From 5821b8cf83366131fb49dd8d1e18f277e2b11c63 Mon Sep 17 00:00:00 2001 From: Ace Taffy Date: Tue, 28 Jul 2026 02:17:01 +0900 Subject: [PATCH 1/2] Add GREAD backbone (TDL Challenge 2026, Track 1) GREAD (Choi et al., ICML 2023, arXiv:2211.14208) models node representations as a reaction-diffusion process integrated with explicit Euler steps, with all reaction terms from the paper (bspm, fisher, allen-cahn, zeldovich, st, fb, fb3, none). The reaction term counteracts the over-smoothing that limits deep message passing, which suits the low-homophily GraphUniverse regimes. Includes five model configs, numerical unit tests covering every reaction term and the Euler horizon edge cases, the 72-run challenge evaluation grid, and a supplementary notebook analysing the out-of-distribution measurements that the grid produces. Generated with [Devin](https://devin.ai) Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com> --- 2026_tdl_challenge/analysis_gread.ipynb | 776 +++ .../outputs/2026-07-25_13-46-02/results.json | 5272 +++++++++++++++++ 2026_tdl_challenge/run_evaluation.ipynb | 2 +- configs/model/graph/gread.yaml | 47 + configs/model/graph/gread_allen_cahn.yaml | 47 + configs/model/graph/gread_fisher.yaml | 47 + configs/model/graph/gread_source_term.yaml | 47 + configs/model/graph/gread_zeldovich.yaml | 47 + test/nn/backbones/graph/test_gread.py | 564 ++ test/pipeline/test_pipeline.py | 8 +- topobench/nn/backbones/graph/gread.py | 427 ++ 11 files changed, 7282 insertions(+), 2 deletions(-) create mode 100644 2026_tdl_challenge/analysis_gread.ipynb create mode 100644 2026_tdl_challenge/outputs/2026-07-25_13-46-02/results.json create mode 100644 configs/model/graph/gread.yaml create mode 100644 configs/model/graph/gread_allen_cahn.yaml create mode 100644 configs/model/graph/gread_fisher.yaml create mode 100644 configs/model/graph/gread_source_term.yaml create mode 100644 configs/model/graph/gread_zeldovich.yaml create mode 100644 test/nn/backbones/graph/test_gread.py create mode 100644 topobench/nn/backbones/graph/gread.py diff --git a/2026_tdl_challenge/analysis_gread.ipynb b/2026_tdl_challenge/analysis_gread.ipynb new file mode 100644 index 000000000..5ba560785 --- /dev/null +++ b/2026_tdl_challenge/analysis_gread.ipynb @@ -0,0 +1,776 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "id": "cell-01", + "metadata": {}, + "source": [ + "# GREAD on GraphUniverse — supplementary transfer analysis\n", + "\n", + "Model: GREAD (Choi et al., ICML 2023, [arXiv:2211.14208](https://arxiv.org/abs/2211.14208)), config `graph/gread`.\n", + "\n", + "This notebook is **supplementary** to the official `run_evaluation.ipynb`. It\n", + "re-trains nothing: it reads the committed `results.json` from the official grid\n", + "and analyses the part of it that the official artifacts summarise only\n", + "partially — the **out-of-distribution (OOD) evaluations**.\n", + "\n", + "The official run trains on each of the 12 GraphUniverse settings and evaluates\n", + "every checkpoint on all 12, so each model yields a full 12x12 transfer matrix\n", + "per task, per seed: 432 measurements per task, of which only the 36 diagonal\n", + "ones appear in the main heatmaps.\n", + "\n", + "We use them to ask three questions the challenge itself poses:\n", + "\n", + "1. Is performance governed by the distribution a model is **evaluated** on, or\n", + " by the **mismatch** between training and evaluation distributions?\n", + "2. Is transfer across homophily levels **symmetric**?\n", + "3. Which structural axis — homophily, average degree, or power-law exponent —\n", + " dominates?\n", + "\n", + "The headline result is that the two tasks answer question 1 in **opposite**\n", + "ways." + ] + }, + { + "cell_type": "code", + "execution_count": 1, + "id": "cell-02", + "metadata": { + "execution": { + "iopub.execute_input": "2026-07-26T18:10:52.074950Z", + "iopub.status.busy": "2026-07-26T18:10:52.074874Z", + "iopub.status.idle": "2026-07-26T18:10:52.686136Z", + "shell.execute_reply": "2026-07-26T18:10:52.685739Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "864 measurements (72 in-distribution, 792 OOD)\n", + "\n", + "triangle-counting MSE/triangles spans 0.0139 to 2215 -> analysed as log10\n" + ] + }, + { + "data": { + "text/html": [ + "
\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", + " \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", + "
