FB2jTlyWIEB6eTpaTQDG6V9CUMofeHT6SGjSL4mORUQkRMHFAnz18FcouFggOgoRmenixYt4+OGHcfHiRdFRTLC4FaCspgwuYUC0/DicSrIRvv8X2JWViI5FRCSEtliL8z+eh7ZYKzoKEZmpuLgYP/74I4qLi0VHMcHiloiIiIisBotbIiIiIrIaLG6JiIiIyGqwuBXAS+WFjF3A17rRyPZth90z30Spl6/oWEREQjj7O6Pv0r5w9ncWHYWIzOTv74+lS5fC399fdBQT3ApMAHc7d+T8Dvw5YTT8vdshf2Q70ZGIiIRx8nZCzMwY0TGIqB68vb0xc+ZM0THqxJlbAUqqS6CJBtrrdkJdmIo2u36AXUmR6FhEREJUFlbizJYzqCysFB2FiMxUWFiILVu2oLCwUHQUEyxuBcioyEDocOBZxX8QkPoXhr7yFDQZqaJjEREJUXSpCFtHbEXRpSLRUYjITJcuXcKIESNw6ZLl7dPP4paIiIiIrAaLWyIiIiKyGkKL20WLFuGOO+6As7MzvLy8MGTIEJw7d86oT1VVFaZMmQJ3d3c4OTnh0UcfRXZ2tlGftLQ0DBgwAA4ODvDy8sJLL72E2tra5rwUIiIiIrIAQovbffv2YcqUKTh8+DB27dqFmpoa9O3bF+Xl5YY+M2bMwI8//ogtW7Zg3759uHr1Kh555BFDu06nw4ABA1BdXY2DBw/is88+w4YNGzB37lwRl2QWe4U9KrOBS/ogaFVqZLXtiFo7lehYRERC2Khs4NPFBzYqbuBD1FKoVCp06dIFKpXl1S8ySZIk0SGuy83NhZeXF/bt24d7770XxcXF8PT0xKZNmzBs2DAAwNmzZxEZGYlDhw7hrrvuws8//4yBAwfi6tWr8Pb2BgCsWbMGr7zyCnJzc6FUKv913JKSEri4uKC4uBhqtbpJrxEAjh07hq5du2Lqxt/gHxnV5ONlJJ3Eh3F9EB8fj+jo6CYfj4iIiKixmVuvWdSa2+LiYgCAm5sbACA+Ph41NTXo06ePoU/btm0RFBSEQ4cOAQAOHTqEjh07GgpbAIiNjUVJSQnOnDlT5zharRYlJSVGLyIiIiJq+SymuNXr9Zg+fTruvvtudOjQAQCQlZUFpVIJjUZj1Nfb2xtZWVmGPv9b2F5vv95Wl0WLFsHFxcXwCgwMbOSrubmzxWfReQGwTOqHyJPf4sXu/vA+e6pZMxARWYrM45l4y+4tZB7PFB2FiMx0/Phx2NnZ4fjx46KjmLCY4nbKlCk4ffo0Nm/e3ORjzZ49G8XFxYZXenp6k4/5vyRJgswGsJXVQiZJsKmpBixndQgRUfOSAF21DuCPQaIWQ5IkVFdXw4JWtxpYxOr9qVOnYseOHdi/fz8CAgIMx318fFBdXY2ioiKj2dvs7Gz4+PgY+hw9etTo/a7vpnC9zz/Z2dnBzs6uka+CiIiIiEQTOnMrSRKmTp2K7777Dr///jtCQ0ON2rt27QpbW1vs3r3bcOzcuXNIS0tDTMy155DHxMQgISEBOTk5hj67du2CWq1Gu3btmudCiIiIiMgiCJ25nTJlCjZt2oQffvgBzs7OhjWyLi4uUKlUcHFxwYQJEzBz5ky4ublBrVbjueeeQ0xMDO666y4AQN++fdGuXTuMGjUKS5YsQVZWFl5//XVMmTKFs7NEREREtxmhxe3q1asBAPfdd5/R8fXr12Ps2LEAgPfeew9yuRyPPvootFotYmNjsWrVKkNfhUKBHTt2YNKkSYiJiYGjoyPGjBmDN998s7kuo95CnUORtAJ4c/EqqCN649Mtf6DIP1h0LCIiITwiPTDp9CS4tnIVHYWIzBQZGYnTp0+jVatWoqOYEFrcmrMI2d7eHitXrsTKlStv2Cc4OBg//fRTY0ZrUvYKe1RdBXLlEVA6uKKqNX+gE9Hty1ZlC6/2XqJjEFE9qFQqtG/fXnSMOlnEB8puN5kVmQh6HBimfwupaS8iasNXOPDUCyjxa94tyYiILEFRahH2L9iPe+fcC02wRnSc21paWhry8vKabTwPDw8EBQU123jUeFJTU7FgwQLMmTMHwcGW9a/PLG4FKKougnsU0Eu+F9sLHkPU9xtxbPg4FrdEdFuqzK/E8bXHccfkO1jcCpSWloa2kZGorKhotjFVDg44m5TEArcFys/Px9q1azF58mQWt0RERGR58vLyUFlRgRFvrYZXaHiTj5dzKRnfvD4JeXl5LG6pUbG4JSIiIgOv0HD4R0aJjkHUYBbzhDIiIiIiolvF4lYAdzt3ZB8E/qsfjHzPMBwa9zzK3flJYSK6PTl6O+LuWXfD0dtRdBQiMpO3tzdmzZoFb29v0VFMcFmCAF4qL1z9Edj5+HPw949C9nOdRUciIhJG7a9Gn0V9RMcgonrw9/fHokWLRMeoE2duBSivLYdTBBCiPwLHkiwE/X0AyvIy0bGIiITQlmpxee9laEu1oqMQkZlKS0uxd+9elJaWio5igsWtAGllaQgfB7wgfw3BKX/iyYlD4Jp2UXQsIiIhCpIL8Nn9n6EguUB0FCIyU3JyMu6//34kJyeLjmKCxS0RERERWQ0Wt0RERERkNVjcEhEREZHVYHErgI3cBjWlQL6kQa2tHUq8fKG3sRUdi4hICLmtHM7+zpDb8q8kopbC1tYW/v7+sLW1vPqFW4EJEK4Ox+mFwPyNW+HfLgrndw4SHYmISBjvjt6YeWWm6BhEVA8dO3bElStXRMeoE39NJiIiIiKrweJWgOSSZHR4FZgvDUPEmR8xuV8neCYnio5FRCREdkI2lgUsQ3ZCtugoRGSmhIQEBAQEICEhQXQUEyxuBajV18LWGXCXFcGmRgt1TibktTWiYxERCaGv0aM0oxT6Gr3oKERkppqaGmRkZKCmxvLqFxa3RERERGQ1WNwSERERkdVgcUtEREREVoPFrQBBTkFIXg8s1b+N1FY9senj71EY1Fp0LCIiIdzC3TBmzxi4hbuJjkJEZgoPD8eePXsQHh4uOooJ7nMrgKONI8rOA5fl3eGv9kF5Nx/RkYiIhLFztkPIfSGiYxBRPTg7O+O+++4THaNOnLkVIKcyB36DgH76FfDOOIFeKxbAKSdTdCwiIiFKMkrw2+zfUJJRIjoKEZkpIyMDs2fPRkZGhugoJljcCpCvzYd3D2CA/Ae4515AzPoP4JifIzoWEZEQ5dnlOPDOAZRnl4uOQkRmys7OxjvvvIPsbMvbn5rFLRERERFZDRa3RERERGQ1WNwSERERkdVgcSuARqlB/klgn/4+FLkF4uSQOFS6cAscIro9qdxV6DKhC1TuKtFRiMhM7u7umDBhAtzd3UVHMcGtwATwdfBF2mZg66DX4R8Uhatzu4uOREQkjCZYg4c/fVh0DCKqh+DgYHz66aeiY9SJM7cCVOmqYO8HeOrPw76iEB4Xz8KmqlJ0LCIiIWoqa5BzJgc1lTWioxCRmSorK3HmzBlUVlpe/cLiVoBLpZcQ+RwwVz4Zrc7vxlPD74H7pfOiYxERCZGXlIfVHVYjLylPdBQiMlNSUhI6dOiApKQk0VFMsLglIiIiIqvB4paIiIiIrAaLWyIiIiKyGixuBZDJZJBqgRrJBpJMhlpbJSCTiY5FRCSGDFAoFQB/DBK1GDKZDEqlEjILrF+4FZgAbV3a4sQcYObGnfCPikLSkUdFRyIiEsa3iy9e174uOgYR1UOXLl2g1WpFx6gTZ26JiIiIyGqwuBXgUukltJ0OzNSPR+tzv2Lskw/APYVbgRHR7Sk3KRcfRX+E3KRc0VGIyExJSUmIjo7mVmB0TZWuCipvIFSeBrvKEvicTYCN1vI2QSYiag61lbXIOp6F2spa0VGIyEyVlZU4fvw4H+JARERERNSUWNwSERERkdVgcUtEREREVoPFrQD+Dv64tAVYo3sRV4LvwHeLP0WRf7DoWEREQmhCNRj2zTBoQjWioxCRmUJDQ/HNN98gNDRUdBQT3OdWALVSjaJjwBlFP/i7BqPkQRa2RHT7Urmq0H54e9ExiKgeXF1dMXz4cNEx6sSZWwHytfnwegDoqfsc7tmJuOPL1XDIzxEdi4hIiLLsMhxadghl2WWioxCRmbKzs7Fs2TJkZ2eLjmKCxa0AOZU58H8QeEzxObwzE9F72Vw452SKjkVEJERpRil+feFXlGaUio5CRGbKyMjACy+8gIyMDNFRTLC4JSIiIiKrweKWiIiIiKwGi1siIiIishosbgVwsnVC8QXgmL4LytTeSL43FlontehYRERC2LnYIWJQBOxc7ERHISIzubi4YNCgQXBxcREdxQS3AhMg0DEQKWuB9Q+8C/9WUUhbfo/oSEREwri1dsMT258QHYOI6qF169bYvn276Bh14sytADX6Gtg4AyopDzbVFVAV5kFeUyM6FhGRELoaHcpzy6Gr0YmOQkRmqqmpQW5uLmossH5hcSvAhZIL6PgqsET2OCISf8K03pHwvJAoOhYRkRA5CTn4j9d/kJPA/b6JWoqEhAR4eXkhISFBdBQTLG6JiIiIyGqwuCUiIiIiq8HiloiIiIisBotbIiIiIrIa3ApMgAiXCJx8A3jh023w7tAdF/bHokblIDoWEZEQ3lHemFU8C7aOtqKjWJS0tDTk5eU123hJSUnNNha1fFFRUSguLoajo6PoKCZY3AqgkCmgrwKqZWrobZSodlKKjkREJIxcIYedmg9w+F9paWloGxmJyooK0VGI6qRQKKBWW+YDqFjcCpBWloawicDT+uk4c2Euui9bhV2z3kFhUGvR0YiIml1+cj5+nvoz+n/YH+7h7qLjWIS8vDxUVlRgxFur4RUa3ixjnjuwG7tWLWqWsajlS05OxtSpU/Hhhx8iPLx5/jdqLha3ApTXlsM5FOgkP43LZflodXgvlOVlomMREQlRXVqNi79eRHVptegoFscrNBz+kVHNMlbOpeRmGYesQ2lpKX799VeUlpaKjmKCxS0RERGRFWjOddqWvEabxS0RERFRCydqnXZmZmazjmcOFrdERERELVxzr9O+vka7qKioyceqLxa3AviofJD+M7DxqQko9++EX195ByU+/qJjEREJoQ5Uo/+H/aEOtMxPXhO1JM21TruypAgA4OPj0+Rj1ReLWwFc7VyRtx84/MwT8PeMQMFjEaIjEREJ4+jpiDun3Ck6BhHVg0qtAQC4urqKDVIHPqFMgOLqYrjeCUTpf4RLwSW0/+8W2BcXio5FRCREZUElTn15CpUFlaKjEJGZqspKAADFxcWCk5hicSvA1YqrCBkKPCV/H/5p8Rg0ZzJcrqaJjkVEJETR5SJ8N+o7FF0uEh2FiMxUkpsNALh69argJKZY3BIRERGR1WBxS0RERERWg8UtEREREVkNFrcCqGxUKL8KJOtbo9LBFRkdu6FG5Sg6FhGRELaOtgi4KwC2jraioxCRmWzt7AEAKpVKcBJT3ApMgBCnEJxfAXyw8SP4R0Th0me9RUciIhLGo40HJhyaIDoGEdWDq18gACAkJERskDpw5paIiIiIrAaLWwGSipLQZRGwAn3Q7sRWzIr2hHfSSdGxiIiEyDyWiTdkbyDzmOU9o56I6pZzKRkAkJSUJDiJKS5LIKuQlpaGvLy8Zh3Tw8MDQUFBzTomERER3RyLW2oSzVlsZmZmYtjw4aiqbN6nG9nZ2+PbrVvh6+vbLONptVrY2dk1y1gixrP2XxZE/ALGe0hEtyMWt9To0tLS0DYyEpUVFc067oi3VsMrNLxZxrp0/Ah+WjYHAwcObJbxAEAml0PS6612PJWDA84mJVllcSTq/xO8h9QSNOc/azf3L2DN+UutJS4PEIXFLTW6vLw8VFZUNFuxee7AbuxatQheoeHwj4xq8vGAa2uNJL2+2a/RWsfLuZSMb16fhD/++AORkZFNPh7QvLOaSUlJzfr/CUDcPczLy2u24qG5Z8M5M924SvOyIZPLMXLkyGYbszl/ARP1Sy2xuBWilXMrJL4HzF36KTRtemHN90dQ6u0nOlaja65i8/qidhGa+xqtdTwRf8k196wm0HzfT6D57+F1DZk90ml1eOD7B3BFe8XsD5WJWI7UnEuRbodZuMrSkmadJGjuX6Kb+5fa67/QNhc3/2AAQKtWrZptTHOxuBXATmEHbQ5QKA+Bg0qD6iCN6EhEQjX3X3KiZsKtmYhfUIDmW44kYinS7cKaf4kGrHeix0apBIBmXddvLha3AmRUZCBkJPCEfi5SLr+Czp98gT8mz0Lx//8tqKk010zA7TDjQE3DWmemRf7rQnO5lV9Q5DlVcNiaiophwdB72Zt1TnMvRxK1FIkaj6hfoq1VSc61f2XJyMhAdHS04DTGrKa4XblyJd59911kZWUhKioKK1aswJ133ik6Vp1Kqkvg2h7oIT+IvKKr6PDzVvw18tkmK25F/bZKRLefhhSbeuSi5sDfcJ58P+SRnmadI+oXBv5C1PLxHjaOqvIyAEBJSYngJKasorj9+uuvMXPmTKxZswbdu3fH8uXLERsbi3PnzsHLy0t0POH42yoRERHdLqyiuF22bBmefvppjBs3DgCwZs0a/Pe//8W6deswa9YsweksB39bJSIiImvX4ovb6upqxMfHY/bs2YZjcrkcffr0waFDh+o8R6vVQqvVGr4uLi4G0HxT6xXlFdBVASUVQEZqMkoApCWdwtWK8iYZL/fytWIzI+kUqptojNtpPBFjcjyOZ+lj3sp48tRyaFCFrKQT0Fc4Nvl4DWHt44kYk+NZx3gVFRXNVj9dH0eSpJt3lFq4jIwMCYB08OBBo+MvvfSSdOedd9Z5zrx58yQAfPHFF1988cUXX3y1sFd6evpNa8MWP3PbELNnz8bMmTMNX+v1ehQUFMDd3R0ymazB71tSUoLAwECkp6dDrVY3RlRqJLw3lov3xnLx3lgu3hvLxXvTdCRJQmlpKfz8bv5sgBZf3Hp4eEChUCA7O9voeHZ2Nnx8fOo8x87OzmRfNo1G02iZ1Go1/wdtoXhvLBfvjeXivbFcvDeWi/emabi4uPxrH3kz5GhSSqUSXbt2xe7duw3H9Ho9du/ejZiYGIHJiIiIiKi5tfiZWwCYOXMmxowZg27duuHOO+/E8uXLUV5ebtg9gYiIiIhuD1ZR3D722GPIzc3F3LlzkZWVhc6dO2Pnzp3w9vZu1hx2dnaYN2+eRT6K7nbHe2O5eG8sF++N5eK9sVy8N+LJJOnf9lMgIiIiImoZWvyaWyIiIiKi61jcEhEREZHVYHFLRERERFaDxS0RERERWQ0Wt41k5cqVCAkJgb29Pbp3746jR4+KjmRVFi1ahDvuuAPOzs7w8vLCkCFDcO7cOaM+VVVVmDJlCtzd3eHk5IRHH33U5OEeaWlpGDBgABwcHODl5YWXXnoJtbW1Rn327t2L6Oho2NnZISwsDBs2bGjqy7Mq77zzDmQyGaZPn244xnsjVkZGBkaOHAl3d3eoVCp07NgRf//9t6FdkiTMnTsXvr6+UKlU6NOnD5KTk43eo6CgAHFxcVCr1dBoNJgwYQLKysqM+pw6dQr33HMP7O3tERgYiCVLljTL9bVUOp0Oc+bMQWhoKFQqFVq3bo0FCxbgfz/nzXvTPPbv349BgwbBz88PMpkM33//vVF7c96HLVu2oG3btrC3t0fHjh3x008/Nfr1Wr2bPpyXzLJ582ZJqVRK69atk86cOSM9/fTTkkajkbKzs0VHsxqxsbHS+vXrpdOnT0snTpyQHnroISkoKEgqKysz9Hn22WelwMBAaffu3dLff/8t3XXXXVKPHj0M7bW1tVKHDh2kPn36SMePH5d++uknycPDQ5o9e7ahT0pKiuTg4CDNnDlTSkxMlFasWCEpFApp586dzXq9LdXRo0elkJAQqVOnTtK0adMMx3lvxCkoKJCCg4OlsWPHSkeOHJFSUlKkX375Rbpw4YKhzzvvvCO5uLhI33//vXTy5Enp4YcflkJDQ6XKykpDn379+klRUVHS4cOHpT/++EMKCwuTnnjiCUN7cXGx5O3tLcXFxUmnT5+WvvrqK0mlUkkfffRRs15vS/L2229L7u7u0o4dO6RLly5JW7ZskZycnKT333/f0If3pnn89NNP0muvvSZt27ZNAiB99913Ru3NdR8OHDggKRQKacmSJVJiYqL0+uuvS7a2tlJCQkKTfw+sCYvbRnDnnXdKU6ZMMXyt0+kkPz8/adGiRQJTWbecnBwJgLRv3z5JkiSpqKhIsrW1lbZs2WLok5SUJAGQDh06JEnStR9ecrlcysrKMvRZvXq1pFarJa1WK0mSJL388stS+/btjcZ67LHHpNjY2Ka+pBavtLRUCg8Pl3bt2iX16tXLUNzy3oj1yiuvSD179rxhu16vl3x8fKR3333XcKyoqEiys7OTvvrqK0mSJCkxMVECIP3111+GPj///LMkk8mkjIwMSZIkadWqVZKrq6vhfl0fu02bNo19SVZjwIAB0vjx442OPfLII1JcXJwkSbw3ovyzuG3O+zBixAhpwIABRnm6d+8uPfPMM416jdaOyxJuUXV1NeLj49GnTx/DMblcjj59+uDQoUMCk1m34uJiAICbmxsAID4+HjU1NUb3oW3btggKCjLch0OHDqFjx45GD/eIjY1FSUkJzpw5Y+jzv+9xvQ/v5b+bMmUKBgwYYPL9470Ra/v27ejWrRuGDx8OLy8vdOnSBZ988omh/dKlS8jKyjL63rq4uKB79+5G90ej0aBbt26GPn369IFcLseRI0cMfe69914olUpDn9jYWJw7dw6FhYVNfZktUo8ePbB7926cP38eAHDy5En8+eef6N+/PwDeG0vRnPeBP+caB4vbW5SXlwedTmfyNDRvb29kZWUJSmXd9Ho9pk+fjrvvvhsdOnQAAGRlZUGpVEKj0Rj1/d/7kJWVVed9ut52sz4lJSWorKxsisuxCps3b8axY8ewaNEikzbeG7FSUlKwevVqhIeH45dffsGkSZPw/PPP47PPPgPwf9/fm/0My8rKgpeXl1G7jY0N3Nzc6nUPydisWbPw+OOPo23btrC1tUWXLl0wffp0xMXFAeC9sRTNeR9u1If3qX6s4vG7dHuZMmUKTp8+jT///FN0FAKQnp6OadOmYdeuXbC3txcdh/5Br9ejW7duWLhwIQCgS5cuOH36NNasWYMxY8YITnd7++abb7Bx40Zs2rQJ7du3x4kTJzB9+nT4+fnx3hDdAs7c3iIPDw8oFAqTT35nZ2fDx8dHUCrrNXXqVOzYsQN79uxBQECA4biPjw+qq6tRVFRk1P9/74OPj0+d9+l62836qNVqqFSqxr4cqxAfH4+cnBxER0fDxsYGNjY22LdvHz744APY2NjA29ub90YgX19ftGvXzuhYZGQk0tLSAPzf9/dmP8N8fHyQk5Nj1F5bW4uCgoJ63UMy9tJLLxlmbzt27IhRo0ZhxowZhn8B4b2xDM15H27Uh/epfljc3iKlUomuXbti9+7dhmN6vR67d+9GTEyMwGTWRZIkTJ06Fd999x1+//13hIaGGrV37doVtra2Rvfh3LlzSEtLM9yHmJgYJCQkGP0A2rVrF9RqteEv/5iYGKP3uN6H9/LGevfujYSEBJw4ccLw6tatG+Li4gx/5r0R5+677zbZNu/8+fMIDg4GAISGhsLHx8foe1tSUoIjR44Y3Z+ioiLEx8cb+vz+++/Q6/Xo3r27oc/+/ftRU1Nj6LNr1y60adMGrq6uTXZ9LVlFRQXkcuO/hhUKBfR6PQDeG0vRnPeBP+caiehPtFmDzZs3S3Z2dtKGDRukxMREaeLEiZJGozH65DfdmkmTJkkuLi7S3r17pczMTMOroqLC0OfZZ5+VgoKCpN9//136+++/pZiYGCkmJsbQfn27qb59+0onTpyQdu7cKXl6eta53dRLL70kJSUlSStXruR2Uw3wv7slSBLvjUhHjx6VbGxspLfffltKTk6WNm7cKDk4OEhffvmloc8777wjaTQa6YcffpBOnTolDR48uM5tjrp06SIdOXJE+vPPP6Xw8HCjbY6Kiookb29vadSoUdLp06elzZs3Sw4ODtxu6ibGjBkj+fv7G7YC27Ztm+Th4SG9/PLLhj68N82jtLRUOn78uHT8+HEJgLRs2TLp+PHjUmpqqiRJzXcfDhw4INnY2Ej/+c9/pKSkJGnevHncCqwBWNw2khUrVkhBQUGSUqmU7rzzTunw4cOiI1kVAHW+1q9fb+hTWVkpTZ48WXJ1dZUcHBykoUOHSpmZmUbvc/nyZal///6SSqWSPDw8pBdeeEGqqakx6rNnzx6pc+fOklKplFq1amU0Bpnnn8Ut741YP/74o9ShQwfJzs5Oatu2rfTxxx8btev1emnOnDmSt7e3ZGdnJ/Xu3Vs6d+6cUZ/8/HzpiSeekJycnCS1Wi2NGzdOKi0tNepz8uRJqWfPnpKdnZ3k7+8vvfPOO01+bS1ZSUmJNG3aNCkoKEiyt7eXWrVqJb322mtGW0Xx3jSPPXv21Pl3zJgxYyRJat778M0330gRERGSUqmU2rdvL/33v/9tsuu2VjJJ+p9HoRARERERtWBcc0tEREREVoPFLRERERFZDRa3RERERGQ1WNwSERERkdVgcUtEREREVoPFLRERERFZDRa3RERERGQ1WNwSERERkdVgcUtERHUaNWoUFi5c2GTvn5iYiICAAJSXlzfZGER0+2FxS0RkhrFjx2LIkCENPn/Dhg3QaDSNlqepnTx5Ej/99BOef/75JhujXbt2uOuuu7Bs2bImG4OIbj8sbomIyMSKFSswfPhwODk5Nek448aNw+rVq1FbW9uk4xDR7YPFLRFRI1i2bBk6duwIR0dHBAYGYvLkySgrKwMA7N27F+PGjUNxcTFkMhlkMhnmz58PANBqtXjxxRfh7+8PR0dHdO/eHXv37jW87/UZ319++QWRkZFwcnJCv379kJmZaTT+unXr0L59e9jZ2cHX1xdTp04FAIwfPx4DBw406ltTUwMvLy+sXbu2zmvR6XTYunUrBg0aZHQ8JCQEb731FkaPHg0nJycEBwdj+/btyM3NxeDBg+Hk5IROnTrh77//NpyTmpqKQYMGwdXVFY6Ojmjfvj1++uknQ/uDDz6IgoIC7Nu3r37fcCKiG2BxS0TUCORyOT744AOcOXMGn332GX7//Xe8/PLLAIAePXpg+fLlUKvVyMzMRGZmJl588UUAwNSpU3Ho0CFs3rwZp06dwvDhw9GvXz8kJycb3ruiogL/+c9/8MUXX2D//v1IS0sznA8Aq1evxpQpUzBx4kQkJCRg+/btCAsLAwA89dRT2Llzp1ExvGPHDlRUVOCxxx6r81pOnTqF4uJidOvWzaTtvffew913343jx49jwIABGDVqFEaPHo2RI0fi2LFjaN26NUaPHg1JkgAAU6ZMgVarxf79+5GQkIDFixcbzQYrlUp07twZf/zxR0O/9URExiQiIvpXY8aMkQYPHmx2/y1btkju7u6Gr9evXy+5uLgY9UlNTZUUCoWUkZFhdLx3797S7NmzDecBkC5cuGBoX7lypeTt7W342s/PT3rttddumKVdu3bS4sWLDV8PGjRIGjt27A37f/fdd5JCoZD0er3R8eDgYGnkyJGGrzMzMyUA0pw5cwzHDh06JAGQMjMzJUmSpI4dO0rz58+/4ViSJElDhw69aR4iovrgzC0RUSP47bff0Lt3b/j7+8PZ2RmjRo1Cfn4+KioqbnhOQkICdDodIiIi4OTkZHjt27cPFy9eNPRzcHBA69atDV/7+voiJycHAJCTk4OrV6+id+/eNxznqaeewvr16wEA2dnZ+PnnnzF+/Pgb9q+srISdnR1kMplJW6dOnQx/9vb2BgB07NjR5Nj1fM8//zzeeust3H333Zg3bx5OnTpl8p4qleqm3yciovpgcUtEdIsuX76MgQMHolOnTvj2228RHx+PlStXAgCqq6tveF5ZWRkUCgXi4+Nx4sQJwyspKQnvv/++oZ+tra3ReTKZzPDP/iqV6l/zjR49GikpKTh06BC+/PJLhIaG4p577rlhfw8PD1RUVNSZ/X+zXC9+6zqm1+sBXCusU1JSMGrUKCQkJKBbt25YsWKF0XsWFBTA09PzX6+DiMgcLG6JiG5RfHw89Ho9li5dirvuugsRERG4evWqUR+lUgmdTmd0rEuXLtDpdMjJyUFYWJjRy8fHx6yxnZ2dERISgt27d9+wj7u7O4YMGYL169djw4YNGDdu3E3fs3PnzgCu7UPbGAIDA/Hss89i27ZteOGFF/DJJ58YtZ8+fRpdunRplLGIiGxEByAiaimKi4tx4sQJo2Pu7u4ICwtDTU0NVqxYgUGDBuHAgQNYs2aNUb+QkBCUlZVh9+7diIqKgoODAyIiIhAXF4fRo0dj6dKl6NKlC3Jzc7F792506tQJAwYMMCvX/Pnz8eyzz8LLywv9+/dHaWkpDhw4gOeee87Q56mnnsLAgQOh0+kwZsyYm76fp6cnoqOj8eeffxoK3YaaPn06+vfvj4iICBQWFmLPnj2IjIw0tF++fBkZGRno06fPLY1DRHQdZ26JiMy0d+9edOnSxej1xhtvICoqCsuWLcPixYvRoUMHbNy4EYsWLTI6t0ePHnj22Wfx2GOPwdPTE0uWLAEArF+/HqNHj8YLL7yANm3aYMiQIfjrr78QFBRkdq4xY8Zg+fLlWLVqFdq3b4+BAwca7bYAAH369IGvry9iY2Ph5+f3r+/51FNPYePGjWZnuBGdTocpU6YgMjIS/fr1Q0REBFatWmVo/+qrr9C3b18EBwff8lhERAAgk64v3CIiIqtVVlYGf39/rF+/Ho888si/9q+srESbNm3w9ddfIyYmpkkyVVdXIzw8HJs2bcLdd9/dJGMQ0e2HyxKIiKyYXq9HXl4eli5dCo1Gg4cfftis81QqFT7//HPk5eU1Wba0tDS8+uqrLGyJqFFx5paIyIpdvnwZoaGhCAgIwIYNG266ZRgRkTVgcUtEREREVoMfKCMiIiIiq8HiloiIiIisBotbIiIiIrIaLG6JiIiIyGqwuCUiIiIiq8HiloiIiIisBotbIiIiIrIaLG6JiIiIyGr8P7EYhwue2sOLAAAAAElFTkSuQmCC", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "#BRAND NEW TESTING SCRIPT FOR ALL METHODS BASED OFF PREVIOUS TWO\n", "#ENHANCED VERSION OF MY PREVIOUS 2 TESTING SCRIPTS WITH EXTRAS\n", "#Testing script for Granite3.2-2B-Instruct using INT8 base + FP16 Adapters\n", "\n", "import os\n", "import torch\n", "import time\n", "import json\n", "import numpy as np\n", "import pandas as pd\n", "import evaluate\n", "import nltk\n", "import gc\n", "import math\n", "import re\n", "import matplotlib.pyplot as plt\n", "import mauve\n", "from transformers import AutoModelForCausalLM, AutoTokenizer, TrainingArguments, Trainer, EarlyStoppingCallback, TrainerCallback, BitsAndBytesConfig\n", "from transformers.trainer_utils import get_last_checkpoint\n", "from peft import LoraConfig, get_peft_model, prepare_model_for_kbit_training, PeftModel\n", "from datasets import Dataset\n", "from sklearn.metrics import accuracy_score, precision_recall_fscore_support\n", "from nltk.translate.bleu_score import sentence_bleu\n", "from rouge_score import rouge_scorer\n", "from torch.utils.data import DataLoader\n", "from sentence_transformers import SentenceTransformer, util\n", "\n", "nltk.download(\"punkt\")\n", "os.environ[\"TOKENIZERS_PARALLELISM\"] = \"false\"\n", "os.environ[\"PYTORCH_CUDA_ALLOC_CONF\"] = \"expandable_segments:True\"\n", "\n", "MODEL_NAME = \"ibm-granite/granite-3.2-2b-instruct\"\n", "ADAPTER_PATH = \"Granite3.2-2B-lora_adapters-FP16\"\n", "TEST_CSV_PATH = \"Testing Dataset RE.csv\"\n", "OUTPUT_JSON_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/INT8/FP16/Granite3.2-2B-INT8-lora-FP16-Evaluation_Results.json\"\n", "OUTPUT_INFER_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/INT8/FP16/Granite3.2-2B-INT8-lora-FP16-Inference_Curve.png\"\n", "OUTPUT_MEMORY_USAGE_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/INT8/FP16/Granite3.2-2B-INT8-lora-FP16-Memory_Usage_Curve.png\"\n", "OUTPUT_LATENCY_HIST_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/INT8/FP16/Granite3.2-2B-INT8-lora-FP16-Latency_Histogram.png\"\n", "OUTPUT_MEMORY_HIST_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/INT8/FP16/Granite3.2-2B-INT8-lora-FP16-Memory_Histogram.png\"\n", "SEMANTIC_MODEL = \"all-MiniLM-L6-v2\"\n", "\n", "tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)\n", "tokenizer.pad_token = tokenizer.eos_token\n", "tokenizer.padding_side = \"right\"\n", "\n", "test_df = pd.read_csv(TEST_CSV_PATH)\n", "\n", "def preprocess_function(examples):\n", " inputs = []\n", " labels = []\n", " \n", " for context, question, answer in zip(\n", " examples.get(\"Context\", [\"\"] * len(examples[\"Question\"])),\n", " examples[\"Question\"], \n", " examples[\"Answer\"]):\n", " \n", " context = context.strip() if context else \"\"\n", " question = question.strip()\n", " answer = answer.strip()\n", "\n", " if context:\n", " prompt = f\"Context: {context}\\nQuestion: {question}\\nAnswer:\"\n", " else:\n", " prompt = f\"Question: {question}\\nAnswer:\"\n", "\n", " full_text = prompt + \" \" + answer\n", " \n", " tokenized = tokenizer(full_text, padding=\"max_length\", truncation=True, max_length=512)\n", " prompt_ids = tokenizer(prompt, truncation=True, max_length=512, add_special_tokens=False)[\"input_ids\"]\n", "\n", " input_ids = tokenized[\"input_ids\"]\n", " attention_mask = tokenized[\"attention_mask\"]\n", " label_ids = input_ids.copy()\n", " label_ids[:len(prompt_ids)] = [-100] * len(prompt_ids)\n", " \n", " if all(id_ == -100 for id_ in label_ids):\n", " continue\n", "\n", " inputs.append({\"input_ids\": input_ids, \"attention_mask\": attention_mask, \"labels\": label_ids})\n", "\n", " return {\"input_ids\": [x[\"input_ids\"] for x in inputs], \"attention_mask\": [x[\"attention_mask\"] for x in inputs],\n", " \"labels\": [x[\"labels\"] for x in inputs]}\n", "\n", "test_dataset = Dataset.from_pandas(test_df).map(preprocess_function, batched=True, batch_size=32,\n", " remove_columns=test_df.columns.tolist())\n", "\n", "bnb_config = BitsAndBytesConfig(\n", " load_in_8bit=True,\n", " llm_int8_threshold=6.0,\n", " llm_int8_skip_modules=None)\n", "\n", "model = AutoModelForCausalLM.from_pretrained(MODEL_NAME, quantization_config=bnb_config, device_map=\"auto\")\n", "model = PeftModel.from_pretrained(model, ADAPTER_PATH).eval()\n", "model.config.pad_token_id = tokenizer.pad_token_id\n", "\n", "# Load semantic similarity model\n", "semantic_model = SentenceTransformer(SEMANTIC_MODEL)\n", "\n", "def compute_loss_and_perplexity():\n", " losses = []\n", " for sample in test_dataset:\n", " with torch.no_grad():\n", " input_ids = torch.tensor(sample[\"input_ids\"]).unsqueeze(0).to(\"cuda\")\n", " attention_mask = torch.tensor(sample[\"attention_mask\"]).unsqueeze(0).to(\"cuda\")\n", " labels = torch.tensor(sample[\"labels\"]).unsqueeze(0).to(\"cuda\")\n", " \n", " loss = model(input_ids=input_ids, attention_mask=attention_mask, labels=labels).loss.item()\n", " losses.append(loss)\n", " \n", " avg_loss = sum(losses) / len(losses)\n", " return avg_loss, math.exp(avg_loss)\n", "\n", "def extract_answer(text):\n", " return text.split(\"Answer:\")[-1].strip() if \"Answer:\" in text else text.strip()\n", "\n", "def normalize(text):\n", " return re.sub(r\"[^\\w\\s]\", \"\", text.strip().lower())\n", "\n", "def compute_metrics(preds, refs):\n", " decoded_preds = tokenizer.batch_decode(preds, skip_special_tokens=True)\n", " #decoded_refs = tokenizer.batch_decode(refs, skip_special_tokens=True)\n", "\n", " # Replace -100s in refs before decoding\n", " safe_refs = [[token if token != -100 else tokenizer.pad_token_id for token in ref] for ref in refs]\n", " decoded_refs = tokenizer.batch_decode(safe_refs, skip_special_tokens=True)\n", "\n", " preds_clean = [normalize(extract_answer(p)) for p in decoded_preds]\n", " refs_clean = [normalize(extract_answer(r)) for r in decoded_refs]\n", "\n", " sim_scores = util.cos_sim(semantic_model.encode(preds_clean, convert_to_tensor=True),\n", " semantic_model.encode(refs_clean, convert_to_tensor=True)).diagonal()\n", " semantic_threshold = 0.8\n", " matches = [1 if sim >= semantic_threshold else 0 for sim in sim_scores]\n", "\n", " accuracy = sum(matches) / len(matches)\n", " precision, recall, f1, _ = precision_recall_fscore_support(matches, matches, average=\"binary\", zero_division=0)\n", " avg_bleu = sum([sentence_bleu([r.split()], p.split()) for r, p in zip(refs_clean, preds_clean)]) / len(preds_clean)\n", "\n", " rouge = rouge_scorer.RougeScorer([\"rouge1\", \"rouge2\", \"rougeL\"], use_stemmer=True)\n", " rouge_scores = [rouge.score(ref, pred) for ref, pred in zip(refs_clean, preds_clean)]\n", " avg_rouge = {k: sum([s[k].fmeasure for s in rouge_scores]) / len(rouge_scores) for k in rouge_scores[0]}\n", "\n", " return {\"accuracy:\": accuracy, \"precision:\": precision, \"recall:\": recall, \"f1:\": f1,\n", " \"bleu:\": avg_bleu, \"rouge:\": avg_rouge, \"semantic_similarity_avg:\": sim_scores.mean().item()}, decoded_preds, decoded_refs\n", "\n", "def measure_inference_and_generate():\n", " preds, latencies, memory_used_per_sample, peak_memories = [], [], [], []\n", "\n", " #Measure model load memory (after full load + preparation)\n", " gc.collect()\n", " torch.cuda.empty_cache()\n", " torch.cuda.reset_peak_memory_stats()\n", " model_load_memory = torch.cuda.memory_allocated() / (1024 ** 3)\n", "\n", " for idx, sample in enumerate(test_dataset):\n", " gc.collect()\n", " torch.cuda.empty_cache()\n", " torch.cuda.reset_peak_memory_stats()\n", "\n", " input_ids = torch.tensor(sample[\"input_ids\"]).unsqueeze(0).to(\"cuda\")\n", " attention_mask = torch.tensor(sample[\"attention_mask\"]).unsqueeze(0).to(\"cuda\")\n", " \n", " # Measure base memory BEFORE\n", " base_memory = torch.cuda.memory_allocated()\n", "\n", " # Wait for everything to settle\n", " torch.cuda.synchronize()\n", " #mem_before = torch.cuda.memory_allocated() / (1024 ** 3)\n", " start_time = time.time()\n", "\n", " with torch.no_grad():\n", " output = model.generate(input_ids=input_ids, attention_mask=attention_mask, max_new_tokens=50,\n", " do_sample=True, top_p=0.9, top_k=50,\n", " temperature=0.7, repetition_penalty=1.1, length_penalty=0.8)\n", "\n", " torch.cuda.synchronize()\n", " end_time = time.time()\n", " #mem_after = torch.cuda.memory_allocated() / (1024 ** 3)\n", " peak_memory = torch.cuda.max_memory_allocated() \n", "\n", " inference_memory = (peak_memory - base_memory) / (1024 ** 3) # in GB\n", "\n", " preds.append(output[0].tolist())\n", " latencies.append((end_time - start_time) * 1000) # ms\n", " memory_used_per_sample.append(inference_memory) # Memory used by this inference\n", " peak_memories.append(peak_memory / (1024 ** 3)) # Peak memory usage during this sample\n", "\n", " # Calculate averages now\n", " avg_inference_memory = np.mean(memory_used_per_sample)\n", "\n", " return preds, latencies, memory_used_per_sample, peak_memories, model_load_memory, avg_inference_memory\n", "\n", "def compute_mauve(pred_texts, ref_texts):\n", " return mauve.compute_mauve(p_text=pred_texts, q_text=ref_texts,\n", " device_id=0, max_text_length=256).mauve\n", "\n", "print(\"Generating predictions...\")\n", "loss, perplexity = compute_loss_and_perplexity()\n", "generated_preds, latencies, memory_used_per_sample, peak_memories, model_load_memory, avg_inference_memory = measure_inference_and_generate()\n", "ref_labels = [sample[\"labels\"] for sample in test_dataset]\n", "metrics, decoded_preds, decoded_refs = compute_metrics(generated_preds, ref_labels)\n", "mauve_score = compute_mauve(decoded_preds, decoded_refs)\n", "\n", "# 1) Plot Inference_Performance curves for latency and memory usage\n", "plt.plot(latencies, label=\"Latency (ms)\")\n", "plt.plot(memory_used_per_sample, label=\"Memory Usage (GB)\")\n", "plt.title(\"Inference_Performance Curve\")\n", "plt.xlabel(\"Sample Index\")\n", "plt.legend()\n", "plt.savefig(OUTPUT_INFER_PATH)\n", "\n", "# 2a) Compute latency stats and then plot the latency histogram\n", "latencies_np = np.array(latencies)\n", "latency_stats = {\n", " \"min_latency_ms\": float(np.min(latencies_np)),\n", " \"max_latency_ms\": float(np.max(latencies_np)),\n", " \"lower_quartile_ms\": float(np.percentile(latencies_np, 25)),\n", " \"median_latency_ms\": float(np.median(latencies_np)),\n", " \"upper_quartile_ms\": float(np.percentile(latencies_np, 75)),\n", " \"avg_latency_ms\": float(np.mean(latencies_np))\n", "}\n", "\n", "# 2b) Plot the Histogram for Latency (ms)\n", "plt.figure(figsize=(8, 6))\n", "plt.hist(latencies, bins=20, color='skyblue', edgecolor='black')\n", "plt.axvline(latency_stats[\"min_latency_ms\"], color='green', linestyle='dashed', linewidth=1, label='Min')\n", "plt.axvline(latency_stats[\"lower_quartile_ms\"], color='orange', linestyle='dashed', linewidth=1, label='Q1')\n", "plt.axvline(latency_stats[\"median_latency_ms\"], color='red', linestyle='dashed', linewidth=1, label='Median')\n", "plt.axvline(latency_stats[\"upper_quartile_ms\"], color='purple', linestyle='dashed', linewidth=1, label='Q3')\n", "plt.axvline(latency_stats[\"max_latency_ms\"], color='black', linestyle='dashed', linewidth=1, label='Max')\n", "plt.title(\"Latency Histogram\")\n", "plt.xlabel(\"Latency (ms)\")\n", "plt.ylabel(\"Frequency\")\n", "plt.legend()\n", "plt.savefig(OUTPUT_LATENCY_HIST_PATH)\n", "\n", "# Line plot focusing on 0.1MB to 1MB\n", "plt.figure(figsize=(8, 6))\n", "plt.plot(memory_used_per_sample, label=\"Memory Usage (GB)\", color=\"teal\")\n", "plt.ylim(0.1, 0.5) # Zoom in to 0.1GB–0.5GB range\n", "plt.title(\"Memory Usage per Sample (Zoomed 100MB–500MB)\")\n", "plt.xlabel(\"Sample Index\")\n", "plt.ylabel(\"Memory (GB)\")\n", "plt.legend()\n", "plt.savefig(OUTPUT_MEMORY_USAGE_PATH)\n", "\n", "# 4) Compute memory stats and Plot the Histogram for memory usage\n", "memory_used_per_sample_np = np.array(memory_used_per_sample)\n", "memory_stats = {\n", " \"min_memory_gb\": float(np.min(memory_used_per_sample_np)),\n", " \"max_memory_gb\": float(np.max(memory_used_per_sample_np)),\n", " \"lower_quartile_gb\": float(np.percentile(memory_used_per_sample_np, 25)),\n", " \"median_memory_gb\": float(np.median(memory_used_per_sample_np)),\n", " \"upper_quartile_gb\": float(np.percentile(memory_used_per_sample_np, 75)),\n", " \"avg_memory_gb\": float(np.mean(memory_used_per_sample_np)),\n", " \"model_load_memory_gb\": model_load_memory,\n", " \"avg_inference_memory_gb\": avg_inference_memory\n", "}\n", "\n", "# Plot the Histogram for memory usage\n", "plt.figure(figsize=(8, 6))\n", "plt.hist(memory_used_per_sample, bins=20, color='lightcoral', edgecolor='black')\n", "plt.axvline(memory_stats[\"min_memory_gb\"], color='green', linestyle='dashed', linewidth=1, label='Min')\n", "plt.axvline(memory_stats[\"lower_quartile_gb\"], color='orange', linestyle='dashed', linewidth=1, label='Q1')\n", "plt.axvline(memory_stats[\"median_memory_gb\"], color='red', linestyle='dashed', linewidth=1, label='Median')\n", "plt.axvline(memory_stats[\"upper_quartile_gb\"], color='purple', linestyle='dashed', linewidth=1, label='Q3')\n", "plt.axvline(memory_stats[\"max_memory_gb\"], color='black', linestyle='dashed', linewidth=1, label='Max')\n", "plt.title(\"Memory Usage Histogram\")\n", "plt.xlabel(\"Memory Usage (GB)\")\n", "plt.ylabel(\"Frequency\")\n", "plt.legend()\n", "plt.savefig(OUTPUT_MEMORY_HIST_PATH)\n", "\n", "# Save all results\n", "results = {\"eval_loss:\": loss, \"perplexity:\": perplexity, \"performance_metrics:\": metrics, \"mauve:\": mauve_score,\n", " \"inference_performance:\": {**latency_stats, **memory_stats}}\n", "\n", "with open(OUTPUT_JSON_PATH, \"w\") as f:\n", " json.dump(results, f, indent=4)\n", "\n", "print(f\"Evaluation Complete. Results saved to {OUTPUT_JSON_PATH}\")\n" ] }, { "cell_type": "code", "execution_count": 4, "id": "4a2951a7-b43c-4395-873a-f0397dae74f3", "metadata": {}, "outputs": [], "source": [ "#3)#######################################################################################################################\n", "#STARTED ABOVE TESTING AT 3:25PM ON 29/04/25\n", "#FEATURISING STARTED AT 4:50PM AND ENDED AT 4:51PM (85 MIN AFTER STARTING)\n", "#ENDED ABOVE TESTING AT 4:55PM (90 MIN AFTER STARTING...)" ] }, { "cell_type": "code", "execution_count": 5, "id": "aa6ecade-94a4-4773-9544-401e55b3cdc0", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "[nltk_data] Downloading package punkt to /home/jovyan/nltk_data...\n", "[nltk_data] Package punkt is already up-to-date!\n", "Map: 100%|██████████| 1500/1500 [00:02<00:00, 608.38 examples/s]\n", "Loading checkpoint shards: 100%|██████████| 2/2 [00:23<00:00, 11.53s/it]\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Generating predictions...\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "/home/jovyan/Falcon1B/lib/python3.11/site-packages/transformers/generation/configuration_utils.py:695: UserWarning: `num_beams` is set to 1. However, `length_penalty` is set to `0.8` -- this flag is only used in beam-based generation modes. You should set `num_beams>1` or unset `length_penalty`.\n", " warnings.warn(\n", "/home/jovyan/Falcon1B/lib/python3.11/site-packages/nltk/translate/bleu_score.py:577: UserWarning: \n", "The hypothesis contains 0 counts of 3-gram overlaps.\n", "Therefore the BLEU score evaluates to 0, independently of\n", "how many N-gram overlaps of lower order it contains.\n", "Consider using lower n-gram order or use SmoothingFunction()\n", " warnings.warn(_msg)\n", "/home/jovyan/Falcon1B/lib/python3.11/site-packages/nltk/translate/bleu_score.py:577: UserWarning: \n", "The hypothesis contains 0 counts of 4-gram overlaps.\n", "Therefore the BLEU score evaluates to 0, independently of\n", "how many N-gram overlaps of lower order it contains.