tasktrainevalseedin_distributionscoretrain_homophilytrain_degreetrain_powerlaweval_homophilyeval_degreeeval_powerlawvalue
0community_detectionh_lo__d_lo__pl_loh_lo__d_lo__pl_lo42True0.314692lolololololo0.314692
1community_detectionh_lo__d_lo__pl_loh_lo__d_lo__pl_hi42False0.312667lololololohi0.312667
2community_detectionh_lo__d_lo__pl_loh_lo__d_hi__pl_lo42False0.328711lolololohilo0.328711
3community_detectionh_lo__d_lo__pl_loh_lo__d_hi__pl_hi42False0.326610lolololohihi0.326610
4community_detectionh_lo__d_lo__pl_loh_mid__d_lo__pl_lo42False0.348079lololomidlolo0.348079
\n", + "
" + ], + "text/plain": [ + " task train eval seed \\\n", + "0 community_detection h_lo__d_lo__pl_lo h_lo__d_lo__pl_lo 42 \n", + "1 community_detection h_lo__d_lo__pl_lo h_lo__d_lo__pl_hi 42 \n", + "2 community_detection h_lo__d_lo__pl_lo h_lo__d_hi__pl_lo 42 \n", + "3 community_detection h_lo__d_lo__pl_lo h_lo__d_hi__pl_hi 42 \n", + "4 community_detection h_lo__d_lo__pl_lo h_mid__d_lo__pl_lo 42 \n", + "\n", + " in_distribution score train_homophily train_degree train_powerlaw \\\n", + "0 True 0.314692 lo lo lo \n", + "1 False 0.312667 lo lo lo \n", + "2 False 0.328711 lo lo lo \n", + "3 False 0.326610 lo lo lo \n", + "4 False 0.348079 lo lo lo \n", + "\n", + " eval_homophily eval_degree eval_powerlaw value \n", + "0 lo lo lo 0.314692 \n", + "1 lo lo hi 0.312667 \n", + "2 lo hi lo 0.328711 \n", + "3 lo hi hi 0.326610 \n", + "4 mid lo lo 0.348079 " + ] + }, + "execution_count": 1, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "import json\n", + "from pathlib import Path\n", + "\n", + "import matplotlib.pyplot as plt\n", + "import numpy as np\n", + "import pandas as pd\n", + "\n", + "STUDY_ID = \"2026-07-25_13-46-02\"\n", + "RESULTS = Path(\"outputs\") / STUDY_ID / \"results.json\"\n", + "if not RESULTS.exists(): # allow running from the repository root\n", + " RESULTS = Path(\"2026_tdl_challenge\") / RESULTS\n", + "\n", + "# metric key, display label, better direction, whether to analyse on a log scale\n", + "TASK_METRIC = {\n", + " \"community_detection\": (\"test_best_rerun_accuracy\", \"accuracy\", \"max\", False),\n", + " \"triangle_counting\": (\"test_mse_by_total_triangles\", \"MSE / triangles\", \"min\", True),\n", + "}\n", + "\n", + "REGIME_ORDER = [\n", + " f\"h_{h}__d_{d}__pl_{p}\"\n", + " for h in (\"lo\", \"mid\", \"hi\")\n", + " for d in (\"lo\", \"hi\")\n", + " for p in (\"lo\", \"hi\")\n", + "]\n", + "HOM_ORDER = [\"lo\", \"mid\", \"hi\"]\n", + "\n", + "\n", + "def load_long(path):\n", + " \"\"\"Flatten results.json into one row per (task, train regime, eval regime, seed).\"\"\"\n", + " payload = json.loads(Path(path).read_text())\n", + " rows = []\n", + " for run in payload[\"results\"]:\n", + " task = run[\"experiment\"]\n", + " metric = TASK_METRIC[task][0]\n", + " train = run[\"run_slug\"]\n", + " rows.append(\n", + " dict(task=task, train=train, eval=train, seed=run[\"train_seed\"],\n", + " in_distribution=True, score=run[metric])\n", + " )\n", + " for eval_slug, metrics in (run.get(\"ood_test\") or {}).items():\n", + " rows.append(\n", + " dict(task=task, train=train, eval=eval_slug, seed=run[\"train_seed\"],\n", + " in_distribution=False, score=metrics[metric])\n", + " )\n", + " frame = pd.DataFrame(rows)\n", + " for col in (\"train\", \"eval\"):\n", + " parts = frame[col].str.split(\"__\", expand=True)\n", + " frame[f\"{col}_homophily\"] = parts[0].str.removeprefix(\"h_\")\n", + " frame[f\"{col}_degree\"] = parts[1].str.removeprefix(\"d_\")\n", + " frame[f\"{col}_powerlaw\"] = parts[2].str.removeprefix(\"pl_\")\n", + " # Analysis scale: triangle-counting errors span several orders of magnitude,\n", + " # so a raw-scale summary would be dominated by a handful of extreme cells.\n", + " frame[\"value\"] = frame.score\n", + " log_tasks = [t for t, spec in TASK_METRIC.items() if spec[3]]\n", + " mask = frame.task.isin(log_tasks)\n", + " frame.loc[mask, \"value\"] = np.log10(frame.loc[mask, \"score\"].clip(lower=1e-6))\n", + " return frame\n", + "\n", + "\n", + "df = load_long(RESULTS)\n", + "print(f\"{len(df)} measurements \"\n", + " f\"({df.in_distribution.sum()} in-distribution, {(~df.in_distribution).sum()} OOD)\")\n", + "print(\"\\ntriangle-counting MSE/triangles spans \"\n", + " f\"{df[df.task == 'triangle_counting'].score.min():.4g} to \"\n", + " f\"{df[df.task == 'triangle_counting'].score.max():.4g} \"\n", + " \"-> analysed as log10\")\n", + "df.head()" + ] + }, + { + "cell_type": "markdown", + "id": "cell-03", + "metadata": {}, + "source": [ + "## 1. Evaluation distribution, or train/eval mismatch?