\n", "Consider using lower n-gram order or use SmoothingFunction()\n", " warnings.warn(_msg)\n", "Featurizing p: 100%|██████████| 1498/1498 [00:42<00:00, 35.07it/s]\n", "Featurizing q: 100%|██████████| 1498/1498 [00:42<00:00, 35.30it/s]\n", "WARNING clustering 2996 points to 150 centroids: please provide at least 5850 training points\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Evaluation Complete. Results saved to Complete_Evaluation_Results/Granite3.2-2B/FP16/FP16/Granite3.2-2B-FP16-lora-FP16-Evaluation_Results.json\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "#BRAND NEW TESTING SCRIPT FOR ALL METHODS BASED OFF PREVIOUS TWO\n", "#ENHANCED VERSION OF MY PREVIOUS 2 TESTING SCRIPTS WITH EXTRAS\n", "#Testing script for Granite3.2-2B-Instruct using FP16 base + FP16 Adapters\n", "\n", "import os\n", "import torch\n", "import time\n", "import json\n", "import numpy as np\n", "import pandas as pd\n", "import evaluate\n", "import nltk\n", "import gc\n", "import math\n", "import re\n", "import matplotlib.pyplot as plt\n", "import mauve\n", "from transformers import AutoModelForCausalLM, AutoTokenizer, TrainingArguments, Trainer, EarlyStoppingCallback, TrainerCallback, BitsAndBytesConfig\n", "from transformers.trainer_utils import get_last_checkpoint\n", "from peft import LoraConfig, get_peft_model, prepare_model_for_kbit_training, PeftModel\n", "from datasets import Dataset\n", "from sklearn.metrics import accuracy_score, precision_recall_fscore_support\n", "from nltk.translate.bleu_score import sentence_bleu\n", "from rouge_score import rouge_scorer\n", "from torch.utils.data import DataLoader\n", "from sentence_transformers import SentenceTransformer, util\n", "\n", "nltk.download(\"punkt\")\n", "os.environ[\"TOKENIZERS_PARALLELISM\"] = \"false\"\n", "os.environ[\"PYTORCH_CUDA_ALLOC_CONF\"] = \"expandable_segments:True\"\n", "\n", "MODEL_NAME = \"ibm-granite/granite-3.2-2b-instruct\"\n", "ADAPTER_PATH = \"Granite3.2-2B-lora_adapters-FP16\"\n", "TEST_CSV_PATH = \"Testing Dataset RE.csv\"\n", "OUTPUT_JSON_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/FP16/FP16/Granite3.2-2B-FP16-lora-FP16-Evaluation_Results.json\"\n", "OUTPUT_INFER_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/FP16/FP16/Granite3.2-2B-FP16-lora-FP16-Inference_Curve.png\"\n", "OUTPUT_MEMORY_USAGE_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/FP16/FP16/Granite3.2-2B-FP16-lora-FP16-Memory_Usage_Curve.png\"\n", "OUTPUT_LATENCY_HIST_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/FP16/FP16/Granite3.2-2B-FP16-lora-FP16-Latency_Histogram.png\"\n", "OUTPUT_MEMORY_HIST_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/FP16/FP16/Granite3.2-2B-FP16-lora-FP16-Memory_Histogram.png\"\n", "SEMANTIC_MODEL = \"all-MiniLM-L6-v2\"\n", "\n", "tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)\n", "tokenizer.pad_token = tokenizer.eos_token\n", "tokenizer.padding_side = \"right\"\n", "\n", "test_df = pd.read_csv(TEST_CSV_PATH)\n", "\n", "def preprocess_function(examples):\n", " inputs = []\n", " labels = []\n", " \n", " for context, question, answer in zip(\n", " examples.get(\"Context\", [\"\"] * len(examples[\"Question\"])),\n", " examples[\"Question\"], \n", " examples[\"Answer\"]):\n", " \n", " context = context.strip() if context else \"\"\n", " question = question.strip()\n", " answer = answer.strip()\n", "\n", " if context:\n", " prompt = f\"Context: {context}\\nQuestion: {question}\\nAnswer:\"\n", " else:\n", " prompt = f\"Question: {question}\\nAnswer:\"\n", "\n", " full_text = prompt + \" \" + answer\n", " \n", " tokenized = tokenizer(full_text, padding=\"max_length\", truncation=True, max_length=512)\n", " prompt_ids = tokenizer(prompt, truncation=True, max_length=512, add_special_tokens=False)[\"input_ids\"]\n", "\n", " input_ids = tokenized[\"input_ids\"]\n", " attention_mask = tokenized[\"attention_mask\"]\n", " label_ids = input_ids.copy()\n", " label_ids[:len(prompt_ids)] = [-100] * len(prompt_ids)\n", " \n", " if all(id_ == -100 for id_ in label_ids):\n", " continue\n", "\n", " inputs.append({\"input_ids\": input_ids, \"attention_mask\": attention_mask, \"labels\": label_ids})\n", "\n", " return {\"input_ids\": [x[\"input_ids\"] for x in inputs], \"attention_mask\": [x[\"attention_mask\"] for x in inputs],\n", " \"labels\": [x[\"labels\"] for x in inputs]}\n", "\n", "test_dataset = Dataset.from_pandas(test_df).map(preprocess_function, batched=True, batch_size=32,\n", " remove_columns=test_df.columns.tolist())\n", "\n", "#Removed BitsAndBytesConfig because the base is loaded in FP16 for LoRA (not QLoRA)\n", "\n", "model = AutoModelForCausalLM.from_pretrained(MODEL_NAME, torch_dtype=torch.float16, device_map=\"auto\",trust_remote_code=True)\n", "model = PeftModel.from_pretrained(model, ADAPTER_PATH).eval()\n", "model.config.pad_token_id = tokenizer.pad_token_id\n", "\n", "# Load semantic similarity model\n", "semantic_model = SentenceTransformer(SEMANTIC_MODEL)\n", "\n", "def compute_loss_and_perplexity():\n", " losses = []\n", " for sample in test_dataset:\n", " with torch.no_grad():\n", " input_ids = torch.tensor(sample[\"input_ids\"]).unsqueeze(0).to(\"cuda\")\n", " attention_mask = torch.tensor(sample[\"attention_mask\"]).unsqueeze(0).to(\"cuda\")\n", " labels = torch.tensor(sample[\"labels\"]).unsqueeze(0).to(\"cuda\")\n", " \n", " loss = model(input_ids=input_ids, attention_mask=attention_mask, labels=labels).loss.item()\n", " losses.append(loss)\n", " \n", " avg_loss = sum(losses) / len(losses)\n", " return avg_loss, math.exp(avg_loss)\n", "\n", "def extract_answer(text):\n", " return text.split(\"Answer:\")[-1].strip() if \"Answer:\" in text else text.strip()\n", "\n", "def normalize(text):\n", " return re.sub(r\"[^\\w\\s]\", \"\", text.strip().lower())\n", "\n", "def compute_metrics(preds, refs):\n", " decoded_preds = tokenizer.batch_decode(preds, skip_special_tokens=True)\n", " #decoded_refs = tokenizer.batch_decode(refs, skip_special_tokens=True)\n", "\n", " # Replace -100s in refs before decoding\n", " safe_refs = [[token if token != -100 else tokenizer.pad_token_id for token in ref] for ref in refs]\n", " decoded_refs = tokenizer.batch_decode(safe_refs, skip_special_tokens=True)\n", "\n", " preds_clean = [normalize(extract_answer(p)) for p in decoded_preds]\n", " refs_clean = [normalize(extract_answer(r)) for r in decoded_refs]\n", "\n", " sim_scores = util.cos_sim(semantic_model.encode(preds_clean, convert_to_tensor=True),\n", " semantic_model.encode(refs_clean, convert_to_tensor=True)).diagonal()\n", " semantic_threshold = 0.8\n", " matches = [1 if sim >= semantic_threshold else 0 for sim in sim_scores]\n", "\n", " accuracy = sum(matches) / len(matches)\n", " precision, recall, f1, _ = precision_recall_fscore_support(matches, matches, average=\"binary\", zero_division=0)\n", " avg_bleu = sum([sentence_bleu([r.split()], p.split()) for r, p in zip(refs_clean, preds_clean)]) / len(preds_clean)\n", "\n", " rouge = rouge_scorer.RougeScorer([\"rouge1\", \"rouge2\", \"rougeL\"], use_stemmer=True)\n", " rouge_scores = [rouge.score(ref, pred) for ref, pred in zip(refs_clean, preds_clean)]\n", " avg_rouge = {k: sum([s[k].fmeasure for s in rouge_scores]) / len(rouge_scores) for k in rouge_scores[0]}\n", "\n", " return {\"accuracy:\": accuracy, \"precision:\": precision, \"recall:\": recall, \"f1:\": f1,\n", " \"bleu:\": avg_bleu, \"rouge:\": avg_rouge, \"semantic_similarity_avg:\": sim_scores.mean().item()}, decoded_preds, decoded_refs\n", "\n", "def measure_inference_and_generate():\n", " preds, latencies, memory_used_per_sample, peak_memories = [], [], [], []\n", "\n", " #Measure model load memory (after full load + preparation)\n", " gc.collect()\n", " torch.cuda.empty_cache()\n", " torch.cuda.reset_peak_memory_stats()\n", " model_load_memory = torch.cuda.memory_allocated() / (1024 ** 3)\n", "\n", " for idx, sample in enumerate(test_dataset):\n", " gc.collect()\n", " torch.cuda.empty_cache()\n", " torch.cuda.reset_peak_memory_stats()\n", "\n", " input_ids = torch.tensor(sample[\"input_ids\"]).unsqueeze(0).to(\"cuda\")\n", " attention_mask = torch.tensor(sample[\"attention_mask\"]).unsqueeze(0).to(\"cuda\")\n", " \n", " # Measure base memory BEFORE\n", " base_memory = torch.cuda.memory_allocated()\n", "\n", " # Wait for everything to settle\n", " torch.cuda.synchronize()\n", " #mem_before = torch.cuda.memory_allocated() / (1024 ** 3)\n", " start_time = time.time()\n", "\n", " with torch.no_grad():\n", " output = model.generate(input_ids=input_ids, attention_mask=attention_mask, max_new_tokens=50,\n", " do_sample=True, top_p=0.9, top_k=50,\n", " temperature=0.7, repetition_penalty=1.1, length_penalty=0.8)\n", "\n", " torch.cuda.synchronize()\n", " end_time = time.time()\n", " #mem_after = torch.cuda.memory_allocated() / (1024 ** 3)\n", " peak_memory = torch.cuda.max_memory_allocated() \n", "\n", " inference_memory = (peak_memory - base_memory) / (1024 ** 3) # in GB\n", "\n", " preds.append(output[0].tolist())\n", " latencies.append((end_time - start_time) * 1000) # ms\n", " memory_used_per_sample.append(inference_memory) # Memory used by this inference\n", " peak_memories.append(peak_memory / (1024 ** 3)) # Peak memory usage during this sample\n", "\n", " # Calculate averages now\n", " avg_inference_memory = np.mean(memory_used_per_sample)\n", "\n", " return preds, latencies, memory_used_per_sample, peak_memories, model_load_memory, avg_inference_memory\n", "\n", "def compute_mauve(pred_texts, ref_texts):\n", " return mauve.compute_mauve(p_text=pred_texts, q_text=ref_texts,\n", " device_id=0, max_text_length=256).mauve\n", "\n", "print(\"Generating predictions...\")\n", "loss, perplexity = compute_loss_and_perplexity()\n", "generated_preds, latencies, memory_used_per_sample, peak_memories, model_load_memory, avg_inference_memory = measure_inference_and_generate()\n", "ref_labels = [sample[\"labels\"] for sample in test_dataset]\n", "metrics, decoded_preds, decoded_refs = compute_metrics(generated_preds, ref_labels)\n", "mauve_score = compute_mauve(decoded_preds, decoded_refs)\n", "\n", "# 1) Plot Inference_Performance curves for latency and memory usage\n", "plt.plot(latencies, label=\"Latency (ms)\")\n", "plt.plot(memory_used_per_sample, label=\"Memory Usage (GB)\")\n", "plt.title(\"Inference_Performance Curve\")\n", "plt.xlabel(\"Sample Index\")\n", "plt.legend()\n", "plt.savefig(OUTPUT_INFER_PATH)\n", "\n", "# 2a) Compute latency stats and then plot the latency histogram\n", "latencies_np = np.array(latencies)\n", "latency_stats = {\n", " \"min_latency_ms\": float(np.min(latencies_np)),\n", " \"max_latency_ms\": float(np.max(latencies_np)),\n", " \"lower_quartile_ms\": float(np.percentile(latencies_np, 25)),\n", " \"median_latency_ms\": float(np.median(latencies_np)),\n", " \"upper_quartile_ms\": float(np.percentile(latencies_np, 75)),\n", " \"avg_latency_ms\": float(np.mean(latencies_np))\n", "}\n", "\n", "# 2b) Plot the Histogram for Latency (ms)\n", "plt.figure(figsize=(8, 6))\n", "plt.hist(latencies, bins=20, color='skyblue', edgecolor='black')\n", "plt.axvline(latency_stats[\"min_latency_ms\"], color='green', linestyle='dashed', linewidth=1, label='Min')\n", "plt.axvline(latency_stats[\"lower_quartile_ms\"], color='orange', linestyle='dashed', linewidth=1, label='Q1')\n", "plt.axvline(latency_stats[\"median_latency_ms\"], color='red', linestyle='dashed', linewidth=1, label='Median')\n", "plt.axvline(latency_stats[\"upper_quartile_ms\"], color='purple', linestyle='dashed', linewidth=1, label='Q3')\n", "plt.axvline(latency_stats[\"max_latency_ms\"], color='black', linestyle='dashed', linewidth=1, label='Max')\n", "plt.title(\"Latency Histogram\")\n", "plt.xlabel(\"Latency (ms)\")\n", "plt.ylabel(\"Frequency\")\n", "plt.legend()\n", "plt.savefig(OUTPUT_LATENCY_HIST_PATH)\n", "\n", "# Line plot focusing on 0.1MB to 1MB\n", "plt.figure(figsize=(8, 6))\n", "plt.plot(memory_used_per_sample, label=\"Memory Usage (GB)\", color=\"teal\")\n", "plt.ylim(0.1, 0.5) # Zoom in to 0.1GB–0.5GB range\n", "plt.title(\"Memory Usage per Sample (Zoomed 100MB–500MB)\")\n", "plt.xlabel(\"Sample Index\")\n", "plt.ylabel(\"Memory (GB)\")\n", "plt.legend()\n", "plt.savefig(OUTPUT_MEMORY_USAGE_PATH)\n", "\n", "# 4) Compute memory stats and Plot the Histogram for memory usage\n", "memory_used_per_sample_np = np.array(memory_used_per_sample)\n", "memory_stats = {\n", " \"min_memory_gb\": float(np.min(memory_used_per_sample_np)),\n", " \"max_memory_gb\": float(np.max(memory_used_per_sample_np)),\n", " \"lower_quartile_gb\": float(np.percentile(memory_used_per_sample_np, 25)),\n", " \"median_memory_gb\": float(np.median(memory_used_per_sample_np)),\n", " \"upper_quartile_gb\": float(np.percentile(memory_used_per_sample_np, 75)),\n", " \"avg_memory_gb\": float(np.mean(memory_used_per_sample_np)),\n", " \"model_load_memory_gb\": model_load_memory,\n", " \"avg_inference_memory_gb\": avg_inference_memory\n", "}\n", "\n", "# Plot the Histogram for memory usage\n", "plt.figure(figsize=(8, 6))\n", "plt.hist(memory_used_per_sample, bins=20, color='lightcoral', edgecolor='black')\n", "plt.axvline(memory_stats[\"min_memory_gb\"], color='green', linestyle='dashed', linewidth=1, label='Min')\n", "plt.axvline(memory_stats[\"lower_quartile_gb\"], color='orange', linestyle='dashed', linewidth=1, label='Q1')\n", "plt.axvline(memory_stats[\"median_memory_gb\"], color='red', linestyle='dashed', linewidth=1, label='Median')\n", "plt.axvline(memory_stats[\"upper_quartile_gb\"], color='purple', linestyle='dashed', linewidth=1, label='Q3')\n", "plt.axvline(memory_stats[\"max_memory_gb\"], color='black', linestyle='dashed', linewidth=1, label='Max')\n", "plt.title(\"Memory Usage Histogram\")\n", "plt.xlabel(\"Memory Usage (GB)\")\n", "plt.ylabel(\"Frequency\")\n", "plt.legend()\n", "plt.savefig(OUTPUT_MEMORY_HIST_PATH)\n", "\n", "# Save all results\n", "results = {\"eval_loss:\": loss, \"perplexity:\": perplexity, \"performance_metrics:\": metrics, \"mauve:\": mauve_score,\n", " \"inference_performance:\": {**latency_stats, **memory_stats}}\n", "\n", "with open(OUTPUT_JSON_PATH, \"w\") as f:\n", " json.dump(results, f, indent=4)\n", "\n", "print(f\"Evaluation Complete. Results saved to {OUTPUT_JSON_PATH}\")\n" ] }, { "cell_type": "code", "execution_count": 6, "id": "e8a348a0-7495-4ae7-b476-2303985e7a74", "metadata": {}, "outputs": [], "source": [ "#4)######################################################################################################################\n", "#STARTED ABOVE TESTING AT 4:55PM ON 29/04/25\n", "#FEATURISING STARTED AT 5:31PM AND ENDED AT 5:32PM (36 MIN AFTER STARTING)\n", "#ENDED ABOVE TESTING AT 5:37PM (42 MIN AFTER STARTING)" ] }, { "cell_type": "code", "execution_count": 7, "id": "77d0c58e-10c4-4b6a-b76c-3c3b9fb53bee", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "[nltk_data] Downloading package punkt to /home/jovyan/nltk_data...\n", "[nltk_data] Package punkt is already up-to-date!\n", "Map: 100%|██████████| 1500/1500 [00:02<00:00, 599.20 examples/s]\n", "Loading checkpoint shards: 100%|██████████| 2/2 [00:23<00:00, 11.61s/it]\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Generating predictions...\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "/home/jovyan/Falcon1B/lib/python3.11/site-packages/transformers/generation/configuration_utils.py:695: UserWarning: `num_beams` is set to 1. However, `length_penalty` is set to `0.8` -- this flag is only used in beam-based generation modes. You should set `num_beams>1` or unset `length_penalty`.\n", " warnings.warn(\n", "Featurizing p: 100%|██████████| 1498/1498 [00:42<00:00, 35.11it/s]\n", "Featurizing q: 100%|██████████| 1498/1498 [00:42<00:00, 35.35it/s]\n", "WARNING clustering 2996 points to 150 centroids: please provide at least 5850 training points\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Evaluation Complete. Results saved to Complete_Evaluation_Results/Granite3.2-2B/FP4/FP16/Granite3.2-2B-FP4-lora-FP16-Evaluation_Results.json\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "#BRAND NEW TESTING SCRIPT FOR ALL METHODS BASED OFF PREVIOUS TWO\n", "#ENHANCED VERSION OF MY PREVIOUS 2 TESTING SCRIPTS WITH EXTRAS\n", "#Testing script for Granite3.2-2B-Instruct using FP4 base + FP16 Adapters\n", "\n", "import os\n", "import torch\n", "import time\n", "import json\n", "import numpy as np\n", "import pandas as pd\n", "import evaluate\n", "import nltk\n", "import gc\n", "import math\n", "import re\n", "import matplotlib.pyplot as plt\n", "import mauve\n", "from transformers import AutoModelForCausalLM, AutoTokenizer, TrainingArguments, Trainer, EarlyStoppingCallback, TrainerCallback, BitsAndBytesConfig\n", "from transformers.trainer_utils import get_last_checkpoint\n", "from peft import LoraConfig, get_peft_model, prepare_model_for_kbit_training, PeftModel\n", "from datasets import Dataset\n", "from sklearn.metrics import accuracy_score, precision_recall_fscore_support\n", "from nltk.translate.bleu_score import sentence_bleu\n", "from rouge_score import rouge_scorer\n", "from torch.utils.data import DataLoader\n", "from sentence_transformers import SentenceTransformer, util\n", "\n", "nltk.download(\"punkt\")\n", "os.environ[\"TOKENIZERS_PARALLELISM\"] = \"false\"\n", "os.environ[\"PYTORCH_CUDA_ALLOC_CONF\"] = \"expandable_segments:True\"\n", "\n", "MODEL_NAME = \"ibm-granite/granite-3.2-2b-instruct\"\n", "ADAPTER_PATH = \"Granite3.2-2B-lora_adapters-FP16\"\n", "TEST_CSV_PATH = \"Testing Dataset RE.csv\"\n", "OUTPUT_JSON_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/FP4/FP16/Granite3.2-2B-FP4-lora-FP16-Evaluation_Results.json\"\n", "OUTPUT_INFER_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/FP4/FP16/Granite3.2-2B-FP4-lora-FP16-Inference_Curve.png\"\n", "OUTPUT_MEMORY_USAGE_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/FP4/FP16/Granite3.2-2B-FP4-lora-FP16-Memory_Usage_Curve.png\"\n", "OUTPUT_LATENCY_HIST_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/FP4/FP16/Granite3.2-2B-FP4-lora-FP16-Latency_Histogram.png\"\n", "OUTPUT_MEMORY_HIST_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/FP4/FP16/Granite3.2-2B-FP4-lora-FP16-Memory_Histogram.png\"\n", "SEMANTIC_MODEL = \"all-MiniLM-L6-v2\"\n", "\n", "tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)\n", "tokenizer.pad_token = tokenizer.eos_token\n", "tokenizer.padding_side = \"right\"\n", "\n", "test_df = pd.read_csv(TEST_CSV_PATH)\n", "\n", "def preprocess_function(examples):\n", " inputs = []\n", " labels = []\n", " \n", " for context, question, answer in zip(\n", " examples.get(\"Context\", [\"\"] * len(examples[\"Question\"])),\n", " examples[\"Question\"], \n", " examples[\"Answer\"]):\n", " \n", " context = context.strip() if context else \"\"\n", " question = question.strip()\n", " answer = answer.strip()\n", "\n", " if context:\n", " prompt = f\"Context: {context}\\nQuestion: {question}\\nAnswer:\"\n", " else:\n", " prompt = f\"Question: {question}\\nAnswer:\"\n", "\n", " full_text = prompt + \" \" + answer\n", " \n", " tokenized = tokenizer(full_text, padding=\"max_length\", truncation=True, max_length=512)\n", " prompt_ids = tokenizer(prompt, truncation=True, max_length=512, add_special_tokens=False)[\"input_ids\"]\n", "\n", " input_ids = tokenized[\"input_ids\"]\n", " attention_mask = tokenized[\"attention_mask\"]\n", " label_ids = input_ids.copy()\n", " label_ids[:len(prompt_ids)] = [-100] * len(prompt_ids)\n", " \n", " if all(id_ == -100 for id_ in label_ids):\n", " continue\n", "\n", " inputs.append({\"input_ids\": input_ids, \"attention_mask\": attention_mask, \"labels\": label_ids})\n", "\n", " return {\"input_ids\": [x[\"input_ids\"] for x in inputs], \"attention_mask\": [x[\"attention_mask\"] for x in inputs],\n", " \"labels\": [x[\"labels\"] for x in inputs]}\n", "\n", "test_dataset = Dataset.from_pandas(test_df).map(preprocess_function, batched=True, batch_size=32,\n", " remove_columns=test_df.columns.tolist())\n", "\n", "bnb_config = BitsAndBytesConfig(\n", " load_in_4bit=True,\n", " bnb_4bit_quant_type=\"fp4\",\n", " bnb_4bit_compute_dtype=torch.float16,\n", " bnb_4bit_use_double_quant=True\n", ")\n", "\n", "model = AutoModelForCausalLM.from_pretrained(MODEL_NAME, quantization_config=bnb_config, device_map=\"auto\")\n", "model = PeftModel.from_pretrained(model, ADAPTER_PATH).eval()\n", "model.config.pad_token_id = tokenizer.pad_token_id\n", "\n", "# Load semantic similarity model\n", "semantic_model = SentenceTransformer(SEMANTIC_MODEL)\n", "\n", "def compute_loss_and_perplexity():\n", " losses = []\n", " for sample in test_dataset:\n", " with torch.no_grad():\n", " input_ids = torch.tensor(sample[\"input_ids\"]).unsqueeze(0).to(\"cuda\")\n", " attention_mask = torch.tensor(sample[\"attention_mask\"]).unsqueeze(0).to(\"cuda\")\n", " labels = torch.tensor(sample[\"labels\"]).unsqueeze(0).to(\"cuda\")\n", " \n", " loss = model(input_ids=input_ids, attention_mask=attention_mask, labels=labels).loss.item()\n", " losses.append(loss)\n", " \n", " avg_loss = sum(losses) / len(losses)\n", " return avg_loss, math.exp(avg_loss)\n", "\n", "def extract_answer(text):\n", " return text.split(\"Answer:\")[-1].strip() if \"Answer:\" in text else text.strip()\n", "\n", "def normalize(text):\n", " return re.sub(r\"[^\\w\\s]\", \"\", text.strip().lower())\n", "\n", "def compute_metrics(preds, refs):\n", " decoded_preds = tokenizer.batch_decode(preds, skip_special_tokens=True)\n", " #decoded_refs = tokenizer.batch_decode(refs, skip_special_tokens=True)\n", "\n", " # Replace -100s in refs before decoding\n", " safe_refs = [[token if token != -100 else tokenizer.pad_token_id for token in ref] for ref in refs]\n", " decoded_refs = tokenizer.batch_decode(safe_refs, skip_special_tokens=True)\n", "\n", " preds_clean = [normalize(extract_answer(p)) for p in decoded_preds]\n", " refs_clean = [normalize(extract_answer(r)) for r in decoded_refs]\n", "\n", " sim_scores = util.cos_sim(semantic_model.encode(preds_clean, convert_to_tensor=True),\n", " semantic_model.encode(refs_clean, convert_to_tensor=True)).diagonal()\n", " semantic_threshold = 0.8\n", " matches = [1 if sim >= semantic_threshold else 0 for sim in sim_scores]\n", "\n", " accuracy = sum(matches) / len(matches)\n", " precision, recall, f1, _ = precision_recall_fscore_support(matches, matches, average=\"binary\", zero_division=0)\n", " avg_bleu = sum([sentence_bleu([r.split()], p.split()) for r, p in zip(refs_clean, preds_clean)]) / len(preds_clean)\n", "\n", " rouge = rouge_scorer.RougeScorer([\"rouge1\", \"rouge2\", \"rougeL\"], use_stemmer=True)\n", " rouge_scores = [rouge.score(ref, pred) for ref, pred in zip(refs_clean, preds_clean)]\n", " avg_rouge = {k: sum([s[k].fmeasure for s in rouge_scores]) / len(rouge_scores) for k in rouge_scores[0]}\n", "\n", " return {\"accuracy:\": accuracy, \"precision:\": precision, \"recall:\": recall, \"f1:\": f1,\n", " \"bleu:\": avg_bleu, \"rouge:\": avg_rouge, \"semantic_similarity_avg:\": sim_scores.mean().item()}, decoded_preds, decoded_refs\n", "\n", "def measure_inference_and_generate():\n", " preds, latencies, memory_used_per_sample, peak_memories = [], [], [], []\n", "\n", " #Measure model load memory (after full load + preparation)\n", " gc.collect()\n", " torch.cuda.empty_cache()\n", " torch.cuda.reset_peak_memory_stats()\n", " model_load_memory = torch.cuda.memory_allocated() / (1024 ** 3)\n", "\n", " for idx, sample in enumerate(test_dataset):\n", " gc.collect()\n", " torch.cuda.empty_cache()\n", " torch.cuda.reset_peak_memory_stats()\n", "\n", " input_ids = torch.tensor(sample[\"input_ids\"]).unsqueeze(0).to(\"cuda\")\n", " attention_mask = torch.tensor(sample[\"attention_mask\"]).unsqueeze(0).to(\"cuda\")\n", " \n", " # Measure base memory BEFORE\n", " base_memory = torch.cuda.memory_allocated()\n", "\n", " # Wait for everything to settle\n", " torch.cuda.synchronize()\n", " #mem_before = torch.cuda.memory_allocated() / (1024 ** 3)\n", " start_time = time.time()\n", "\n", " with torch.no_grad():\n", " output = model.generate(input_ids=input_ids, attention_mask=attention_mask, max_new_tokens=50,\n", " do_sample=True, top_p=0.9, top_k=50,\n", " temperature=0.7, repetition_penalty=1.1, length_penalty=0.8)\n", "\n", " torch.cuda.synchronize()\n", " end_time = time.time()\n", " #mem_after = torch.cuda.memory_allocated() / (1024 ** 3)\n", " peak_memory = torch.cuda.max_memory_allocated() \n", "\n", " inference_memory = (peak_memory - base_memory) / (1024 ** 3) # in GB\n", "\n", " preds.append(output[0].tolist())\n", " latencies.append((end_time - start_time) * 1000) # ms\n", " memory_used_per_sample.append(inference_memory) # Memory used by this inference\n", " peak_memories.append(peak_memory / (1024 ** 3)) # Peak memory usage during this sample\n", "\n", " # Calculate averages now\n", " avg_inference_memory = np.mean(memory_used_per_sample)\n", "\n", " return preds, latencies, memory_used_per_sample, peak_memories, model_load_memory, avg_inference_memory\n", "\n", "def compute_mauve(pred_texts, ref_texts):\n", " return mauve.compute_mauve(p_text=pred_texts, q_text=ref_texts,\n", " device_id=0, max_text_length=256).mauve\n", "\n", "print(\"Generating predictions...\")\n", "loss, perplexity = compute_loss_and_perplexity()\n", "generated_preds, latencies, memory_used_per_sample, peak_memories, model_load_memory, avg_inference_memory = measure_inference_and_generate()\n", "ref_labels = [sample[\"labels\"] for sample in test_dataset]\n", "metrics, decoded_preds, decoded_refs = compute_metrics(generated_preds, ref_labels)\n", "mauve_score = compute_mauve(decoded_preds, decoded_refs)\n", "\n", "# 1) Plot Inference_Performance curves for latency and memory usage\n", "plt.plot(latencies, label=\"Latency (ms)\")\n", "plt.plot(memory_used_per_sample, label=\"Memory Usage (GB)\")\n", "plt.title(\"Inference_Performance Curve\")\n", "plt.xlabel(\"Sample Index\")\n", "plt.legend()\n", "plt.savefig(OUTPUT_INFER_PATH)\n", "\n", "# 2a) Compute latency stats and then plot the latency histogram\n", "latencies_np = np.array(latencies)\n", "latency_stats = {\n", " \"min_latency_ms\": float(np.min(latencies_np)),\n", " \"max_latency_ms\": float(np.max(latencies_np)),\n", " \"lower_quartile_ms\": float(np.percentile(latencies_np, 25)),\n", " \"median_latency_ms\": float(np.median(latencies_np)),\n", " \"upper_quartile_ms\": float(np.percentile(latencies_np, 75)),\n", " \"avg_latency_ms\": float(np.mean(latencies_np))\n", "}\n", "\n", "# 2b) Plot the Histogram for Latency (ms)\n", "plt.figure(figsize=(8, 6))\n", "plt.hist(latencies, bins=20, color='skyblue', edgecolor='black')\n", "plt.axvline(latency_stats[\"min_latency_ms\"], color='green', linestyle='dashed', linewidth=1, label='Min')\n", "plt.axvline(latency_stats[\"lower_quartile_ms\"], color='orange', linestyle='dashed', linewidth=1, label='Q1')\n", "plt.axvline(latency_stats[\"median_latency_ms\"], color='red', linestyle='dashed', linewidth=1, label='Median')\n", "plt.axvline(latency_stats[\"upper_quartile_ms\"], color='purple', linestyle='dashed', linewidth=1, label='Q3')\n", "plt.axvline(latency_stats[\"max_latency_ms\"], color='black', linestyle='dashed', linewidth=1, label='Max')\n", "plt.title(\"Latency Histogram\")\n", "plt.xlabel(\"Latency (ms)\")\n", "plt.ylabel(\"Frequency\")\n", "plt.legend()\n", "plt.savefig(OUTPUT_LATENCY_HIST_PATH)\n", "\n", "# Line plot focusing on 0.1MB to 1MB\n", "plt.figure(figsize=(8, 6))\n", "plt.plot(memory_used_per_sample, label=\"Memory Usage (GB)\", color=\"teal\")\n", "plt.ylim(0.1, 0.5) # Zoom in to 0.1GB–0.5GB range\n", "plt.title(\"Memory Usage per Sample (Zoomed 100MB–500MB)\")\n", "plt.xlabel(\"Sample Index\")\n", "plt.ylabel(\"Memory (GB)\")\n", "plt.legend()\n", "plt.savefig(OUTPUT_MEMORY_USAGE_PATH)\n", "\n", "# 4) Compute memory stats and Plot the Histogram for memory usage\n", "memory_used_per_sample_np = np.array(memory_used_per_sample)\n", "memory_stats = {\n", " \"min_memory_gb\": float(np.min(memory_used_per_sample_np)),\n", " \"max_memory_gb\": float(np.max(memory_used_per_sample_np)),\n", " \"lower_quartile_gb\": float(np.percentile(memory_used_per_sample_np, 25)),\n", " \"median_memory_gb\": float(np.median(memory_used_per_sample_np)),\n", " \"upper_quartile_gb\": float(np.percentile(memory_used_per_sample_np, 75)),\n", " \"avg_memory_gb\": float(np.mean(memory_used_per_sample_np)),\n", " \"model_load_memory_gb\": model_load_memory,\n", " \"avg_inference_memory_gb\": avg_inference_memory\n", "}\n", "\n", "# Plot the Histogram for memory usage\n", "plt.figure(figsize=(8, 6))\n", "plt.hist(memory_used_per_sample, bins=20, color='lightcoral', edgecolor='black')\n", "plt.axvline(memory_stats[\"min_memory_gb\"], color='green', linestyle='dashed', linewidth=1, label='Min')\n", "plt.axvline(memory_stats[\"lower_quartile_gb\"], color='orange', linestyle='dashed', linewidth=1, label='Q1')\n", "plt.axvline(memory_stats[\"median_memory_gb\"], color='red', linestyle='dashed', linewidth=1, label='Median')\n", "plt.axvline(memory_stats[\"upper_quartile_gb\"], color='purple', linestyle='dashed', linewidth=1, label='Q3')\n", "plt.axvline(memory_stats[\"max_memory_gb\"], color='black', linestyle='dashed', linewidth=1, label='Max')\n", "plt.title(\"Memory Usage Histogram\")\n", "plt.xlabel(\"Memory Usage (GB)\")\n", "plt.ylabel(\"Frequency\")\n", "plt.legend()\n", "plt.savefig(OUTPUT_MEMORY_HIST_PATH)\n", "\n", "# Save all results\n", "results = {\"eval_loss:\": loss, \"perplexity:\": perplexity, \"performance_metrics:\": metrics, \"mauve:\": mauve_score,\n", " \"inference_performance:\": {**latency_stats, **memory_stats}}\n", "\n", "with open(OUTPUT_JSON_PATH, \"w\") as f:\n", " json.dump(results, f, indent=4)\n", "\n", "print(f\"Evaluation Complete. Results saved to {OUTPUT_JSON_PATH}\")\n" ] }, { "cell_type": "code", "execution_count": 8, "id": "c837ff27-429a-4194-94d4-feae38765ca5", "metadata": {}, "outputs": [], "source": [ "#5)#####################################################################################################################\n", "#STARTED ABOVE TESTING AT 5:37PM ON 29/04/25\n", "#ENDED ABOVE TESTING AT 6:25PM (47 MIN AFTER STARTING)" ] }, { "cell_type": "code", "execution_count": 9, "id": "f8b5ca7f-f865-4886-b73d-47e7df250231", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "[nltk_data] Downloading package punkt to /home/jovyan/nltk_data...\n", "[nltk_data] Package punkt is already up-to-date!\n", "Map: 100%|██████████| 1500/1500 [00:02<00:00, 612.69 examples/s]\n", "Loading checkpoint shards: 100%|██████████| 2/2 [00:02<00:00, 1.24s/it]\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Generating predictions...\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "/home/jovyan/Falcon1B/lib/python3.11/site-packages/transformers/generation/configuration_utils.py:695: UserWarning: `num_beams` is set to 1. However, `length_penalty` is set to `0.8` -- this flag is only used in beam-based generation modes. You should set `num_beams>1` or unset `length_penalty`.\n", " warnings.warn(\n", "Featurizing p: 100%|██████████| 1498/1498 [00:42<00:00, 34.94it/s]\n", "Featurizing q: 100%|██████████| 1498/1498 [00:42<00:00, 35.22it/s]\n", "WARNING clustering 2996 points to 150 centroids: please provide at least 5850 training points\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Evaluation Complete. Results saved to Complete_Evaluation_Results/Granite3.2-2B/BF16/FP16/Granite3.2-2B-BF16-lora-FP16-Evaluation_Results.json\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAArcAAAIjCAYAAAAZajMiAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjEsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvc2/+5QAAAAlwSFlzAAAPYQAAD2EBqD+naQAAf0ZJREFUeJzt3XlYVGXjPvB7GGZgWIZhB5FNAcVdsJS0VRJNrays3nDXLIVKbdMWtXpfzUrzzVxa3ErN0qxsM83UytzCDRUVVxTZ93WGmTm/P/w533ecIRWB5zjen+uaSznLPPfAEW8OZ56jkCRJAhERERGRA3ASHYCIiIiIqLGw3BIRERGRw2C5JSIiIiKHwXJLRERERA6D5ZaIiIiIHAbLLRERERE5DJZbIiIiInIYLLdERERE5DBYbomIiIjIYbDcEhHdZBQKBaZPny46BhFRk2C5JSLZW7ZsGRQKBf7+++/rfq7q6mpMnz4dW7duvf5gMjB9+nQoFAoUFhbaXR8REYEBAwZc9zirVq3C3Llzr/t5iIiamrPoAEREzam6uhpvvPEGAOCuu+4SG0aQmpoaODtf27f/VatW4dChQ5gwYULThCIiaiQ8c0tEdJNxdXW95nIrWlVVlegIRHSDYLklIodgMBgwdepUxMfHw8vLC+7u7rj99tuxZcsWyzZnzpyBv78/AOCNN96AQqGwuf706NGjeOSRR+Dj4wNXV1d069YN69evtxrr0mUS27dvx6RJk+Dv7w93d3cMGjQIBQUFNtl+/vln3HnnnfD09IRWq8Utt9yCVatWAQCmTZsGlUpld7+xY8dCp9Ohtra2MT5FFpe/5oqKCkyYMAERERFwcXFBQEAA7r33XuzduxfAxTPcP/74I86ePWv5nEVERFj2z8/Px+jRoxEYGAhXV1d07twZy5cvtxm3qKgIQ4cOhVarhU6nw/Dhw3HgwAEoFAosW7bMst2IESPg4eGBkydP4r777oOnpyeSk5MBAH/88QcGDx6MsLAwuLi4IDQ0FBMnTkRNTY3VWJeeIysrCwMGDICHhwdCQkIwf/58AEB6ejruueceuLu7Izw83PL1IKIb3431ozsRUT3Ky8vx6aef4l//+heefPJJVFRUYPHixUhKSsLu3bvRpUsX+Pv7Y+HChRg3bhwGDRqEhx56CADQqVMnAMDhw4fRs2dPhISEYPLkyXB3d8dXX32FBx98EF9//TUGDRpkNeYzzzwDb29vTJs2DWfOnMHcuXORmpqKL7/80rLNsmXLMGrUKLRv3x5TpkyBTqfDvn37sGHDBjzxxBMYOnQo3nzzTXz55ZdITU217GcwGLB27Vo8/PDDcHV1veLrLy4utrvcbDZfcd+nn34aa9euRWpqKtq1a4eioiL8+eefyMjIQFxcHF599VWUlZXh/PnzeP/99wEAHh4eAC5e4nDXXXfhxIkTSE1NRWRkJNasWYMRI0agtLQUzz33nCXHwIEDsXv3bowbNw5t27bFd999h+HDh9vNZDQakZSUhF69euG9996Dm5sbAGDNmjWorq7GuHHj4Ovri927d2PevHk4f/481qxZY/UcJpMJ/fr1wx133IF33nkHK1euRGpqKtzd3fHqq68iOTkZDz30EBYtWoRhw4YhISEBkZGRV/x8EZHMSUREMrd06VIJgLRnz556tzEajZJer7daVlJSIgUGBkqjRo2yLCsoKJAASNOmTbN5jt69e0sdO3aUamtrLcvMZrN02223SdHR0TZ5EhMTJbPZbFk+ceJESalUSqWlpZIkSVJpaank6ekpde/eXaqpqbEa63/3S0hIkLp37261ft26dRIAacuWLfW+ZkmSpGnTpkkA/vHRv39/q30uf/1eXl5SSkrKP47Tv39/KTw83Gb53LlzJQDSihUrLMsMBoOUkJAgeXh4SOXl5ZIkSdLXX38tAZDmzp1r2c5kMkn33HOPBEBaunSpZfnw4cMlANLkyZNtxquurrZZNnPmTEmhUEhnz561eY4ZM2ZYlpWUlEgajUZSKBTS6tWrLcuPHj1a7zFBRDceXpZARA5BqVRCrVYDuHiWsLi4GEajEd26dbP8ev2fFBcX47fffsOjjz6KiooKFBYWorCwEEVFRUhKSkJmZiays7Ot9hk7diwUCoXl49tvvx0mkwlnz54FAGzatAkVFRWYPHmyzdnX/91v2LBh2LVrF06ePGlZtnLlSoSGhuLOO++8qtf/9ddfY9OmTTaPwMDAK+6r0+mwa9cuXLhw4arG+l8//fQTgoKC8K9//cuyTKVS4dlnn0VlZSW2bdsGANiwYQNUKhWefPJJy3ZOTk5ISUmp97nHjRtns0yj0Vj+XlVVhcLCQtx2222QJAn79u2z2X7MmDGWv+t0OrRp0wbu7u549NFHLcvbtGkDnU6HU6dOXeWrJiI542UJROQwli9fjtmzZ+Po0aOoq6uzLL+aXzWfOHECkiTh9ddfx+uvv253m/z8fISEhFg+DgsLs1rv7e0NACgpKQEAS1nt0KHDP4792GOPYcKECVi5ciWmTp2KsrIy/PDDD5g4caJVCf4nd9xxB/z8/GyWX80lDe+88w6GDx+O0NBQxMfH47777sOwYcPQqlWrK+579uxZREdHw8nJ+lxJbGysZf2lP4ODgy2XF1wSFRVl93mdnZ3RsmVLm+VZWVmYOnUq1q9fb/k8X1JWVmb1saurq+Ua60u8vLzQsmVLm8+rl5eXzfMR0Y2J5ZaIHMKKFSswYsQIPPjgg3jxxRcREBAApVKJmTNnWp0Rrc+la1NfeOEFJCUl2d3m8iKmVCrtbidJ0jVl9/b2xoABAyzldu3atdDr9RgyZMg1PU9DPfroo7j99tvxzTffYOPGjXj33Xcxa9YsrFu3Dv369WuWDJdzcXGxKcwmkwn33nsviouL8fLLL6Nt27Zwd3dHdnY2RowYYXN9cX1fn8b6uhGRPLHcEpFDWLt2LVq1aoV169ZZnZWbNm2a1Xb1nQm9dJZSpVIhMTGxUTK1bt0aAHDo0KF6z1BeMmzYMDzwwAPYs2cPVq5cia5du6J9+/aNkuNqBAcHY/z48Rg/fjzy8/MRFxeH//znP5ZyW9/nLTw8HAcPHoTZbLYqo0ePHrWsv/Tnli1bUF1dbXX29sSJE1edMT09HcePH8fy5csxbNgwy/JNmzZd/QslIofHa26JyCFcOhv3v2ffdu3ahR07dlhtd6lYlZaWWi0PCAjAXXfdhY8++gg5OTk2z29vqq4r6dOnDzw9PTFz5kyb6bwuP0vYr18/+Pn5YdasWdi2bVuznbU1mUw2v84PCAhAixYtoNfrLcvc3d1ttgOA++67D7m5uVYzRBiNRsybNw8eHh6Wa4aTkpJQV1eHTz75xLKd2Wy2TM11Nex9jSVJwn//+9+rfg4icnw8c0tEN4wlS5Zgw4YNNsufe+45DBgwAOvWrcOgQYPQv39/nD59GosWLUK7du1QWVlp2Vaj0aBdu3b48ssvERMTAx8fH3To0AEdOnTA/Pnz0atXL3Ts2BFPPvkkWrVqhby8POzYsQPnz5/HgQMHrimvVqvF+++/jzFjxuCWW27BE088AW9vbxw4cADV1dVWc8GqVCo8/vjj+PDDD6FUKq3eoNWUKioq0LJlSzzyyCPo3LkzPDw88Ouvv2LPnj2YPXu2Zbv4+Hh8+eWXmDRpEm655RZ4eHhg4MCBGDt2LD766COMGDECaWlpiIiIwNq1a7F9+3bMnTsXnp6eAIAHH3wQt956K55//nmcOHECbdu2xfr16y1TmF3NtcVt27ZF69at8cILLyA7OxtarRZff/01r5UlImviJmogIro6l6bequ9x7tw5yWw2SzNmzJDCw8MlFxcXqWvXrtIPP/wgDR8+3GYKq7/++kuKj4+X1Gq1zRRQJ0+elIYNGyYFBQVJKpVKCgkJkQYMGCCtXbvWJs/lU5Nt2bLF7vRd69evl2677TZJo9FIWq1WuvXWW6UvvvjC5nXu3r1bAiD16dPnqj83l6YCKygosLs+PDz8H6cC0+v10osvvih17txZ8vT0lNzd3aXOnTtLCxYssNqnsrJSeuKJJySdTicBsPqc5uXlSSNHjpT8/PwktVotdezY0Wpqr0sKCgqkJ554QvL09JS8vLykESNGSNu3b5cAWE3NNXz4cMnd3d3u6zly5IiUmJgoeXh4SH5+ftKTTz4pHThwwO50Yvae484775Tat29/VZ8nIroxKSSJV9ATEcnBgQMH0KVLF3z22WcYOnSo6DjN4ttvv8WgQYPw559/omfPnqLjEJED4DW3REQy8cknn8DDw8Ny5zRHc/ktck0mE+bNmwetVou4uDhBqYjI0fCaWyIiwb7//nscOXIEH3/8seX2sI7omWeeQU1NDRISEqDX67Fu3Tr89ddfmDFjhtXNGYiIrgcvSyAiEiwiIgJ5eXlISkrC559/bnkTlqNZtWoVZs+ejRMnTqC2thZRUVEYN24cUlNTRUcjIgfCcktEREREDoPX3BIRERGRw2C5JSIiIiKHwTeU4eJdci5cuABPT8+rmkiciIiIiJqXJEmoqKhAixYtrG73fTmWWwAXLlxAaGio6BhEREREdAXnzp1Dy5Yt613PcgtY3pl87tw5aLXaJh/v2Ikv0WbvWByL+xhtKtsAd94JbNsGdOnS5GMTEd0IcvfnYumdSzFy20gEdQkSHYeILrN//37ceeed2LZtG7o0U38pLy9HaGjoFWeUYbnF/93TXKvVNku59fBwg9bt//8Jj0sLgWYYm4joRlDlUQVXuMLTw7NZvi8T0bXx8PCw/Nnc/0avdAkp31BGRERERA6D5ZaIiIiIHAbLLRERERE5DN6hDBcvUPby8kJZWVmzXDdiMFSiqCQDvt6xUEMN5OcDAQGAWt3kYxMR3QhMBhOq8qvgHuAOpVopOg45OEmSYDQaYTKZREe5YRgMBhQXF8PHxwfqRuovSqUSzs7O9V5Te7V9jW8oE0Ct9kBw4C3/t+AfprMgIroZKdVKaFvyjWTU9AwGA3JyclBdXS06yg0pOzu7UZ/Pzc0NwcHB11WYWW4FyMreiuzfhyDkjhUI04cBL78MzJoFtGolOhoRkSyUnCrBry//isRZifBu5S06Djkos9mM06dPQ6lUokWLFlCr1byZ01UyGAzIzc1FUFBQo5y5lSQJBoMBBQUFOH36NKKjo//xRg3/hOVWgKqqbCRI2cioygYqtcDatcCUKaJjERHJRm1pLY6sPYJeU3qJjkIOzGAwwGw2IzQ0FG5ubqLj3FBMJhPKy8sREhICV1fXRnlOjUYDlUqFs2fPwmAwNPh5+YYyIiIiuqk19AwhNb7G+Frwq0lEREREDoPlloiIiIgcBsutAL4+HbHVqw98fToCLVoAM2Zc/JOIiAAAni08cc+Me+DZ4p/vIU9EV3bXXXdhwoQJjfqcarUaISEhjTYNWGNiuRUgwK8T7ur/CwL8OgFBQRffTBYUJDoWEZFseAR54PYpt8MjyEN0FCJZGjFiBBQKBZ5++mmbdSkpKVAoFBgxYgQAYN26dXjrrbcadXyVSoXg4GCoVKpGfd7GwHIrQFn5Weza9SrKys8CpaXA+vUX/yQiIgAXZ0s4tv4YaktrRUchkq3Q0FCsXr0aNTU1lmW1tbVYtWoVwsLCLMt8fHzg6dm4vwUxGo0oLS2F0Whs1OdtDCy3AlzI/RPdT87Ahdw/gVOngAceuPgnEREBuDjP7eoHVqPkVInoKESyFRcXh9DQUKxbt86ybN26dQgLC0PXrl0tyy6/LCEiIgIzZszAqFGj4OnpibCwMHz88cfXNLZer8eJEyeg1+uv+3U0Ns5zS0RERHSZnIoc5FTmWC3zdvVGpHckao21OFJwxGafuOA4AMCxwmOoqquyWhehi4CPxgcFVQU4V37Oal2wRzCCPYMblHPUqFFYunQpkpOTAQBLlizByJEjsXXr1n/cb/bs2XjrrbfwyiuvYO3atRg3bhzuvPNOtGnTpkE55ITlloiIiOgyH6V9hDe2vWG1LLljMlY8tALny88j/uN4m32kaRIAYMR3I7Dz/E6rdZ8P+hxDOg3BV4e/QurPqVbrpt05DdPvmt6gnEOGDMGUKVNw9uxZAMD27duxevXqK5bb++67D+PHjwcAvPzyy3j//fexZcsWllsiIiIiR/RU/FO4v839Vsu8XS/eCrqltiXSxqbVu++yB5bZPXMLAI+2fxQJoQlW64I9GnbWFgD8/f3Rv39/LFu2DJIkoX///vDz87vifp06dbL8XaFQICgoCPn5+Q3OIScstwKoVJ44YXKBSuUJuLoC7dpd/JOIiAAAzq7O8G/nD2dX/jdFYgR71n+pgKuzq+USBHva+NV/9tPf3R/+7v7Xne9/jRo1CqmpF88Gz58//6r2uXyWA4VCAbPZfNVjOjk5wdXVVZZ3d+N3DQGiIu9HlvI4CgsLsRe1wOefA7W1wN69TTamn5+f1TsniYjkzL+dP8YfHi86BtENoW/fvjAYDFAoFEhKSmqWMTUaDTp06NAsY10rllsBsrKy0DY2FjXV1c02psbNDUczMlhwiYiIHIxSqURGRobl7zc7llsBDh9didxF1XgpZzQijLfgubcm4b+vz0F2RFSTjJd/OhNfvTYOhYWFLLdEdEPI3Z+LpXcsxcjfRyKoC29yQ3QlWq22Wcerrq7G0aNH0bZtW7i5uTXr2FfCciuAJJmgVQI6/wAEukRBU1uNwIgoILaz6GhERLIgmSUYKgyQzJLoKESytGzZsn9c/+2331r+fvnMCWfOnLHZfv/+/dc0viRJMJvNkCT5/RuV31XAREREREQNxHJLRERERA6D5ZaIiIiIHAbLrQDeXgmIex+oUsWjKCIaS1f+iqKIaNGxiIhkw6+tH8amjYVf2ytPRk9Ezc/V1RWxsbFwleE8/Sy3AriovbHvbwBOOhg1bsiL7QyjRl7vNCQiEknlpkJwXDBUbqorb0xEzU6pVMLd3V2WU4+x3ApQUn4AH04DVHWHoc05j3tnvgRtznnRsYiIZKMsqww/pvyIsqwy0VGIyA69Xo+zZ89Cr9eLjmKD5VYAvT4bKTGA2pwLTWkR4tcshaa0SHQsIiLZqC6sxt8L/kZ1YfPd7IaIrp7RaERBQQGMRqPoKDZYbomIiIjIYbDcEhEREZHDYLklIiIiIofBciuAxjUcc44ABqcQVPv4Y3fy06j28Rcdi4hINtwD3NFjYg+4B7iLjkIkW+fOncOoUaPQokULqNVqhIeH47nnnkNR0f+9j2fdunXo06cPfH19oVAorvk2u/VRqVQIDAyESiW/GU1YbgXw8myP5/8D1KnaoiKwBX57/i1UBLYQHYuISDa0LbVImpMEbUut6ChEsnTq1Cl069YNmZmZ+OKLL3DixAksWrQImzdvRkJCAoqLiwEAVVVV6NWrF2bNmtWo46vVaoSGhkKtVjfq8zYGZ9EBbka1hkL0uA1QmIuhqq6Ef2YGCqJjUefmIToaEZEsGCoNyEvPQ2DHQKg95PefJ5FoKSkpUKvV2LhxIzQaDQAgLCwMXbt2RevWrfHqq69i4cKFGDp0KADgzJkzjTq+yWRCTU0NNBqN7Oa65ZlbAUrLdmNHCuBWtw8+Z09i2Mj74HP2pOhYRESyUXS8CEtuW4Ki45wmkQSpyQGK91o/Kk9fXGeqtV1XvPf/9i0/ZrtOf/FMKmoLbNfV5FxTtOLiYvzyyy8YP368pdheEhQUhOTkZHz55ZeQJOl6PgP/qLa2FkePHkVtbW2TjdFQPHNLREREdLnMj4BDb1gvi0gGblsBVJ8HNsTb7vPE/y+TO0YARTut1yV8DkQOAbK+Av5OtV7XYRrQafrVR8vMhCRJiI2Ntbs+NjYWJSUlKCgoQEBAwFU/r6NguSUiIiK6XPRTQMv7rZepvS/+6dYS6JtW/74JywBjlfUy94iLf4Y9CvglWK/TBDco4pXOzMrxetjmwHJLREREdDlNcP2lU+kK+MTVv6+2Tf3rXP0vPq5DVFQUFAoFMjIyMGjQIJv1GRkZ8Pf3h06nu65xblS85lYAJyc1CoyApFDBrHRGtc4XZiV/ziAiusTJ2Qlufm5wcuZ/U0SX8/X1xb333osFCxagpqbGal1ubi5WrlyJESNGNGkGhUIBZ2dnKBSKJh2nIfhdQ4AA30QEDAeq1XegIKY9PvjtKApi2ouORUQkG4GdAvFiwYsI7BQoOgqRLH344YfQ6/VISkrC77//jnPnzmHDhg249957ERMTg6lTpwK4+Oaz/fv348iRIwCAY8eOYf/+/cjNzb2u8d3c3NClSxe4ubld92tpbCy3RERERDeY6Oho7NmzB61atcKjjz6K8PBw9OvXDzExMdi+fTs8PC5OL7p+/Xp07doV/fv3BwA8/vjj6Nq1KxYtWiQyfpNiuRWgoHgrMhcDrobt8Dt5FE/dfwv8Th4VHYuISDbyD+fjg6gPkH84X3QUItmKiIjAsmXLkJubC7PZjKlTp2Ljxo04ePCgZZsRI0ZAkiSbx/Tp069r7JqaGqSnp9tcFiEHvNBTAJOpGlGugFJfC6VBD+/zZ6A06EXHIiKSDZPehJKTJTDpTaKjEN0w3njjDURERGDnzp249dZb4eTUdOcwzWYz9Ho9zGZzk43RUCy3RERERA5i5MiRoiMIx8sSiIiIiMhhsNwSERERkcNguRXASxuPpM+AGudOKAlthS8//BIloa1ExyIikg2fKB8kb0iGT5SP6ChEZIerqyuio6Ph6uoqOooNllsBNC6B2PgLYFb6w+DhidO33QODh6foWEREsuGidUFUUhRctC6ioxCRHUqlEl5eXlAqlaKj2GC5FaCs8iimTQKcjSfgXpCLXovegXvB9U2mTETkSCpyKrB1+lZU5FSIjkJEdhgMBly4cAEGg0F0FBsstwLU1JzC9HjAxXQWHoV56PXxu/AozBMdi4hINipzKrHtjW2ozKkUHYWI7Kirq8OFCxdQV1cnOooNllsiIiIichgst0RERERkZevWrVAoFCgtLQUALFu2DDqdTmimq8VyS0RERHSDGTFiBBQKBZ5++mmbdSkpKVAoFBgxYkSjjffYY4/h+PHjjfZ8TYnlVgC1SyBWZAF1Tn6o1epwqN8jqNXqRMciIpINV29XdEzuCFdv+U0zRCQXoaGhWL16NWpqaizLamtrsWrVKoSFhTXqWBqNBgEBAZaPnZ2d4ePjA2dn+d3sluVWAB9tPIZOAQyqzigLCccP/1mIspBw0bGIiGTDO9IbD614CN6R3qKjEMlWXFwcQkNDsW7dOsuydevWISwsDF27drUsM5vNmDlzJiIjI6HRaNC5c2esXbvW6rl++uknxMTEQKPR4O6778aZM2es1l9+WcL58+cxceJEhIWFwcPDA7fccgt+/fVXq30iIiIwY8YMjBo1Cp6enggLC8PHH3/ceJ+AerDcCmCoK0frGADmSij1tdBlnYJSXys6FhGRbBhrjSg+UQxjrVF0FLpZ5eQAe/daP06fvriuttZ23d69/7fvsWO264qLL64rKLBdl5PT4JijRo3C0qVLLR8vWbIEI0eOtNpm5syZ+Oyzz7Bo0SIcPnwYEydOxJAhQ7Bt2zYAwLlz5/DQQw9h4MCB2L9/P8aMGYPJkyf/47jl5eW49957sWnTJuzbtw99+/bFwIEDkZWVZbXd7Nmz0a1bN+zbtw/jx4/HuHHjcOzYsQa/3qvBcitAcemfODENcK/bBb9Tx/D0g93hd6ppv9BERDeSgiMFmBc9DwVHCkRHoZvVRx8B8fHWj9dfv7ju/HnbdfHx/7fviBG263766eK6r76yXffRRw2OOWTIEPz55584e/Yszp49i+3bt2PIkCGW9Xq9HjNmzMCSJUuQlJSEVq1aYcSIERgyZAg++v/jLly4EK1bt8bs2bPRpk0bJCcnX/F63ZiYGPTo0QOtW7dGdHQ03nrrLbRu3Rrr16+32u6+++7D+PHjERUVhZdffhl+fn7YsmVLg1/v1ZDfhRJEREREoj31FHD//dbLvP//ZTItWwJpafXvu2wZUFVlvSwi4uKfjz4KJCRYrwsObnBMf39/9O/fH8uWLYMkSejfvz/8/Pws60+cOIHq6mrce++9VvsZDAbLpQsZGRno3r271fqEyzNeprKyEnPnzsWePXuQl5cHo9GImpoamzO3nTp1svxdoVAgKCgI+fn5DXqtV4vlloiIiOhywcH1l05XVyAurv5927Spf52//8VHIxo1ahRSU1MBAPPnz7daV1l58UYoP/74I0JCQqzWubg0/PbWr7zyCrZu3Yp3330XHTp0gEajwSOPPGJzxzKVSmX1sUKhgNlsbvC4V4PlloiIiOgG1rdvXxgMBigUCiQlJVmta9euHVxcXJCVlYU777zT7v6xsbE2lxPs3LnzH8fcuXMnBgwYgPvvvx/u7u6orKy0eROaKCy3RERERDcwpVKJjIwMy9//l6enJ1544QVMnDgRZrMZvXr1QllZGbZv3w6tVovhw4fj6aefxuzZs/Hiiy9izJgxSEtLw7Jly/5xzNatW2PLli04ePAg3Nzc8Prrrzf5GdmrxTeUCRDkfx8UyUCVS2/kxXbG23sLkBfbWXQsIiLZCI4LxjRpGoLjGn4tItHNRKvVQqvV2l331ltv4fXXX8fMmTMRGxuLvn374scff0RkZCQAICwsDF9//TW+/fZbdO7cGYsWLcKMGTP+cbwPPvgAISEh6N27NwYOHIikpCTE/dOlGs1IIUmSJDqEaOXl5fDy8kJZWVm9B0Zj2rt3L+Lj45G68leENEOpzc44gA+TE5GWliabA4+IiEi02tpanD59GpGRkXB15Q1D5OCfviZX29d45laAwpLt+GsB4GrYDZ8zJzB0eD/4nDkhOhYRkWwUHivE4oTFKDxWKDoKEdlRW1uLjIwM1NbKb55+llsBjMYyJHgBSqkCqpoqhKT/DVVN1ZV3JCK6SdRV1eH8zvOoq6oTHYWI7DCZTKiqqoLJZBIdxQbLLRERERE5DJZbIiIiInIYLLdERERE5DBYbgXw9OiIId8DNc5tUdYiDN+/tQBlLcJExyIikg1dhA6DPh8EXYROdBQissPFxQWRkZHXdZezpsJyK4C7JhQrVwNmZQhqvbxxuP9g1Hp5i45FRCQbGh8NOg3pBI2PRnQUIrLD2dkZvr6+cHaW3/3AWG4FqKw+jfGjAaXxLDQlhYj7cjE0JZzuhojokqqCKuyevxtVBZxJhkiO6urqkJ+fj7o6+c1ownIrQGVVBubfA7iaTkCbm40+syZDm5stOhYRkWyUnyvHz6k/o/xcuegoRGSHwWBAVlYWDAaD6Cg2WG6JiIiIyGGw3BIRERHdgM6dO4dRo0ahRYsWUKvVCA8Px3PPPYeioiLLNtOnT0fbtm3h7u4Ob29vJCYmYteuXQJTNz2WWyIiIqIbzKlTp9CtWzdkZmbiiy++wIkTJ7Bo0SJs3rwZCQkJKC4uBgDExMTgww8/RHp6Ov78809ERESgT58+KCgoEPwKmo783uJ2E1A5++CXAsDo5QWDuwdO9bgLBncP0bGIiGRD7alG6z6tofZUi45CJEspKSlQq9XYuHEjNJqLs4qEhYWha9euaN26NV599VUsXLgQTzzxhNV+c+bMweLFi3Hw4EH07t27weMrlUpotVoolcrreh1NgeVWAF/vHoifAKSu7IaSsNb4asEa0ZGIiGTFN9oXQ34ZIjoG3cQqcipQmVNptczV2xXekd4w1hpRcMT2zGdwXDAAoPBYIeqqrGcR0EXooPHRoKqgyuaNkh7BHvAM9rzqbMXFxfjll1/wn//8x1JsLwkKCkJycjK+/PJLLFiwAAqFwrLOYDDg448/hpeXFzp37nzV49nj6uqKmJiY63qOpsJyK4DJbICnFyBJBihMJqhqqlGncYMkw59+iIhEMJvMqKuqg8pdBSclr6Cj5pf2URq2vbHNalnH5I54aMVDKD9fjo/jP7bZZ5o0DQDw3YjvcH7neat1gz4fhE5DOuHwV4fxc+rPVuvunHYn7pp+11Vny8zMhCRJiI2Ntbs+NjYWJSUlKCgoQEBAAH744Qc8/vjjqK6uRnBwMDZt2gQ/P7+rHs8eSZJgNpvh5ORkVaDlgOVWgIKiX1G+AJis/wMBx9UYmZyIpSt/RV7s9f0URUTkKPIO5OHj+I8xNm2s5WwYUXOKfyoebe5vY7XM1dsVAKBtqcXYtLH17vvAsgfsnrkFgPaPtkdoQqjVOo/ghl2aKEnSP65Xqy9e1nP33Xdj//79KCwsxCeffIJHH30Uu3btQkBAQIPGBYDq6mpkZGQgNjYW7u7uDX6epiD0x2GTyYTXX38dkZGR0Gg0aN26Nd566y2rL5YkSZg6dSqCg4Oh0WiQmJiIzMxMq+cpLi5GcnIytFotdDodRo8ejcrKysuHIyIiIroqnsGeCI4Ltnp4R168m6izq7PNuv/9IcyvjZ/Nukt323P3d7dZdy2XJABAVFQUFAoFMjIy7K7PyMiAv78/dDrdxTHd3REVFYUePXpg8eLFcHZ2xuLFixvwWbkxCC23s2bNwsKFC/Hhhx8iIyMDs2bNwjvvvIN58+ZZtnnnnXfwwQcfYNGiRdi1axfc3d2RlJSE2tpayzbJyck4fPgwNm3ahB9++AG///47xo6t/ycqIiIiohuVr68v7r33XixYsAA1NTVW63Jzc7Fy5UqMGDGi3v3NZjP0en0TpxRHaLn966+/8MADD6B///6IiIjAI488gj59+mD37t0ALp61nTt3Ll577TU88MAD6NSpEz777DNcuHAB3377LYCLP51s2LABn376Kbp3745evXph3rx5WL16NS5cuCDw1RERERE1jQ8//BB6vR5JSUn4/fffce7cOWzYsAH33nsvYmJiMHXqVFRVVeGVV17Bzp07cfbsWaSlpWHUqFHIzs7G4MGDRb+EJiO03N52223YvHkzjh8/DgA4cOAA/vzzT/Tr1w8AcPr0aeTm5iIxMdGyj5eXF7p3744dO3YAAHbs2AGdTodu3bpZtklMTISTk1O9kxTr9XqUl5dbPYiIiIhuFNHR0dizZw9atWqFRx99FOHh4ejXrx9iYmKwfft2eHh4QKlU4ujRo3j44YcRExODgQMHoqioCH/88Qfat28v+iU0GaFvKJs8eTLKy8vRtm1bKJVKmEwm/Oc//0FycjKAi6fWASAwMNBqv8DAQMu63NxcmwuinZ2d4ePjY9nmcjNnzsQbb7zR2C/nqvn53A3/YcBjc3qiIKod/rs5A3oPL2F5iIjkJqBjAF7IfwGuOlfRUYhkKyIiAsuWLbN8PG3aNMyZMwcHDx5Ejx494OrqinXr1jXJ2BqNBp07d5blPLdCz9x+9dVXWLlyJVatWoW9e/di+fLleO+997B8+fImHXfKlCkoKyuzPM6dO9ek413OWalBYT6gULjCrFKhxtsPZpWqWTMQEcmZUqWEu787lCr5/cdJJFdvvPEGPvjgA+zcuRNms7lJx3JycoJKpYKTk/ym6hOa6MUXX8TkyZPx+OOPo2PHjhg6dCgmTpyImTNnArg4ETEA5OXlWe2Xl5dnWRcUFIT8/Hyr9UajEcXFxZZtLufi4gKtVmv1aE5FZbvx3WzAxbAPunOn8fCEIdCdO92sGYiI5Kz4ZDG+uP8LFJ8sFh2F6IYycuRITJgwoclLZ21tLTIzM63e4C8XQsttdXW1zSdfqVRaftqIjIxEUFAQNm/ebFlfXl6OXbt2ISEhAQCQkJCA0tJSpKWlWbb57bffYDab0b1792Z4FdeuzlCI+4MAZ6kYLpXliP79F7hU8rpfIqJL9GV6HP/+OPRljvuObqIbmclkQllZGUwmk+goNoRecztw4ED85z//QVhYGNq3b499+/Zhzpw5GDVqFABAoVBgwoQJ+Pe//43o6GhERkbi9ddfR4sWLfDggw8CuHgXjr59++LJJ5/EokWLUFdXh9TUVDz++ONo0aKFwFdHRERERM1NaLmdN28eXn/9dYwfPx75+flo0aIFnnrqKUydOtWyzUsvvYSqqiqMHTsWpaWl6NWrFzZs2ABX1/97k8HKlSuRmpqK3r17w8nJCQ8//DA++OADES+JiIiIiAQSWm49PT0xd+5czJ07t95tFAoF3nzzTbz55pv1buPj44NVq1Y1QUIiIiIiupHI7y1uNwF39xhM+hPQK1uhIiAYmye9iYoA3judiOgSzxBP9JndB54h13ZbUiJqHmq1Gi1btoRarRYdxQbLrQCeblF4fyFgdI5EtW8A9gwZh2rfgCvvSER0k/AI9EDCpAR4BHqIjkJEdqhUKgQFBUElw6lMWW4FqK7JxiMPAU6mHLiUl6LNpu/gUl4qOhYRkWzUlNTg8JrDqCmpER2FiOy4NO2q0WgUHcUGy60A5ZUHsOZhQGM8Al32WQx6eQx02WdFxyIiko3S06VY++halJ4uFR2FiOzQ6/U4deoU9Hr5TdfHcktERER0gxkxYgQUCgWefvppm3UpKSlQKBQYMWJE8weTAZZbIiIiohtQaGgoVq9ejZqa/7t8p7a2FqtWrUJYWJjAZGKx3BIRERHdgOLi4hAaGop169ZZlq1btw5hYWHo2rWrZdmGDRvQq1cv6HQ6+Pr6YsCAATh58qRl/WeffQYPDw9kZmZalo0fPx5t27ZFdXV187yYRiR0ntublbPSE3srAJOLG4wuGuS27Qiji0Z0LCIi2XDWOCOoaxCcNfxvisTIyclBTk6O1TJvb29ERkaitrYWR44csdknLi4OAHDs2DFUVVVZrYuIiICPjw8KCgpw7tw5q3XBwcEIDm7YlKCjRo3C0qVLkZycDABYsmQJRo4cia1bt1q2qaqqwqRJk9CpUydUVlZi6tSpGDRoEPbv3w8nJycMGzYMP/zwA5KTk/HXX3/hl19+waeffoodO3bAzc3N7rhOTk5wc3ODk5P8zpPyu4YAfj63I/5eIHVlAopaxWDZqt9ERyIikhX/WH88tfcp0THoJvbRRx/hjTfesFqWnJyMFStW4Pz584iPj7fZR5IkABevh925c6fVus8//xxDhgzBV199hdTUVKt106ZNw/Tp0xuUc8iQIZgyZQrOnr34xvTt27dj9erVVuX24YcfttpnyZIl8Pf3x5EjR9ChQwfL6+3UqROeffZZrFu3DtOnT7f7Gi/RaDRo165dgzI3NZZbIiIioss89dRTuP/++62WeXt7AwBatmyJtLS0evddtmyZ3TO3APDoo48iISHBal1Dz9oCgL+/P/r3749ly5ZBkiT0798ffn5+VttkZmZi6tSp2LVrFwoLC2E2mwEAWVlZlnLr7e2NxYsXIykpCbfddhsmT57c4EyisdwKkFe4AbWfA9P0vyHwqAJDh/fD58t/Rl7bTqKjERHJQs6+HCzusRijd45GcFfewZGa3z9dKuDq6mq5BMGeNm3a1LvO398f/v7+153vf40aNcpyNnj+/Pk26wcOHIjw8HB88sknaNGiBcxmMzp06ACDwWC13e+//w6lUomcnBxUVVXB07P+OwRWV1cjIyMDsbGx9V66IIr8LpS4CUiSGS5OACABkgTnOgPw/3+VQUREACTAZDAB/NZIdEV9+/aFwWBAXV0dkpKSrNYVFRXh2LFjeO2119C7d2/ExsaipKTE5jn++usvzJo1C99//z08PDxsLp24nCRJlofc8MwtERER0Q1MqVQiIyPD8vf/5e3tDV9fX3z88ccIDg5GVlaWzSUHFRUVGDp0KJ599ln069cPLVu2xC233IKBAwfikUceabbX0Vh45paIiIjoBqfVaqHVam2WOzk5YfXq1UhLS0OHDh0wceJEvPvuu1bbPPfcc3B3d8eMGTMAAB07dsSMGTPw1FNPITs7u1nyNyaeuSUiIiK6wSxbtuwf13/77beWvycmJtpMXfa/lxMsWbLEZv9JkyZh0qRJ15VRFJ65FcBH1xPt3waqVbegKDIGn675A0WRMaJjERHJhl+sH8YdGge/WL8rb0xEzU6j0aB9+/bQaOQ3Tz/LrQBqlReOpAOSkxZGVw0KW7eF0VV+BwcRkSgqjQoB7QOg0qhERyEiO5ycnKDRaGR5Ewf5JboJlJYfwCdvAaq6Q9BeOId+b06A9sK5K+9IRHSTKD1bivVj1qP0bKnoKERkh16vx5kzZ6DX60VHscFyK0CtPhtjWgFqcx40ZcXo/O1KaMqKRcciIpKNmqIa7Fu8DzVFNaKjEJEdRqMRhYWFMBqNoqPYYLklIiKim5oc52q9WTXG14LlloiIiG5KKtXFa7qrq6sFJ6FLLn0tLn1tGoJTgREREdFNSalUQqfTIT8/HwDg5uYGhUIhONWN4dK1tnq93ubGEQ0hSRKqq6uRn58PnU53Xc/JciuAxjUCMw8A+vahqPINwI6Rz6LKN0B0LCIi2XAPdEfPyT3hHuguOgo5uKCgIACwFFy6OkajEQaDAdnZ2XB2brw6qdPpLF+ThmK5FcDLsx1eeQdIXRmDyoBgbHvmddGRiIhkRRuiReLMRNEx6CagUCgQHByMgIAA1NXViY5zU1OpVI1yFpjlVoBaQwHuvBtQmAuhrqpEUMYB5MZ2hsHdQ3Q0IiJZ0FfokZOWg+D4YLh4uoiOQzcBpVLZKMXqZlFRUYG0tDTEx8fD09NTdBwrfEOZAKVle7B1DOBWdwDeWSfxxNgH4Z11UnQsIiLZKM4sxvK7l6M4k9MkEslRZmYm7r77bmRmZoqOYoPlloiIiIgcBsstERERETkMllsiIiIichgstwI4ObngvAEwK1QwO6tQHhAMs3PDJysmInI0TioneIZ4wknF/6aI5EilUiEkJOS6brbQVDhbggABvr0R2gdIXXkHCqLbYcGGg6IjERHJSmDHQEw6P0l0DCKqR8eOHXH+/HnRMezij8RERERE5DBYbgXIL9qMc0sBjeF3+Gcewfi+neCfeUR0LCIi2chLz8OclnOQl54nOgoR2ZGeno6WLVsiPT1ddBQbLLcCmM16tFQDTlIdnIx10ObnwMnIu6IQEV1irjOjIrsC5jqz6ChEZEddXR2ys7NleVc3llsiIiIichgst0RERETkMFhuiYiIiMhhsNwKoPO6BXd9ClSrOqMkrDVWffwtSsJai45FRCQbPtE+GL5lOHyifURHISI7oqOjsWXLFkRHR4uOYoPlVgBXtT+2bQEkJz8Y3D2Q1a0nDO4eomMREcmGi6cLIu6KgIuni+goRGSHp6cn7rrrLnh6eoqOYoPlVoCyiiOY8RLgbDwOj/wc3DnvLXjk54iORUQkG+XZ5fh1yq8ozy4XHYWI7MjOzsaUKVOQnZ0tOooNllsBamrPYEpnwMV0Du5F+UhY+gHci/JFxyIiko2qvCpsf3s7qvKqREchIjvy8vLw9ttvIy9PfnNRs9wSERERkcNguSUiIiIih8FyS0REREQOg+VWAFeXEHx6CjA4BaLGywcHHkxGjRenuyEiukTjq0HX0V2h8dWIjkJEdvj6+mL06NHw9fUVHcWGs+gANyOdtjOefB1IXdkB5S1C8fPUuaIjERHJii5ch/s/vV90DCKqR3h4OD799FPRMezimVsBDHVlaNcRUJjL4VxbA7+TR+FcWyM6FhGRbNTV1CH/cD7qaupERyEiO2pqanD48GHU1Mivv7DcClBcuh2HJwNudXvge/o4xgy+Hb6nj4uORUQkG4UZhVjYYSEKMwpFRyEiOzIyMtChQwdkZGSIjmKD5ZaIiIiIHAbLLRERERE5DJZbIiIiInIYLLcCKBRO0JsBQAEoFDCq1IBCIToWEZF8KAClWgnwWyORLCkUCqjVaihk2F84FZgAgX594Zr0KlJX3oO8tp3w3q5s0ZGIiGQluGswXtO/JjoGEdWja9eu0Ov1omPYxTO3REREROQwWG4FKCz+A2mLAFfDDvieOo4RT9wD31OcCoyI6JKCjAJ8FPcRCjIKREchIjsyMjIQFxfHqcDoIqOpAnGegFKqhrO+BkFH0+Gsl98kyEREohhrjMjdlwtjjVF0FCKyo6amBvv27eNNHIiIiIiImhLLLRERERE5DJZbIiIiInIYLLcCaD06Y/DXQI1zO5SGhOObWZ+iNCRcdCwiItnQRerwyFePQBepEx2FiOyIjIzEV199hcjISNFRbLDcCuCmCcHadYBZGQy9Vodj9z4AvVYnOhYRkWxovDVoP7g9NN4a0VGIyA5vb28MHjwY3t7eoqPYYLkVoKL6BCaOA5yNp+FWlI9bViyEW1G+6FhERLJRmVeJHXN2oDKvUnQUIrIjLy8Pc+bMQV5enugoNlhuBaiqOo45vQAX0yl45ueg95yp8MzPER2LiEg2KrIrsPH5jajIrhAdhYjsyM7OxvPPP4/sbPndZZXlloiIiIgcBsstERERETkMllsiIiIichgstwKo1H5YnwsYFT7Qe2iReUcS9B5a0bGIiGTDxcsFMQNj4OLlIjoKEdnh5eWFgQMHwsvLS3QUG86iA9yMfL1uRfzzQOrKrigNjcTXc1eIjkREJCs+rX3wr/X/Eh2DiOrRunVrrF+/XnQMu3jmVgCjqQZ+AYAk1cKprg6akkI41dWJjkVEJBumOhOqCqpgqjOJjkJEdtTV1aGgoAB1MuwvLLcCFBZvQcH7gIdhO/xPHMFzvWPhf+KI6FhERLKRn56P9wLeQ3465wAnkqP09HQEBAQgPT1ddBQbLLdERERE5DBYbomIiIjIYbDcEhEREZHDYLklIiIiIofBciuAv28itOOBSvXtyI/pgDm/n0J+TAfRsYiIZCOwcyAml01GYOdA0VGIyI7OnTujrKwMnTt3Fh3FBue5FUDppEZFGaBQqCEplTB4eIqOREQkK05KJ7hoeQMHIrlSKpXQauV5AyqeuRWgqGQnNswFXAx/wzvrJB4dPxjeWSdFxyIiko2izCKsSFqBoswi0VGIyI7MzEwkJSUhMzNTdBQbLLcC1BmLkeQPOEtlUFdVotXOrVBXVYqORUQkG4YKA05uPAlDhUF0FCKyo6KiAhs3bkRFRYXoKDZYbomIiIjIYbDcEhEREZHDYLklIiIiIofBciuAh3ssUn4DapVRKA8KwcaX30Z5UIjoWEREsqEN1aLfh/2gDZXnu7GJbnahoaH48MMPERoaKjqKDZZbATzcIrFgMWByDkeNtx/2PjYaNd5+omMREcmGu787bk25Fe7+7qKjEJEd/v7+SElJgb+/v+goNoSX2+zsbAwZMgS+vr7QaDTo2LEj/v77b8t6SZIwdepUBAcHQ6PRIDEx0WbaieLiYiQnJ0Or1UKn02H06NGorJTv7ANVNeeQ/DjgZMqGa1kJ2v+4Bq5lJaJjERHJRk1xDQ6uOIia4hrRUYjIjuLiYqxYsQLFxcWio9gQWm5LSkrQs2dPqFQq/Pzzzzhy5Ahmz54Nb29vyzbvvPMOPvjgAyxatAi7du2Cu7s7kpKSUFtba9kmOTkZhw8fxqZNm/DDDz/g999/x9ixY0W8pKtSUZmOFQMBjfEovC5kYeDr4+F1IUt0LCIi2Sg9U4pvhn6D0jOloqMQkR1nzpzB0KFDcebMGdFRbAi9Q9msWbMQGhqKpUuXWpZFRkZa/i5JEubOnYvXXnsNDzzwAADgs88+Q2BgIL799ls8/vjjyMjIwIYNG7Bnzx5069YNADBv3jzcd999eO+999CiRYvmfVFEREREJIzQM7fr169Ht27dMHjwYAQEBKBr16745JNPLOtPnz6N3NxcJCYmWpZ5eXmhe/fu2LFjBwBgx44d0Ol0lmILAImJiXBycsKuXbvsjqvX61FeXm71ICIiIqIbn9Bye+rUKSxcuBDR0dH45ZdfMG7cODz77LNYvnw5ACA3NxcAEBgYaLVfYGCgZV1ubi4CAgKs1js7O8PHx8eyzeVmzpwJLy8vy0OO7/QjIiIiomsntNyazWbExcVhxowZ6Nq1K8aOHYsnn3wSixYtatJxp0yZgrKyMsvj3LlzTTre5ZydvbCjDDApPFGncUd2x26o0/AdwUREl6jcVWjZoyVU7irRUYjIDnd3d/To0QPu7vLrL0LLbXBwMNq1a2e1LDY2FllZF99cFRQUBADIy8uz2iYvL8+yLigoCPn5+VbrjUYjiouLLdtczsXFBVqt1urRnPy8e+K28UCt+lYUR0Th8+U/ozgiqlkzEBHJmV8bP4zeMRp+bThNIpEctWnTBjt27ECbNm1ER7EhtNz27NkTx44ds1p2/PhxhIeHA7j45rKgoCBs3rzZsr68vBy7du1CQkICACAhIQGlpaVIS0uzbPPbb7/BbDaje/fuzfAqiIiIiEguhJbbiRMnYufOnZgxYwZOnDiBVatW4eOPP0ZKSgoAQKFQYMKECfj3v/+N9evXIz09HcOGDUOLFi3w4IMPArh4prdv37548sknsXv3bmzfvh2pqal4/PHHZTtTQm7BT5BWAu76zQjMOIDJcf4IzDggOhYRkWzk7M3BG4o3kLM3R3QUIrJj7969UCgU2Lt3r+goNoROBXbLLbfgm2++wZQpU/Dmm28iMjISc+fORXJysmWbl156CVVVVRg7dixKS0vRq1cvbNiwAa6urpZtVq5cidTUVPTu3RtOTk54+OGH8cEHH4h4SUREREQkkNByCwADBgzAgAED6l2vUCjw5ptv4s0336x3Gx8fH6xataop4hERERHRDUT47XeJiIiIiBoLyy0REREROQyWWwF8dL0Q9QZQpeqOwlZtsOjbXShsJb+pNIiIRPFv549nMp+Bfzt/0VGIyI527dohMzPTZkpXORB+ze3NSK3S4uRxAE4eMLm4ojSslehIRESy4uzqDJ8oH9ExiKgerq6uiIqS5xz9PHMrQHF5Gj6fCajrDsAr+ywGvDoOXtlnRcciIpKNktMlWDdkHUpOl4iOQkR2nD59GkOGDMHp06dFR7HBciuAQZ+HIWGAylwI1/JSdPh5LVzLS0XHIiKSjdqSWqSvTEdtSa3oKERkR0lJCVauXImSEvn9AMpyS0REREQOg+WWiIiIiBwGyy0REREROQyWWwE0mlaYngboleGo9AvEn2NfRKVfoOhYRESy4RHsgTun3QmPYA/RUYjIjuDgYEybNg3BwcGio9hoULk9depUY+e4qXh5tMUbcwCjcxSq/IPw59Mvoco/SHQsIiLZ8Az2xF3T74JnsKfoKERkR3BwMKZPn+445TYqKgp33303VqxYgdpavpP1WtXo89AnCXAyFUBdWYHIv36DurJCdCwiItnQl+tx4pcT0JfrRUchIjvKy8vxyy+/oLy8XHQUGw0qt3v37kWnTp0wadIkBAUF4amnnsLu3bsbO5vDKitPwy/DAI3xILzPncJjqY/B+xzPhhMRXVJ8ohgr+65E8Yli0VGIyI4TJ06gb9++OHHihOgoNhpUbrt06YL//ve/uHDhApYsWYKcnBz06tULHTp0wJw5c1BQUNDYOYmIiIiIrui63lDm7OyMhx56CGvWrMGsWbNw4sQJvPDCCwgNDcWwYcOQk5PTWDmJiIiIiK7ousrt33//jfHjxyM4OBhz5szBCy+8gJMnT2LTpk24cOECHnjggcbKSURERER0Rc4N2WnOnDlYunQpjh07hvvuuw+fffYZ7rvvPjg5XezKkZGRWLZsGSIiIhozq8NQKt1wohYwObnCpHZBScsImNQuomMREcmG0kUJ79beULooRUchIjtcXFzQunVruLjIr780qNwuXLgQo0aNwogRI+qdAiIgIACLFy++rnCOyt/nLkTfC6Su7InC1m3x0fo9oiMREclKQPsAPHviWdExiKge7du3l+WbyYAGltvMzMwrbqNWqzF8+PCGPD0RERERUYM06JrbpUuXYs2aNTbL16xZg+XLl193KEeXX/Qr8pcDbobf4X/8MJ69py38jx8WHYuISDbyDubhXf93kXcwT3QUIrLj4MGD8Pf3x8GDB0VHsdGgcjtz5kz4+fnZLA8ICMCMGTOuO5SjM5sN8HcGFFIdnExGuJUWwclkFB2LiEg2zEYzqgurYTaaRUchIjuMRiMKCwthNMqvvzSo3GZlZSEyMtJmeXh4OLKysq47FBERERFRQzSo3AYEBNg9DX3gwAH4+vpedygiIiIiooZoULn917/+hWeffRZbtmyByWSCyWTCb7/9hueeew6PP/54Y2ckIiIiIroqDZot4a233sKZM2fQu3dvODtffAqz2Yxhw4bxmturoPO6FQlvAreM64ri8Nb4bOlPKA5vLToWEZFs+Mb4YtRfo+Abw98GEslRTEwM/vrrL8TExIiOYqNB5VatVuPLL7/EW2+9hQMHDkCj0aBjx44IDw9v7HwOyVXth51/Ad1SfFDn5oELnW8RHYmISFbUHmqEJoSKjkFE9fDw8EBCQoLoGHZd1+13Y2JiMHjwYAwYMIDF9hqUVRzG7FcBVd1ReOZdwD2zX4dn3gXRsYiIZKP8fDl+mfQLys+Xi45CRHacP38ekyZNwvnz50VHsdGgcmsymbB48WI88cQTSExMxD333GP1oH9WU3sWk9oBanM23IoLcOvKRXArLhAdi4hINqryq7Dz/Z2oyq8SHYWI7MjPz8f777+P/Px80VFsNOiyhOeeew7Lli1D//790aFDBygUisbORURERER0zRpUblevXo2vvvoK9913X2PnISIiIiJqsAZdlqBWqxEVFdXYWYiIiIiIrkuDyu3zzz+P//73v5AkqbHz3BRcXEIw/zhgcApCjc4XaYNHokbH6W6IiC5x83NDt/Hd4ObnJjoKEdnh5+eH8ePHw8/PT3QUGwqpAQ110KBB2LJlC3x8fNC+fXuoVCqr9evWrWu0gM2hvLwcXl5eKCsrg1arbfLx9u7di/j4eKSu/BUhsZ2bfLzsjAP4MDkRaWlpiIuLa/LxiIiIiBrb1fa1Bl1zq9PpMGjQoAaHu9npDSXo2g2AuRTONdXwPZOJoohoGDU8Q0FEBAB11XUoPFoIv7Z+ULmprrwDETWr6upqHD16FG3btoWbm7z6S4PK7dKlSxs7x02lpGwH9k4EJuvT4HtGh5HJiVi68lfkNcNZXCKiG0Hh0UJ8HP8xxqaNRXBcsOg4RHSZo0ePIj4+Xpa/FW7wTRyMRiN+/fVXfPTRR6ioqAAAXLhwAZWVlY0WjoiIiIjoWjTozO3Zs2fRt29fZGVlQa/X495774WnpydmzZoFvV6PRYsWNXZOIiIiIqIratCZ2+eeew7dunVDSUkJNBqNZfmgQYOwefPmRgtHRERERHQtGnTm9o8//sBff/0FtVpttTwiIgLZ2dmNEsyRKRRKlJsASeEEyckJencPSE4NvkKEiMjhKJwUUHuqoXDiHTCJ5MjJyQmenp5wkmF/aVC5NZvNMJlMNsvPnz8PT0/P6w7l6AL9kuCV9ApSV96N/DYd8f4fp0VHIiKSlaAuQZhSPkV0DCKqR5cuXVBeXi46hl0Nqtt9+vTB3LlzLR8rFApUVlZi2rRpvCUvEREREQnToHI7e/ZsbN++He3atUNtbS2eeOIJyyUJs2bNauyMDqegeBsOfQJoDDvge+oYRj/SC76njomORUQkGwVHCrCg/QIUHCkQHYWI7Dhy5Ajat2+PI0eOiI5io0GXJbRs2RIHDhzA6tWrcfDgQVRWVmL06NFITk62eoMZ2WcyVaG9G+Ckr4azvhb+p47BWV8rOhYRkWwYa40oOFIAY61RdBQisqO2thZHjhxBba38+kuDyi0AODs7Y8iQIY2ZhYiIiIjoujSo3H722Wf/uH7YsGENCkNEREREdD0aVG6fe+45q4/r6upQXV0NtVoNNzc3llsiIiIiEqJBbygrKSmxelRWVuLYsWPo1asXvvjii8bO6HC0nl1w/2qgxrk9SltGYO37n6O0ZYToWEREsuHdyhuPf/c4vFt5i45CRHa0atUK3333HVq1aiU6io1Gm3k3Ojoab7/9ts1ZXbLl5toC338PmJVB0Ht64cSdfaH39BIdi4hINlx1rmhzfxu46lxFRyEiO3Q6He6//37odDrRUWw06m0lnJ2dceHChcZ8SodUXnkck58BlMZTcC/MQ48lc+FemCc6FhGRbFTmVuKPmX+gMrdSdBQisiM3NxczZ85Ebm6u6Cg2GnTN7fr1660+liQJOTk5+PDDD9GzZ89GCebIqmtOYGYPYLL+NDwKWuGuD/+D0wl3o8ovUHQ0IiJZqLhQgd9e+Q1RSVHwCPIQHYeILnPhwgW88sorSEpKQlBQkOg4VhpUbh988EGrjxUKBfz9/XHPPfdg9uzZjZGLiIiIiOiaNajcms3mxs5BRERERHTdGvWaWyIiIiIikRp05nbSpElXve2cOXMaMoRDU6sDsCYbqAvwRa2nF44mDkQtZ0sgIrJw1bmi3SPtOFsCkUzpdDo88sgjspwtoUHldt++fdi3bx/q6urQpk0bAMDx48ehVCoRFxdn2U6hUDROSgfj49UN974EpK7sgrKWEfj2nSWiIxERyYp3K28MXjNYdAwiqkerVq2wZs0a0THsalC5HThwIDw9PbF8+XJ4e1+cYLukpAQjR47E7bffjueff75RQzoao6kaIWEApGo41RngXlyIKh8/mFVq0dGIiGTBZDChKr8K7gHuUKqVouMQ0WUMBgPy8/MREBAAtVpe/aVB19zOnj0bM2fOtBRbAPD29sa///1vzpZwFQqLt+L8TMDdsAP+JzKQ0q8z/E9kiI5FRCQb+Yfy8X7o+8g/lC86ChHZcejQIYSGhuLQoUOio9hoULktLy9HQUGBzfKCggJUVFRcdygiIiIiooZoULkdNGgQRo4ciXXr1uH8+fM4f/48vv76a4wePRoPPfRQY2ckIiIiIroqDbrmdtGiRXjhhRfwxBNPoK6u7uITOTtj9OjRePfddxs1IBERERHR1WpQuXVzc8OCBQvw7rvv4uTJkwCA1q1bw93dvVHDERERERFdi+u6iUNOTg5ycnIQHR0Nd3d3SJLUWLkcWoBvH7iMAqrUdyKvTUe8u/M88tp0FB2LiEg2groE4dXaVxHURV73rCeii7p06YLa2lp06dJFdBQbDSq3RUVF6N27N2JiYnDfffchJycHADB69GhOA3YVnJycYdADUDgDTk4wqV0AJ94sjojoEoWTAs4uzlA4cb50IjlycnKCi4sLnGTYXxqUaOLEiVCpVMjKyoKbm5tl+WOPPYYNGzY0WjhHVVSyA1vmAS6Gv+F99iSeePIBeJ89KToWEZFsFB0vwrK7lqHoeJHoKERkx/Hjx3HXXXfh+PHjoqPYaNA1txs3bsQvv/yCli1bWi2Pjo7G2bNnGyWYI6szluBeH2CDvgzq6kqEpf0FdXWl6FhERLJhqDTg7LazMFQaREchIjsqKyuxbds2VFbKr7806MxtVVWV1RnbS4qLi+Hi4nLdoYiIiIiIGqJB5fb222/HZ599ZvlYoVDAbDbjnXfewd13391o4YiIiIiIrkWDLkt455130Lt3b/z9998wGAx46aWXcPjwYRQXF2P79u2NnZGIiIiI6Ko06Mxthw4dcPz4cfTq1QsPPPAAqqqq8NBDD2Hfvn1o3bp1Y2d0OB7u7TBmI1DrHIPyoJb46fU5KA9qeeUdiYhuEl5hXhj4yUB4hXmJjkJEdoSFheGTTz5BWFiY6Cg2rvnMbV1dHfr27YtFixbh1VdfbYpMDs/DLQKLlwOpfUJR4+2Lg4OGio5ERCQrbn5uiBsTJzoGEdXDz88PY8aMER3Drms+c6tSqXDw4MGmyHLTqKw+g9HDAaXpHDQlRej0zefQlHC6GyKiS6oLq7H3072oLqwWHYWI7CgsLMSnn36KwsJC0VFsNOiyhCFDhmDx4sWNneWmUVl1BJ/2AVyNx6HNPY/73poEbe550bGIiGSjLKsM3z/5PcqyykRHISI7srKy8OSTTyIrK0t0FBsNekOZ0WjEkiVL8OuvvyI+Ph7u7u5W6+fMmdMo4YiIiIiIrsU1ldtTp04hIiIChw4dQlzcxWuhLr8zhULBWyUSERERkRjXVG6jo6ORk5ODLVu2ALh4u90PPvgAgYGBTRKOiIiIiOhaXNM1t5IkWX38888/o6qqqlED3QxUzt7YWgwYFV4wuHkgK/42GNw8RMciIpINtYca4XeGQ+2hFh2FiOzw8PDAnXfeCQ8P+fWXBl1ze8nlZZeujq93AuKfAVJXdkNJeGus+uQ70ZGIiGTFN8YXI7aOEB2DiOoRExODrVu3io5h1zWduVUoFDbX1PIa22tnNhuhdgEgGQGzGUqDHjCbRcciIpINySzBqDdCMvMkCpEcmc1m6PV6mGXYX67pzK0kSRgxYgRcXFwAALW1tXj66adtZktYt25d4yV0QPlFG6FfAkzWb0PgMWeMTE7E0pW/Ii+2s+hoRESykLs/Fx/Hf4yxaWMRHBcsOg4RXWb//v2Ij49HWlqaZZIBubimcjt8+HCrj4cMGdKoYYiIiIiIrsc1ldulS5c2VQ4iIiIiouvWoDuUERERERHJEcstERERETkM2ZTbt99+GwqFAhMmTLAsq62tRUpKCnx9feHh4YGHH34YeXl5VvtlZWWhf//+cHNzQ0BAAF588UUYjcZmTn9t/HzuQsspQJU6AQVRsZj/8wEURMWKjkVEJBsBHQIw8dxEBHQIEB2FiOzo0KEDzp07hw4dOoiOYkMW5XbPnj346KOP0KlTJ6vlEydOxPfff481a9Zg27ZtuHDhAh566CHLepPJhP79+8NgMOCvv/7C8uXLsWzZMkydOrW5X8I1cVa6ITsLgMINZpUaFYEtYFZxonIiokuUaiW0LbVQqpWioxCRHWq1Gi1btoRaLb/+IrzcVlZWIjk5GZ988gm8vb0ty8vKyrB48WLMmTMH99xzD+Lj47F06VL89ddf2LlzJwBg48aNOHLkCFasWIEuXbqgX79+eOuttzB//nwYDIZ6x9Tr9SgvL7d6NKfisr/x1TuAum4/vM6fwYMvjYLX+TPNmoGISM5KTpVgzeA1KDlVIjoKEdlx6tQpDB48GKdOnRIdxYbwcpuSkoL+/fsjMTHRanlaWhrq6uqslrdt2xZhYWHYsWMHAGDHjh3o2LEjAgMDLdskJSWhvLwchw8frnfMmTNnwsvLy/IIDQ1t5Ff1zwyGfAwOAVTmIrhWlKHtr9/DtaKsWTMQEclZbWktjqw9gtrSWtFRiMiO0tJSrF27FqWlpaKj2BBablevXo29e/di5syZNutyc3OhVquh0+mslgcGBiI3N9eyzf8W20vrL62rz5QpU1BWVmZ5nDt37jpfCRERERHJwTXNc9uYzp07h+eeew6bNm2Cq6trs47t4uJiucsaERERETkOYWdu09LSkJ+fj7i4ODg7O8PZ2Rnbtm3DBx98AGdnZwQGBsJgMNic7s7Ly0NQUBAAICgoyGb2hEsfX9qGiIiIiG4ewspt7969kZ6ejv3791se3bp1Q3JysuXvKpUKmzdvtuxz7NgxZGVlISEhAQCQkJCA9PR05OfnW7bZtGkTtFot2rVr1+yv6Wq5aaIwZSdQq4xEpX8Qtqa+ikp/lnEioks8W3jinhn3wLOFp+goRGRHixYtMGPGDLRo0UJ0FBvCLkvw9PS0mRvN3d0dvr6+luWjR4/GpEmT4OPjA61Wi2eeeQYJCQno0aMHAKBPnz5o164dhg4dinfeeQe5ubl47bXXkJKSIuvLDrQeMXh7HpC6shWq/AKxc9QE0ZGIiGTFI8gDt0+5XXQMIqpHUFAQpkyZIjqGXcJnS/gn77//PgYMGICHH34Yd9xxB4KCgrBu3TrLeqVSiR9++AFKpRIJCQkYMmQIhg0bhjfffFNg6iurrr2AgQMBJ1MuXCrKELVtA1w4WwIRkUVtaS2OrT/G2RKIZKq0tBTr16/nbAlXsnXrVsydO9fysaurK+bPn4/i4mJUVVVh3bp1NtfShoeH46effkJ1dTUKCgrw3nvvwdlZ2Anpq1JesR/rHwc0xsPQnT+DRyYOhY7z3BIRWZScKsHqB1ZznlsimTp16hQeeOABznNLRERERNSUWG6JiIiIyGGw3BIRERGRw2C5FUCpdMfhasCscIPRxRUFrdrA6NK8N7IgIpIzZ1dn+Lfzh7OrvN9DQXSzcnV1Rbt27Zr9RlxXg981BPD3uRMd7gVSVyagqFUbLF77p+hIRESy4t/OH+MPjxcdg4jq0a5dOxw+fFh0DLt45paIiIiIHAbLrQB5hb+g7DPAzbAFAcfSMfH2SAQcSxcdi4hINnL352KmdiZy9+eKjkJEduzfvx9arRb79+8XHcUGy60AkmSCVgkoJDMUZjNcqiqhMJtFxyIikg3JLMFQYYBklkRHISI7zGYzKioqYJZhf2G5JSIiIiKHwXJLRERERA6D5ZaIiIiIHAbLrQDeXgmIex+oUsWjKCIaS1f+iqKIaNGxiIhkw6+tH8amjYVfWz/RUYjIjrZt2yItLQ1t27YVHcUG57kVwEXtjX1/Az2ddDBq3JAX21l0JCIiWVG5qRAcFyw6BhHVw83NDXFxcaJj2MUztwKUlB/Ah9MAVd1haHPO496ZL0Gbc150LCIi2SjLKsOPKT+iLKtMdBQisiMrKwspKSnIysoSHcUGy60Aen02UmIAtTkXmtIixK9ZCk1pkehYRESyUV1Yjb8X/I3qwmrRUYjIjsLCQixYsACFhYWio9hguSUiIiIih8FyS0REREQOg+WWiIiIiBwGy60AGtdwzDkCGJxCUO3jj93JT6Pax190LCIi2XAPcEePiT3gHuAuOgoR2REQEICJEyciICBAdBQbnApMAC/P9nj+P0DqyraoCGyB355/S3QkIiJZ0bbUImlOkugYRFSPli1bYs6cOaJj2MUztwLUGgrR4zZAYS6GqroSLQ7sgaq6UnQsIiLZMFQacG7HORgqDaKjEJEdlZWV2LFjByor5ddfWG4FKC3bjR0pgFvdPvicPYlhI++Dz9mTomMREclG0fEiLLltCYqOc5pEIjk6fvw4brvtNhw/flx0FBsst0RERETkMFhuiYiIiMhhsNwSERERkcNguRXAyUmNAiMgKVQwK51RrfOFWcmJK4iILnFydoKbnxucnPnfFJEcOTs7w8/PD87O8usv8kt0EwjwTURAHyB15R0oiGmPD347KjoSEZGsBHYKxIsFL4qOQUT16NSpEwoKCkTHsIs/EhMRERGRw2C5FaCgeCsyFwOuhu3wO3kUT91/C/xO8uwtEdEl+Yfz8UHUB8g/nC86ChHZcfjwYURFReHw4cOio9hguRXAZKpGlCuglGqhNOjhff4MlAa96FhERLJh0ptQcrIEJr1JdBQiskOv1+PkyZPQ6+XXX1huiYiIiMhhsNwSERERkcNguSUiIiIih8FyK4CXNh5JnwE1zp1QEtoKX374JUpCW4mORUQkGz5RPkjekAyfKB/RUYjIjqioKGzYsAFRUVGio9hguRVA4xKIjb8AZqU/DB6eOH3bPTB4eIqORUQkGy5aF0QlRcFF6yI6ChHZodVqkZSUBK1WKzqKDZZbAcoqj2LaJMDZeALuBbnotegduBfkio5FRCQbFTkV2Dp9KypyKkRHISI7cnJyMH36dOTk5IiOYoPlVoCamlOYHg+4mM7CozAPvT5+Fx6FeaJjERHJRmVOJba9sQ2VOZWioxCRHTk5OXjjjTdYbomIiIiImhLLLRERERE5DJZbIiIiInIYLLcCqF0CsSILqHPyQ61Wh0P9HkGtVic6FhGRbLh6u6Jjcke4eruKjkJEdnh7eyM5ORne3t6io9hwFh3gZuSjjce9U4DUlZ1RFhKOH/6zUHQkIiJZ8Y70xkMrHhIdg4jqERkZiRUrVoiOYRfP3ApgqCtH6xgA5koo9bXQZZ2CUl8rOhYRkWwYa40oPlEMY61RdBQisqO2thYnTpxAba38+gvLrQDFpX/ixDTAvW4X/E4dw9MPdoffqWOiYxERyUbBkQLMi56HgiMFoqMQkR1HjhxBdHQ0jhw5IjqKDZZbIiIiInIYLLdERERE5DBYbomIiIjIYbDcEhEREZHD4FRgAgT53wdF39eRurI38mI74+29fMMEEdH/Co4LxjRpmugYRFSPuLg4SJIkOoZdPHNLRERERA6D5VaAwpLt+GsB4GrYDZ8zJzB0eD/4nDkhOhYRkWwUHivE4oTFKDxWKDoKEdlx7NgxJCQk4Ngx+U1lynIrgNFYhgQvQClVQFVThZD0v6GqqRIdi4hINuqq6nB+53nUVdWJjkJEdlRVVWHnzp2oqpJff2G5JSIiIiKHwXJLRERERA6D5ZaIiIiIHAbLrQCeHh0x5HugxrktylqE4fu3FqCsRZjoWEREsqGL0GHQ54Ogi9CJjkJEdkRERODzzz9HRESE6Cg2WG4FcNeEYuVqwKwMQa2XNw73H4xaL2/RsYiIZEPjo0GnIZ2g8dGIjkJEdvj4+GDIkCHw8fERHcUGy60AldWnMX40oDSehaakEHFfLoamhNPdEBFdUlVQhd3zd6OqQH7vxCYioKCgAPPnz0dBgfxuRMVyK0BlVQbm3wO4mk5Am5uNPrMmQ5ubLToWEZFslJ8rx8+pP6P8XLnoKERkx7lz55Camopz586JjmKD5ZaIiIiIHAbLLRERERE5DJZbIiIiInIYLLcCqJx98EsBYFR4weDugVM97oLB3UN0LCIi2VB7qtG6T2uoPdWioxCRHZ6enujTpw88PT1FR7HhLDrAzcjXuwfiJwCpK7uhJKw1vlqwRnQkIiJZ8Y32xZBfhoiOQUT1iI6Oxi+//CI6hl08cyuAyWyApxcgSQYoTCaoKyugMJlExyIikg2zyQx9uR5mk1l0FCKyw2Qyoby8HCYZ9heWWwEKin5F+QLAw/AHAo4fwqQ7WiHg+CHRsYiIZCPvQB7e9nobeQfyREchIjsOHDgALy8vHDhwQHQUGyy3REREROQwWG6JiIiIyGGw3BIRERGRw2C5JSIiIiKHwanABPDzuRv+w4DH5vREQVQ7/HdzBvQeXqJjERHJRkDHALyQ/wJcda6ioxCRHR07dkR+fj50Op3oKDZYbgVwVmpQmA8oFK4wq1So8fYTHYmISFaUKiXc/d1FxyCieqhUKvj7+4uOYRcvSxCgqGw3vpsNuBj2QXfuNB6eMAS6c6dFxyIiko3ik8X44v4vUHyyWHQUIrLj5MmTuP/++3Hy5EnRUWyw3ApQZyjE/UGAs1QMl8pyRP/+C1wqy0XHIiKSDX2ZHse/Pw59mV50FCKyo6ysDN9//z3KyspER7HBcktEREREDoPlloiIiIgcBsstERERETkMllsB3N1jMOlPQK9shYqAYGye9CYqAoJFxyIikg3PEE/0md0HniGeoqMQkR0hISGYPXs2QkJCREexwanABPB0i8L7C4HUXpGo9g3AniHjREciIpIVj0APJExKEB2DiOoRGBiISZMmiY5hF8/cClBdk41HHgKcTDlwKS9Fm03fwaW8VHQsIiLZqCmpweE1h1FTUiM6ChHZUVJSgjVr1qCkpER0FBsstwKUVx7AmocBjfEIdNlnMejlMdBlnxUdi4hINkpPl2Lto2tRerpUdBQisuP06dN49NFHcfq0/ObpZ7klIiIiIochtNzOnDkTt9xyCzw9PREQEIAHH3wQx44ds9qmtrYWKSkp8PX1hYeHBx5++GHk5eVZbZOVlYX+/fvDzc0NAQEBePHFF2E0GpvzpRARERGRDAgtt9u2bUNKSgp27tyJTZs2oa6uDn369EFVVZVlm4kTJ+L777/HmjVrsG3bNly4cAEPPfSQZb3JZEL//v1hMBjw119/Yfny5Vi2bBmmTp0q4iURERERkUBCZ0vYsGGD1cfLli1DQEAA0tLScMcdd6CsrAyLFy/GqlWrcM899wAAli5ditjYWOzcuRM9evTAxo0bceTIEfz6668IDAxEly5d8NZbb+Hll1/G9OnToVarRby0f+Ss9MTeCsDk4gajiwa5bTvC6KIRHYuISDacNc4I6hoEZw0n9SGSI41Gg65du0KjkV9/kdU1t5fuT+zj4wMASEtLQ11dHRITEy3btG3bFmFhYdixYwcAYMeOHejYsSMCAwMt2yQlJaG8vByHDx+2O45er0d5ebnVozn5+dyO+KeBWnUCilrFYNmq31DUKqZZMxARyZl/rD+e2vsU/GP9RUchIjtiY2Oxd+9exMbGio5iQzbl1mw2Y8KECejZsyc6dOgAAMjNzYVarYZOp7PaNjAwELm5uZZt/rfYXlp/aZ09M2fOhJeXl+URGhrayK+GiIiIiESQze97UlJScOjQIfz5559NPtaUKVOsJh4uLy9v1oKbV7gBtZ8D0/S/IfCoAkOH98Pny39GXttOzZaBiEjOcvblYHGPxRi9czSCu96cd3DMyspCYWFhs47p5+eHsLCwZh2Tbkz79u1Djx49sHPnTnTt2lV0HCuyKLepqan44Ycf8Pvvv6Nly5aW5UFBQTAYDCgtLbU6e5uXl4egoCDLNrt377Z6vkuzKVza5nIuLi5wcXFp5Fdx9STJDBcnAJAASYJznQGQJGF5iIhkRwJMBhNwk35rzMrKQtvYWNRUVzfruBo3NxzNyGDBpSuSJAkGgwGSDPuL0HIrSRKeeeYZfPPNN9i6dSsiIyOt1sfHx0OlUmHz5s14+OGHAQDHjh1DVlYWEhIu3pYxISEB//nPf5Cfn4+AgAAAwKZNm6DVatGuXbvmfUFERESNoLCwEDXV1Xj03wsREBndLGPmn87EV6+NQ2FhIcst3dCEltuUlBSsWrUK3333HTw9PS3XyHp5eUGj0cDLywujR4/GpEmT4OPjA61Wi2eeeQYJCQno0aMHAKBPnz5o164dhg4dinfeeQe5ubl47bXXkJKSIvTsLBER0fUKiIxGSGxn0TGIbihCy+3ChQsBAHfddZfV8qVLl2LEiBEAgPfffx9OTk54+OGHodfrkZSUhAULFli2VSqV+OGHHzBu3DgkJCTA3d0dw4cPx5tvvtlcL4OIiIiIZEL4ZQlX4urqivnz52P+/Pn1bhMeHo6ffvqpMaM1KR9dT7R/Eej90i0oiozBp2v+QGlIuOhYRESy4Rfrh3GHxsG7lbfoKERkR2xsLA4dOoRWrVqJjmJDFm8ou9moVV44kg7c46SF0VWDwtZtRUciIpIVlUaFgPYBomMQUT00Gg3at28vOoZdspnn9mZSWn4An7wFqOoOQXvhHPq9OQHaC+dExyIiko3Ss6VYP2Y9Ss+Wio5CRHacPXsWY8aMwdmzZ0VHscFyK0CtPhtjWgFqcx40ZcXo/O1KaMqKRcciIpKNmqIa7Fu8DzVFNaKjEJEdRUVFWLx4MYqKikRHscFyS0REREQOg+WWiIiIiBwGyy0REREROQyWWwE0rhGYeQDQK0NR5RuAHSOfRZUv3xVMRHSJe6A7ek7uCfdAd9FRiMiOwMBATJ48GYGBgaKj2OBUYAJ4ebbDK+8AqStjUBkQjG3PvC46EhGRrGhDtEicmSg6BhHVIyQkBDNnzhQdwy6euRWg1lCAO+8GFOZCqKsqEfb3dqirKkXHIiKSDX2FHme2noG+Qi86ChHZUVFRga1bt6KiokJ0FBsstwKUlu3B1jGAW90BeGedxBNjH4R31knRsYiIZKM4sxjL716O4kxOk0gkR5mZmbj77ruRmZkpOooNllsiIiIichgst0RERETkMFhuiYiIiMhhsNwK4OTkgvMGwKxQweysQnlAMMzOKtGxiIhkw0nlBM8QTzip+N8UkRypVCqEhIRApZJff+FUYAIE+PZGaB8gdeUdKIhuhwUbDoqOREQkK4EdAzHp/CTRMYioHh07dsT58+dFx7CLPxITERERkcNguRUgv2gzzi0FNIbf4Z95BOP7doJ/5hHRsYiIZCMvPQ9zWs5BXnqe6ChEZEd6ejpatmyJ9PR00VFssNwKYDbr0VINOEl1cDLWQZufAydjnehYRESyYa4zoyK7AuY6s+goRGRHXV0dsrOzUVcnv/7CcktEREREDoPlloiIiIgcBsstERERETkMllsBdF634K5PgWpVZ5SEtcaqj79FSVhr0bGIiGTDJ9oHw7cMh0+0j+goRGRHdHQ0tmzZgujoaNFRbLDcCuCq9se2LYDk5AeDuweyuvWEwd1DdCwiItlw8XRBxF0RcPF0ER2FiOzw9PTEXXfdBU9PT9FRbLDcClBWcQQzXgKcjcfhkZ+DO+e9BY/8HNGxiIhkozy7HL9O+RXl2eWioxCRHdnZ2ZgyZQqys7NFR7HBcitATe0ZTOkMuJjOwb0oHwlLP4B7Ub7oWEREslGVV4Xtb29HVV6V6ChEZEdeXh7efvtt5OXJby5qllsiIiIichgst0RERETkMFhuiYiIiMhhsNwK4OoSgk9PAQanQNR4+eDAg8mo8eJ0N0REl2h8Neg6uis0vhrRUYjIDl9fX4wePRq+vr6io9hwFh3gZqTTdsaTrwOpKzugvEUofp46V3QkIiJZ0YXrcP+n94uOQUT1CA8Px6effio6hl08cyuAoa4M7ToCCnM5nGtr4HfyKJxra0THIiKSjbqaOuQfzkddTZ3oKERkR01NDQ4fPoyaGvn1F5ZbAYpLt+PwZMCtbg98Tx/HmMG3w/f0cdGxiIhkozCjEAs7LERhRqHoKERkR0ZGBjp06ICMjAzRUWyw3BIRERGRw2C5JSIiIiKHwXJLRERERA6D5VYAhcIJejMAKACFAkaVGlAoRMciIpIPBaBUKwF+aySSJYVCAbVaDYUM+wunAhMg0K8vXJNeRerKe5DXthPe25UtOhIRkawEdw3Ga/rXRMcgonp07doVer1edAy7eOaWiIiIiBwGy60AhcV/IG0R4GrYAd9TxzHiiXvge4pTgRERXVKQUYCP4j5CQUaB6ChEZEdGRgbi4uI4FRhdZDRVIM4TUErVcNbXIOhoOpz18psEmYhIFGONEbn7cmGsMYqOQkR21NTUYN++fbyJAxERERFRU2K5JSIiIiKHwXJLRERERA6D5VYArUdnDP4aqHFuh9KQcHwz61OUhoSLjkVEJBu6SB0e+eoR6CJ1oqMQkR2RkZH46quvEBkZKTqKDZZbAdw0IVi7DjArg6HX6nDs3geg1+pExyIikg2NtwbtB7eHxlsjOgoR2eHt7Y3BgwfD29tbdBQbLLcCVFSfwMRxgLPxNNyK8nHLioVwK8oXHYuISDYq8yqxY84OVOZVio5CRHbk5eVhzpw5yMvLEx3FBsutAFVVxzGnF+BiOgXP/Bz0njMVnvk5omMREclGRXYFNj6/ERXZFaKjEJEd2dnZeP7555GdLb+7rLLcEhEREZHDYLklIiIiIofBcktEREREDoPlVgCV2g/rcwGjwgd6Dy0y70iC3kMrOhYRkWy4eLkgZmAMXLxcREchIju8vLwwcOBAeHl5iY5iw1l0gJuRr9etiH8eSF3ZFaWhkfh67grRkYiIZMWntQ/+tf5fomMQUT1at26N9evXi45hF8/cCmA01cAvAJCkWjjV1UFTUginujrRsYiIZMNUZ0JVQRVMdSbRUYjIjrq6OhQUFKBOhv2F5VaAwuItKHgf8DBsh/+JI3iudyz8TxwRHYuISDby0/PxXsB7yE/nHOBEcpSeno6AgACkp6eLjmKDlyUQERFdhaysLBQWFjbLWBkZGc0yDpEjYrklIiK6gqysLLSNjUVNdbXoKER0BSy3REREV1BYWIia6mo8+u+FCIiMbvLxjm3fjE0LZjb5OESOiOWWiIjoKgVERiMktnOTj5N/OrPJxyByVHxDmQD+vonQjgcq1bcjP6YD5vx+CvkxHUTHIiKSjcDOgZhcNhmBnQNFRyEiOzp37oyysjJ07tz0P+xdK565FUDppEZFGaBQqCEplTB4eIqOREQkK05KJ7hoeQMHIrlSKpXQauV5AyqeuRWgqGQnNswFXAx/wzvrJB4dPxjeWSdFxyIiko2izCKsSFqBoswi0VGIyI7MzEwkJSUhM1N+l9Cw3ApQZyxGkj/gLJVBXVWJVju3Ql1VKToWEZFsGCoMOLnxJAwVBtFRiMiOiooKbNy4ERUVFaKj2GC5JSIiIiKHwXJLRERERA6D5ZaIiIiIHAbLrQAe7rFI+Q2oVUahPCgEG19+G+VBIaJjERHJhjZUi34f9oM2VJ7vxia62YWGhuLDDz9EaGio6Cg2WG4F8HCLxILFgMk5HDXeftj72GjUePuJjkVEJBvu/u64NeVWuPu7i45CRHb4+/sjJSUF/v7+oqPY4Dy3AlTVnEPy44CTKRuuZWFo/eevONkrEbVe3qKjERHJQk1xDTJ/ykT0fdHQ+GhExyG6IWRlZaGwsLBZxiorK8OBAwcwbNgw+Pj4NMuYV4vlVoCKynSsGAhM1h+F14UQDHx9PJau/JXllojo/ys9U4pvhn6DsWljWW6JrkJWVhbaxsaiprq6WceNjo5G//79m3XMK2G5JSIiIrrBFRYWoqa6Go/+eyECIqObfLxj2zdj04KZKC0tbfKxrhXLLREREZGDCIiMRkhs5yYfJ/+0/O5MdgnfUEZEREREDoPlVgBnZy/sKANMCk/UadyR3bEb6jR8RzAR0SUqdxVa9mgJlbtKdBQiskPl4goA0Gjkd008L0sQwM+7J+LHA6krb0VxRBQ+X/6z6EhERLLi18YPo3eMFh2D6Lo05+wFGRkZzTLOJd4tLs5vGxER0azjXg2WWyKyqzm/KQOAn58fwsLCmm08IqKmJGr2AmK5FSK34CdIK4HJ+s0IzABGJidi6cpfkdcMF4ATXQ0R35Q1bm44mpHBgksAgJy9Ofg4/mOMTRuL4Lhg0XGIrpmo2Quay6U3lGVkZCAuLq7Zxr0aLLfUJHjW78bW3N+U809n4qvXxqGwsJBfx0bEf4fUEM356+2b4Zjh7AXNj+WWGp2Is34urq74eu1aBAc33xkevV4PFxeXZhtPxH8CzfVNmRofz77TtaoozIPCyQlDhgxptjF5zFBTYLmlRtfcZ/1O79uFn+a8jgEDBjT5WP9L4eQEyWxutvH4n8CNrbnPomZkZPDsO12TmopySGYzjxm64bHc3kSa61dNl8Zpzl/FNOc3ZOD/rm3ifwI3ruYsmzk5OXhk8GDU1tQ0y3j/q7nPvjfW95nSjFLL8+Ugx+42N8OvtEXgb2zoRsdyK4CPrheingX6vd4dha3aYNG3u1AR2KLJxhPxqyYRmvMb8qVrm/ifQONqrh/ARJVNET+ANZfG/j7jDGdoocW/h/wbRhjtbtOclyM19zRLRHLnExIOAGjVqpXgJLZYbgVQq7Q4eRyAkwdMLq4oDWvaA6O5f9XU3P+p3kya++x7cxH1A1hz/5sQ8QNYc2nu7zOiLkcioouc1WoAaNb3nlwtllsBisvT8PlMYF/dAXhl63D7grfxx/jJKPv/PwU1Fb5j88bl6GffRf0Axn8Tja+xPqdSdjmMC3bBeXx3KEK0Nuub+3Ik/tBOZK08/+LlQtnZ2ZwKrKnMnz8f7777LnJzc9G5c2fMmzcPt956q+hYdhn0eRgSBhzSF8K1vBQdfl6LPUOebvJySzeum+XsO8smXSKV62H++TikIV2gCKl/Ox4zNz5OPXZjqq2qBACUl5cLTmLLIcrtl19+iUmTJmHRokXo3r075s6di6SkJBw7dgwBAQGi4xE1Gv5HTkSOQsRvpHid9s3BIcrtnDlz8OSTT2LkyJEAgEWLFuHHH3/EkiVLMHnyZMHpiIiI6HK8Tpuayg1fbg0GA9LS0jBlyhTLMicnJyQmJmLHjh1299Hr9dDr9ZaPy8rKADTfqfXqqlqUVwOFF7KgqTuIcgBZGQdxobqqScYrOHPxLFx2xkEYmmiMm2k8EWNyPI4n9zEbezyns1XQoRa5GfthrnZv8vGuxNHHEzHmpfHqamuaZbzq0iJIZjNuH5YCXdA/XOvSSM4f3o99P37lsMfMpfGqq6ubrT9dGkeSpH/eULrBZWdnSwCkv/76y2r5iy++KN16661295k2bZoEgA8++OCDDz744IOPG+xx7ty5f+yGN/yZ24aYMmUKJk2aZPnYbDajuLgYvr6+UCgU1/Xc5eXlCA0Nxblz56DV2r7Dl4jHCF0NHid0NXic0JU40jEiSRIqKirQosU/3xvghi+3fn5+UCqVyMvLs1qel5eHoKAgu/u4uLjYzMum0+kaNZdWq73hDyJqWjxG6GrwOKGrweOErsRRjhEvL68rbuPUDDmalFqtRnx8PDZv3mxZZjabsXnzZiQkJAhMRkRERETN7YY/cwsAkyZNwvDhw9GtWzfceuutmDt3LqqqqiyzJxARERHRzcEhyu1jjz2GgoICTJ06Fbm5uejSpQs2bNiAwMDAZs/i4uKCadOmyfJ2dCQPPEboavA4oavB44Su5GY8RhSSdKX5FIiIiIiIbgw3/DW3RERERESXsNwSERERkcNguSUiIiIih8FyS0REREQOg+W2Ec2fPx8RERFwdXVF9+7dsXv3btGRqIn8/vvvGDhwIFq0aAGFQoFvv/3War0kSZg6dSqCg4Oh0WiQmJiIzMxMq22Ki4uRnJwMrVYLnU6H0aNHo7Ky0mqbgwcP4vbbb4erqytCQ0PxzjvvNPVLo0Y0c+ZM3HLLLfD09ERAQAAefPBBHDt2zGqb2tpapKSkwNfXFx4eHnj44YdtbkqTlZWF/v37w83NDQEBAXjxxRdhNBqtttm6dSvi4uLg4uKCqKgoLFu2rKlfHjWChQsXolOnTpYJ9hMSEvDzzz9b1vP4oMu9/fbbUCgUmDBhgmUZj5PL/OPNeemqrV69WlKr1dKSJUukw4cPS08++aSk0+mkvLw80dGoCfz000/Sq6++Kq1bt04CIH3zzTdW699++23Jy8tL+vbbb6UDBw5I999/vxQZGSnV1NRYtunbt6/UuXNnaefOndIff/whRUVFSf/6178s68vKyqTAwEApOTlZOnTokPTFF19IGo1G+uijj5rrZdJ1SkpKkpYuXSodOnRI2r9/v3TfffdJYWFhUmVlpWWbp59+WgoNDZU2b94s/f3331KPHj2k2267zbLeaDRKHTp0kBITE6V9+/ZJP/30k+Tn5ydNmTLFss2pU6ckNzc3adKkSdKRI0ekefPmSUqlUtqwYUOzvl66duvXr5d+/PFH6fjx49KxY8ekV155RVKpVNKhQ4ckSeLxQdZ2794tRURESJ06dZKee+45y3IeJ9ZYbhvJrbfeKqWkpFg+NplMUosWLaSZM2cKTEXN4fJyazabpaCgIOndd9+1LCstLZVcXFykL774QpIkSTpy5IgEQNqzZ49lm59//llSKBRSdna2JEmStGDBAsnb21vS6/WWbV5++WWpTZs2TfyKqKnk5+dLAKRt27ZJknTxuFCpVNKaNWss22RkZEgApB07dkiSdPEHKScnJyk3N9eyzcKFCyWtVms5Nl566SWpffv2VmM99thjUlJSUlO/JGoC3t7e0qeffsrjg6xUVFRI0dHR0qZNm6Q777zTUm55nNjiZQmNwGAwIC0tDYmJiZZlTk5OSExMxI4dOwQmIxFOnz6N3Nxcq+PBy8sL3bt3txwPO3bsgE6nQ7du3SzbJCYmwsnJCbt27bJsc8cdd0CtVlu2SUpKwrFjx1BSUtJMr4YaU1lZGQDAx8cHAJCWloa6ujqrY6Vt27YICwuzOlY6duxodVOapKQklJeX4/Dhw5Zt/vc5Lm3D7z83FpPJhNWrV6OqqgoJCQk8PshKSkoK+vfvb/O15HFiyyHuUCZaYWEhTCaTzR3RAgMDcfToUUGpSJTc3FwAsHs8XFqXm5uLgIAAq/XOzs7w8fGx2iYyMtLmOS6t8/b2bpL81DTMZjMmTJiAnj17okOHDgAufh3VajV0Op3VtpcfK/aOpUvr/mmb8vJy1NTUQKPRNMVLokaSnp6OhIQE1NbWwsPDA9988w3atWuH/fv38/ggAMDq1auxd+9e7Nmzx2Ydv4/YYrklImoGKSkpOHToEP7880/RUUhm2rRpg/3796OsrAxr167F8OHDsW3bNtGxSCbOnTuH5557Dps2bYKrq6voODcEXpbQCPz8/KBUKm3emZiXl4egoCBBqUiUS1/zfzoegoKCkJ+fb7XeaDSiuLjYaht7z/G/Y9CNITU1FT/88AO2bNmCli1bWpYHBQXBYDCgtLTUavvLj5UrHQf1baPVam+osy03K7VajaioKMTHx2PmzJno3Lkz/vvf//L4IAAXLzvIz89HXFwcnJ2d4ezsjG3btuGDDz6As7MzAgMDeZxchuW2EajVasTHx2Pz5s2WZWazGZs3b0ZCQoLAZCRCZGQkgoKCrI6H8vJy7Nq1y3I8JCQkoLS0FGlpaZZtfvvtN5jNZnTv3t2yze+//466ujrLNps2bUKbNm14ScINQpIkpKam4ptvvsFvv/1mc5lJfHw8VCqV1bFy7NgxZGVlWR0r6enpVj8Mbdq0CVqtFu3atbNs87/PcWkbfv+5MZnNZuj1eh4fBADo3bs30tPTsX//fsujW7duSE5Otvydx8llRL+jzVGsXr1acnFxkZYtWyYdOXJEGjt2rKTT6azemUiOo6KiQtq3b5+0b98+CYA0Z84cad++fdLZs2clSbo4FZhOp5O+++476eDBg9IDDzxgdyqwrl27Srt27ZL+/PNPKTo62moqsNLSUikwMFAaOnSodOjQIWn16tWSm5sbpwK7gYwbN07y8vKStm7dKuXk5Fge1dXVlm2efvppKSwsTPrtt9+kv//+W0pISJASEhIs6y9N4dOnTx9p//790oYNGyR/f3+7U/i8+OKLUkZGhjR//vwbdgqfm83kyZOlbdu2SadPn5YOHjwoTZ48WVIoFNLGjRslSeLxQfb972wJksTj5HIst41o3rx5UlhYmKRWq6Vbb71V2rlzp+hI1ES2bNkiAbB5DB8+XJKki9OBvf7661JgYKDk4uIi9e7dWzp27JjVcxQVFUn/+te/JA8PD0mr1UojR46UKioqrLY5cOCA1KtXL8nFxUUKCQmR3n777eZ6idQI7B0jAKSlS5datqmpqZHGjx8veXt7S25ubtKgQYOknJwcq+c5c+aM1K9fP0mj0Uh+fn7S888/L9XV1Vlts2XLFqlLly6SWq2WWrVqZTUGydeoUaOk8PBwSa1WS/7+/lLv3r0txVaSeHyQfZeXWx4n1hSSJElizhkTERERETUuXnNLRERERA6D5ZaIiIiIHAbLLRERERE5DJZbIiIiInIYLLdERERE5DBYbomIiIjIYbDcEhEREZHDYLklIiIiIofBcktERHYNHToUM2bMaLLnP3LkCFq2bImqqqomG4OIbj4st0REV2HEiBF48MEHG7z/smXLoNPpGi1PUztw4AB++uknPPvss002Rrt27dCjRw/MmTOnycYgopsPyy0REdmYN28eBg8eDA8PjyYdZ+TIkVi4cCGMRmOTjkNENw+WWyKiRjBnzhx07NgR7u7uCA0Nxfjx41FZWQkA2Lp1K0aOHImysjIoFAooFApMnz4dAKDX6/HCCy8gJCQE7u7u6N69O7Zu3Wp53ktnfH/55RfExsbCw8MDffv2RU5OjtX4S5YsQfv27eHi4oLg4GCkpqYCAEaNGoUBAwZYbVtXV4eAgAAsXrzY7msxmUxYu3YtBg4caLU8IiIC//73vzFs2DB4eHggPDwc69evR0FBAR544AF4eHigU6dO+Pvvvy37nD17FgMHDoS3tzfc3d3Rvn17/PTTT5b19957L4qLi7Ft27Zr+4QTEdWD5ZaIqBE4OTnhgw8+wOHDh7F8+XL89ttveOmllwAAt912G+bOnQutVoucnBzk5OTghRdeAACkpqZix44dWL16NQ4ePIjBgwejb9++yMzMtDx3dXU13nvvPXz++ef4/fffkZWVZdkfABYuXIiUlBSMHTsW6enpWL9+PaKiogAAY8aMwYYNG6zK8A8//IDq6mo89thjdl/LwYMHUVZWhm7dutmse//999GzZ0/s27cP/fv3x9ChQzFs2DAMGTIEe/fuRevWrTFs2DBIkgQASElJgV6vx++//4709HTMmjXL6mywWq1Gly5d8McffzT0U09EZE0iIqIrGj58uPTAAw9c9fZr1qyRfH19LR8vXbpU8vLystrm7NmzklKplLKzs62W9+7dW5oyZYplPwDSiRMnLOvnz58vBQYGWj5u0aKF9Oqrr9abpV27dtKsWbMsHw8cOFAaMWJEvdt/8803klKplMxms9Xy8PBwaciQIZaPc3JyJADS66+/blm2Y8cOCYCUk5MjSZIkdezYUZo+fXq9Y0mSJA0aNOgf8xARXQueuSUiagS//vorevfujZCQEHh6emLo0KEoKipCdXV1vfukp6fDZDIhJiYGHh4else2bdtw8uRJy3Zubm5o3bq15ePg4GDk5+cDAPLz83HhwgX07t273nHGjBmDpUuXAgDy8vLw888/Y9SoUfVuX1NTAxcXFygUCpt1nTp1svw9MDAQANCxY0ebZZfyPfvss/j3v/+Nnj17Ytq0aTh48KDNc2o0mn/8PBERXQuWWyKi63TmzBkMGDAAnTp1wtdff420tDTMnz8fAGAwGOrdr7KyEkqlEmlpadi/f7/lkZGRgf/+97+W7VQqldV+CoXC8mt/jUZzxXzDhg3DqVOnsGPHDqxYsQKRkZG4/fbb693ez88P1dXVdrP/b5ZL5dfeMrPZDOBisT516hSGDh2K9PR0dOvWDfPmzbN6zuLiYvj7+1/xdRARXQ2WWyKi65SWlgaz2YzZs2ejR48eiImJwYULF6y2UavVMJlMVsu6du0Kk8mE/Px8REVFWT2CgoKuamxPT09ERERg8+bN9W7j6+uLBx98EEuXLsWyZcswcuTIf3zOLl26ALg4D21jCA0NxdNPP41169bh+eefxyeffGK1/tChQ+jatWujjEVE5Cw6ABHRjaKsrAz79++3Wubr64uoqCjU1dVh3rx5GDhwILZv345FixZZbRcREYHKykps3rwZnTt3hpubG2JiYpCcnIxhw4Zh9uzZ6Nq1KwoKCrB582Z06tQJ/fv3v6pc06dPx9NPP42AgAD069cPFRUV2L59O5555hnLNmPGjMGAAQNgMpkwfPjwf3w+f39/xMXF4c8//7QU3YaaMGEC+vXrh5iYGJSUlGDLli2IjY21rD9z5gyys7ORmJh4XeMQEV3CM7dERFdp69at6Nq1q9XjjTfeQOfOnTFnzhzMmjULHTp0wMqVKzFz5kyrfW+77TY8/fTTeOyxx+Dv74933nkHALB06VIMGzYMzz//PNq0aYMHH3wQe/bsQVhY2FXnGj58OObOnYsFCxagffv2GDBggNVsCwCQmJiI4OBgJCUloUWLFld8zjFjxmDlypVXnaE+JpMJKSkpiI2NRd++fRETE4MFCxZY1n/xxRfo06cPwsPDr3ssIiIAUEiXLtwiIiKHVVlZiZCQECxduhQPPfTQFbevqalBmzZt8OWXXyIhIaFJMhkMBkRHR2PVqlXo2bNnk4xBRDcfXpZAROTAzGYzCgsLMXv2bOh0Otx///1XtZ9Go8Fnn32GwsLCJsuWlZWFV155hcWWiBoVz9wSETmwM2fOIDIyEi1btsSyZcv+ccowIiJHwHJLRERERA6DbygjIiIiIofBcktEREREDoPlloiIiIgcBsstERERETkMllsiIiIichgst0RERETkMFhuiYiIiMhhsNwSERERkcP4f/e74U7n5mmyAAAAAElFTkSuQmCC", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "#BRAND NEW TESTING SCRIPT FOR ALL METHODS BASED OFF PREVIOUS TWO\n", "#ENHANCED VERSION OF MY PREVIOUS 2 TESTING SCRIPTS WITH EXTRAS\n", "#Testing script for Granite3.2-2B-Instruct using BF16 base + FP16 Adapters\n", "\n", "import os\n", "import torch\n", "import time\n", "import json\n", "import numpy as np\n", "import pandas as pd\n", "import evaluate\n", "import nltk\n", "import gc\n", "import math\n", "import re\n", "import matplotlib.pyplot as plt\n", "import mauve\n", "from transformers import AutoModelForCausalLM, AutoTokenizer, TrainingArguments, Trainer, EarlyStoppingCallback, TrainerCallback, BitsAndBytesConfig\n", "from transformers.trainer_utils import get_last_checkpoint\n", "from peft import LoraConfig, get_peft_model, prepare_model_for_kbit_training, PeftModel\n", "from datasets import Dataset\n", "from sklearn.metrics import accuracy_score, precision_recall_fscore_support\n", "from nltk.translate.bleu_score import sentence_bleu\n", "from rouge_score import rouge_scorer\n", "from torch.utils.data import DataLoader\n", "from sentence_transformers import SentenceTransformer, util\n", "\n", "nltk.download(\"punkt\")\n", "os.environ[\"TOKENIZERS_PARALLELISM\"] = \"false\"\n", "os.environ[\"PYTORCH_CUDA_ALLOC_CONF\"] = \"expandable_segments:True\"\n", "\n", "MODEL_NAME = \"ibm-granite/granite-3.2-2b-instruct\"\n", "ADAPTER_PATH = \"Granite3.2-2B-lora_adapters-FP16\"\n", "TEST_CSV_PATH = \"Testing Dataset RE.csv\"\n", "OUTPUT_JSON_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/BF16/FP16/Granite3.2-2B-BF16-lora-FP16-Evaluation_Results.json\"\n", "OUTPUT_INFER_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/BF16/FP16/Granite3.2-2B-BF16-lora-FP16-Inference_Curve.png\"\n", "OUTPUT_MEMORY_USAGE_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/BF16/FP16/Granite3.2-2B-BF16-lora-FP16-Memory_Usage_Curve.png\"\n", "OUTPUT_LATENCY_HIST_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/BF16/FP16/Granite3.2-2B-BF16-lora-FP16-Latency_Histogram.png\"\n", "OUTPUT_MEMORY_HIST_PATH = \"Complete_Evaluation_Results/Granite3.2-2B/BF16/FP16/Granite3.2-2B-BF16-lora-FP16-Memory_Histogram.png\"\n", "SEMANTIC_MODEL = \"all-MiniLM-L6-v2\"\n", "\n", "tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)\n", "tokenizer.pad_token = tokenizer.eos_token\n", "tokenizer.padding_side = \"right\"\n", "\n", "test_df = pd.read_csv(TEST_CSV_PATH)\n", "\n", "def preprocess_function(examples):\n", " inputs = []\n", " labels = []\n", " \n", " for context, question, answer in zip(\n", " examples.get(\"Context\", [\"\"] * len(examples[\"Question\"])),\n", " examples[\"Question\"], \n", " examples[\"Answer\"]):\n", " \n", " context = context.strip() if context else \"\"\n", " question = question.strip()\n", " answer = answer.strip()\n", "\n", " if context:\n", " prompt = f\"Context: {context}\\nQuestion: {question}\\nAnswer:\"\n", " else:\n", " prompt = f\"Question: {question}\\nAnswer:\"\n", "\n", " full_text = prompt + \" \" + answer\n", " \n", " tokenized = tokenizer(full_text, padding=\"max_length\", truncation=True, max_length=512)\n", " prompt_ids = tokenizer(prompt, truncation=True, max_length=512, add_special_tokens=False)[\"input_ids\"]\n", "\n", " input_ids = tokenized[\"input_ids\"]\n", " attention_mask = tokenized[\"attention_mask\"]\n", " label_ids = input_ids.copy()\n", " label_ids[:len(prompt_ids)] = [-100] * len(prompt_ids)\n", " \n", " if all(id_ == -100 for id_ in label_ids):\n", " continue\n", "\n", " inputs.append({\"input_ids\": input_ids, \"attention_mask\": attention_mask, \"labels\": label_ids})\n", "\n", " return {\"input_ids\": [x[\"input_ids\"] for x in inputs], \"attention_mask\": [x[\"attention_mask\"] for x in inputs],\n", " \"labels\": [x[\"labels\"] for x in inputs]}\n", "\n", "test_dataset = Dataset.from_pandas(test_df).map(preprocess_function, batched=True, batch_size=32,\n", " remove_columns=test_df.columns.tolist())\n", "\n", "#Removed BitsAndBytesConfig because the base is loaded in BF16 for LoRA (not QLoRA)\n", "\n", "model = AutoModelForCausalLM.from_pretrained(MODEL_NAME, torch_dtype=torch.bfloat16, device_map=\"auto\",trust_remote_code=True)\n", "model = PeftModel.from_pretrained(model, ADAPTER_PATH).eval()\n", "model.config.pad_token_id = tokenizer.pad_token_id\n", "\n", "# Load semantic similarity model\n", "semantic_model = SentenceTransformer(SEMANTIC_MODEL)\n", "\n", "def compute_loss_and_perplexity():\n", " losses = []\n", " for sample in test_dataset:\n", " with torch.no_grad():\n", " input_ids = torch.tensor(sample[\"input_ids\"]).unsqueeze(0).to(\"cuda\")\n", " attention_mask = torch.tensor(sample[\"attention_mask\"]).unsqueeze(0).to(\"cuda\")\n", " labels = torch.tensor(sample[\"labels\"]).unsqueeze(0).to(\"cuda\")\n", " \n", " loss = model(input_ids=input_ids, attention_mask=attention_mask, labels=labels).loss.item()\n", " losses.append(loss)\n", " \n", " avg_loss = sum(losses) / len(losses)\n", " return avg_loss, math.exp(avg_loss)\n", "\n", "def extract_answer(text):\n", " return text.split(\"Answer:\")[-1].strip() if \"Answer:\" in text else text.strip()\n", "\n", "def normalize(text):\n", " return re.sub(r\"[^\\w\\s]\", \"\", text.strip().lower())\n", "\n", "def compute_metrics(preds, refs):\n", " decoded_preds = tokenizer.batch_decode(preds, skip_special_tokens=True)\n", " #decoded_refs = tokenizer.batch_decode(refs, skip_special_tokens=True)\n", "\n", " # Replace -100s in refs before decoding\n", " safe_refs = [[token if token != -100 else tokenizer.pad_token_id for token in ref] for ref in refs]\n", " decoded_refs = tokenizer.batch_decode(safe_refs, skip_special_tokens=True)\n", "\n", " preds_clean = [normalize(extract_answer(p)) for p in decoded_preds]\n", " refs_clean = [normalize(extract_answer(r)) for r in decoded_refs]\n", "\n", " sim_scores = util.cos_sim(semantic_model.encode(preds_clean, convert_to_tensor=True),\n", " semantic_model.encode(refs_clean, convert_to_tensor=True)).diagonal()\n", " semantic_threshold = 0.8\n", " matches = [1 if sim >= semantic_threshold else 0 for sim in sim_scores]\n", "\n", " accuracy = sum(matches) / len(matches)\n", " precision, recall, f1, _ = precision_recall_fscore_support(matches, matches, average=\"binary\", zero_division=0)\n", " avg_bleu = sum([sentence_bleu([r.split()], p.split()) for r, p in zip(refs_clean, preds_clean)]) / len(preds_clean)\n", "\n", " rouge = rouge_scorer.RougeScorer([\"rouge1\", \"rouge2\", \"rougeL\"], use_stemmer=True)\n", " rouge_scores = [rouge.score(ref, pred) for ref, pred in zip(refs_clean, preds_clean)]\n", " avg_rouge = {k: sum([s[k].fmeasure for s in rouge_scores]) / len(rouge_scores) for k in rouge_scores[0]}\n", "\n", " return {\"accuracy:\": accuracy, \"precision:\": precision, \"recall:\": recall, \"f1:\": f1,\n", " \"bleu:\": avg_bleu, \"rouge:\": avg_rouge, \"semantic_similarity_avg:\": sim_scores.mean().item()}, decoded_preds, decoded_refs\n", "\n", "def measure_inference_and_generate():\n", " preds, latencies, memory_used_per_sample, peak_memories = [], [], [], []\n", "\n", " #Measure model load memory (after full load + preparation)\n", " gc.collect()\n", " torch.cuda.empty_cache()\n", " torch.cuda.reset_peak_memory_stats()\n", " model_load_memory = torch.cuda.memory_allocated() / (1024 ** 3)\n", "\n", " for idx, sample in enumerate(test_dataset):\n", " gc.collect()\n", " torch.cuda.empty_cache()\n", " torch.cuda.reset_peak_memory_stats()\n", "\n", " input_ids = torch.tensor(sample[\"input_ids\"]).unsqueeze(0).to(\"cuda\")\n", " attention_mask = torch.tensor(sample[\"attention_mask\"]).unsqueeze(0).to(\"cuda\")\n", " \n", " # Measure base memory BEFORE\n", " base_memory = torch.cuda.memory_allocated()\n", "\n", " # Wait for everything to settle\n", " torch.cuda.synchronize()\n", " #mem_before = torch.cuda.memory_allocated() / (1024 ** 3)\n", " start_time = time.time()\n", "\n", " with torch.no_grad():\n", " output = model.generate(input_ids=input_ids, attention_mask=attention_mask, max_new_tokens=50,\n", " do_sample=True, top_p=0.9, top_k=50,\n", " temperature=0.7, repetition_penalty=1.1, length_penalty=0.8)\n", "\n", " torch.cuda.synchronize()\n", " end_time = time.time()\n", " #mem_after = torch.cuda.memory_allocated() / (1024 ** 3)\n", " peak_memory = torch.cuda.max_memory_allocated() \n", "\n", " inference_memory = (peak_memory - base_memory) / (1024 ** 3) # in GB\n", "\n", " preds.append(output[0].tolist())\n", " latencies.append((end_time - start_time) * 1000) # ms\n", " memory_used_per_sample.append(inference_memory) # Memory used by this inference\n", " peak_memories.append(peak_memory / (1024 ** 3)) # Peak memory usage during this sample\n", "\n", " # Calculate averages now\n", " avg_inference_memory = np.mean(memory_used_per_sample)\n", "\n", " return preds, latencies, memory_used_per_sample, peak_memories, model_load_memory, avg_inference_memory\n", "\n", "def compute_mauve(pred_texts, ref_texts):\n", " return mauve.compute_mauve(p_text=pred_texts, q_text=ref_texts,\n", " device_id=0, max_text_length=256).mauve\n", "\n", "print(\"Generating predictions...\")\n", "loss, perplexity = compute_loss_and_perplexity()\n", "generated_preds, latencies, memory_used_per_sample, peak_memories, model_load_memory, avg_inference_memory = measure_inference_and_generate()\n", "ref_labels = [sample[\"labels\"] for sample in test_dataset]\n", "metrics, decoded_preds, decoded_refs = compute_metrics(generated_preds, ref_labels)\n", "mauve_score = compute_mauve(decoded_preds, decoded_refs)\n", "\n", "# 1) Plot Inference_Performance curves for latency and memory usage\n", "plt.plot(latencies, label=\"Latency (ms)\")\n", "plt.plot(memory_used_per_sample, label=\"Memory Usage (GB)\")\n", "plt.title(\"Inference_Performance Curve\")\n", "plt.xlabel(\"Sample Index\")\n", "plt.legend()\n", "plt.savefig(OUTPUT_INFER_PATH)\n", "\n", "# 2a) Compute latency stats and then plot the latency histogram\n", "latencies_np = np.array(latencies)\n", "latency_stats = {\n", " \"min_latency_ms\": float(np.min(latencies_np)),\n", " \"max_latency_ms\": float(np.max(latencies_np)),\n", " \"lower_quartile_ms\": float(np.percentile(latencies_np, 25)),\n", " \"median_latency_ms\": float(np.median(latencies_np)),\n", " \"upper_quartile_ms\": float(np.percentile(latencies_np, 75)),\n", " \"avg_latency_ms\": float(np.mean(latencies_np))\n", "}\n", "\n", "# 2b) Plot the Histogram for Latency (ms)\n", "plt.figure(figsize=(8, 6))\n", "plt.hist(latencies, bins=20, color='skyblue', edgecolor='black')\n", "plt.axvline(latency_stats[\"min_latency_ms\"], color='green', linestyle='dashed', linewidth=1, label='Min')\n", "plt.axvline(latency_stats[\"lower_quartile_ms\"], color='orange', linestyle='dashed', linewidth=1, label='Q1')\n", "plt.axvline(latency_stats[\"median_latency_ms\"], color='red', linestyle='dashed', linewidth=1, label='Median')\n", "plt.axvline(latency_stats[\"upper_quartile_ms\"], color='purple', linestyle='dashed', linewidth=1, label='Q3')\n", "plt.axvline(latency_stats[\"max_latency_ms\"], color='black', linestyle='dashed', linewidth=1, label='Max')\n", "plt.title(\"Latency Histogram\")\n", "plt.xlabel(\"Latency (ms)\")\n", "plt.ylabel(\"Frequency\")\n", "plt.legend()\n", "plt.savefig(OUTPUT_LATENCY_HIST_PATH)\n", "\n", "# Line plot focusing on 0.1MB to 1MB\n", "plt.figure(figsize=(8, 6))\n", "plt.plot(memory_used_per_sample, label=\"Memory Usage (GB)\", color=\"teal\")\n", "plt.ylim(0.1, 0.5) # Zoom in to 0.1GB–0.5GB range\n", "plt.title(\"Memory Usage per Sample (Zoomed 100MB–500MB)\")\n", "plt.xlabel(\"Sample Index\")\n", "plt.ylabel(\"Memory (GB)\")\n", "plt.legend()\n", "plt.savefig(OUTPUT_MEMORY_USAGE_PATH)\n", "\n", "# 4) Compute memory stats and Plot the Histogram for memory usage\n", "memory_used_per_sample_np = np.array(memory_used_per_sample)\n", "memory_stats = {\n", " \"min_memory_gb\": float(np.min(memory_used_per_sample_np)),\n", " \"max_memory_gb\": float(np.max(memory_used_per_sample_np)),\n", " \"lower_quartile_gb\": float(np.percentile(memory_used_per_sample_np, 25)),\n", " \"median_memory_gb\": float(np.median(memory_used_per_sample_np)),\n", " \"upper_quartile_gb\": float(np.percentile(memory_used_per_sample_np, 75)),\n", " \"avg_memory_gb\": float(np.mean(memory_used_per_sample_np)),\n", " \"model_load_memory_gb\": model_load_memory,\n", " \"avg_inference_memory_gb\": avg_inference_memory\n", "}\n", "\n", "# Plot the Histogram for memory usage\n", "plt.figure(figsize=(8, 6))\n", "plt.hist(memory_used_per_sample, bins=20, color='lightcoral', edgecolor='black')\n", "plt.axvline(memory_stats[\"min_memory_gb\"], color='green', linestyle='dashed', linewidth=1, label='Min')\n", "plt.axvline(memory_stats[\"lower_quartile_gb\"], color='orange', linestyle='dashed', linewidth=1, label='Q1')\n", "plt.axvline(memory_stats[\"median_memory_gb\"], color='red', linestyle='dashed', linewidth=1, label='Median')\n", "plt.axvline(memory_stats[\"upper_quartile_gb\"], color='purple', linestyle='dashed', linewidth=1, label='Q3')\n", "plt.axvline(memory_stats[\"max_memory_gb\"], color='black', linestyle='dashed', linewidth=1, label='Max')\n", "plt.title(\"Memory Usage Histogram\")\n", "plt.xlabel(\"Memory Usage (GB)\")\n", "plt.ylabel(\"Frequency\")\n", "plt.legend()\n", "plt.savefig(OUTPUT_MEMORY_HIST_PATH)\n", "\n", "# Save all results\n", "results = {\"eval_loss:\": loss, \"perplexity:\": perplexity, \"performance_metrics:\": metrics, \"mauve:\": mauve_score,\n", " \"inference_performance:\": {**latency_stats, **memory_stats}}\n", "\n", "with open(OUTPUT_JSON_PATH, \"w\") as f:\n", " json.dump(results, f, indent=4)\n", "\n", "print(f\"Evaluation Complete. Results saved to {OUTPUT_JSON_PATH}\")\n" ] }, { "cell_type": "code", "execution_count": 10, "id": "aa1f317a-b13d-451f-b169-2367525305ec", "metadata": {}, "outputs": [], "source": [ "#STARTED ABOVE TESTING AT 6:25PM ON 29/04/25\n", "#FEATURISING STARTED AT 7:00PM AND ENDED AT 7:02PM (35 MIN AFTER STARTING)\n", "#FINISHED ABOVE TESTING AT 7:06PM (41 MIN AFTER)" ] }, { "cell_type": "code", "execution_count": null, "id": "c3262e03-4bb3-4bfe-8cf0-ee77e3e78094", "metadata": {}, "outputs": [], "source": [ "#EVALUATION CODE FOR Granite3.2-2B-FP16" ] }, { "cell_type": "code", "execution_count": null, "id": "d57948cf-26b2-4cb8-9822-44f3b37859e6", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "/home/jovyan/Falcon1B/lib/python3.11/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n", " from .autonotebook import tqdm as notebook_tqdm\n", "[nltk_data] Downloading package punkt to /home/jovyan/nltk_data...\n", "[nltk_data] Package punkt is already up-to-date!\n", "Loading checkpoint shards: 0%| | 0/2 [00:00> Type your question below. Type 'exit' to quit.\")\n", " while True:\n", " user_input = input(\"\\nYour Question: \")\n", " if user_input.lower() in [\"exit\", \"quit\"]:\n", " break\n", " prompt = f\"Question: {user_input.strip()}\\nAnswer:\"\n", " print(\"\\nModel Response:\\n\", generate_response(prompt))\n", "\n", "if __name__ == \"__main__\":\n", " print(\"Testing Granite3.2-2B-Instruct with LoRA merged...\")\n", " test_on_random_prompts(n=15) # change n=10/15/20 as needed\n", " interactive_mode()\n", "\n", "\n", "#######################################################################################\n", "#######################################################################################\n", "#######################################################################################\n", "#######################################################################################\n", "###### I AM WORRIED ABOUT DATA LEAKS EVEN THOUGH I AM NOT USING THE ANSWER COLUMN #####\n", "\n" ] } ], "metadata": { "kernelspec": { "display_name": "Python (Falcon1B)", "language": "python", "name": "falcon1b" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.11.7" } }, "nbformat": 4, "nbformat_minor": 5 }