\n", + "\n", + "An OOD drop is usually read as a transfer failure. But the same drop appears if\n", + "the target distribution is simply harder, wherever the model was trained. The\n", + "two explanations make different predictions:\n", + "\n", + "- *Mismatch matters*: scores depend on the **training** regime.\n", + "- *Target difficulty matters*: scores depend on the **evaluation** regime.\n", + "\n", + "We quantify this as the fraction of score variance explained by grouping the\n", + "432 measurements by evaluation regime, versus by training regime." + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "id": "cell-04", + "metadata": { + "execution": { + "iopub.execute_input": "2026-07-26T18:10:52.687651Z", + "iopub.status.busy": "2026-07-26T18:10:52.687518Z", + "iopub.status.idle": "2026-07-26T18:10:52.694178Z", + "shell.execute_reply": "2026-07-26T18:10:52.693786Z" + } + }, + "outputs": [ + { + "data": { + "text/html": [ + "
\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", + "
analysedby_eval_regimeby_train_regime
task
community_detectionaccuracy0.7670.038
triangle_countinglog10(MSE / triangles)0.1050.427
\n", + "
" + ], + "text/plain": [ + " analysed by_eval_regime by_train_regime\n", + "task \n", + "community_detection accuracy 0.767 0.038\n", + "triangle_counting log10(MSE / triangles) 0.105 0.427" + ] + }, + "execution_count": 2, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "def variance_explained(frame, by, col=\"value\"):\n", + " \"\"\"Fraction of variance in `col` explained by grouping on `by`.\"\"\"\n", + " group_mean = frame.groupby(by)[col].transform(\"mean\")\n", + " resid = ((frame[col] - group_mean) ** 2).mean()\n", + " total = ((frame[col] - frame[col].mean()) ** 2).mean()\n", + " return 1.0 - resid / total\n", + "\n", + "\n", + "summary = []\n", + "for task, sub in df.groupby(\"task\"):\n", + " metric, label, _, is_log = TASK_METRIC[task]\n", + " summary.append(\n", + " dict(\n", + " task=task,\n", + " analysed=f\"log10({label})\" if is_log else label,\n", + " by_eval_regime=variance_explained(sub, \"eval\"),\n", + " by_train_regime=variance_explained(sub, \"train\"),\n", + " )\n", + " )\n", + "variance_table = pd.DataFrame(summary).set_index(\"task\").round(3)\n", + "variance_table" + ] + }, + { + "cell_type": "markdown", + "id": "cell-05", + "metadata": {}, + "source": [ + "The two tasks behave in opposite ways, and this is the main result of the\n", + "notebook.\n", + "\n", + "**Community detection** is governed by the evaluation regime: it explains an\n", + "order of magnitude more variance than the training regime. The model transfers\n", + "well; the regimes differ in intrinsic difficulty. Reporting an \"OOD gap\" here\n", + "without this decomposition would attribute to distribution shift what is mostly\n", + "a difficulty effect.\n", + "\n", + "**Triangle counting** is the reverse: the training regime dominates. This is a\n", + "genuine distribution-shift failure. A count is an extensive quantity whose\n", + "typical magnitude changes by orders of magnitude across regimes, so a regressor\n", + "calibrated on one regime systematically mispredicts the scale elsewhere — a\n", + "calibration failure rather than a failure to perceive structure." + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "id": "cell-06", + "metadata": { + "execution": { + "iopub.execute_input": "2026-07-26T18:10:52.695582Z", + "iopub.status.busy": "2026-07-26T18:10:52.695502Z", + "iopub.status.idle": "2026-07-26T18:10:53.111369Z", + "shell.execute_reply": "2026-07-26T18:10:53.110836Z" + } + }, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "fig, axes = plt.subplots(1, 2, figsize=(18, 7.5))\n", + "for ax, (task, (metric, label, direction, is_log)) in zip(axes, TASK_METRIC.items()):\n", + " sub = df[df.task == task]\n", + " matrix = (\n", + " sub.pivot_table(index=\"train\", columns=\"eval\", values=\"value\", aggfunc=\"mean\")\n", + " .reindex(index=REGIME_ORDER, columns=REGIME_ORDER)\n", + " )\n", + " data = matrix.to_numpy()\n", + " im = ax.imshow(data, cmap=\"RdBu_r\" if direction == \"max\" else \"RdBu\")\n", + " ax.set_xticks(range(len(REGIME_ORDER)))\n", + " ax.set_yticks(range(len(REGIME_ORDER)))\n", + " ax.set_xticklabels(REGIME_ORDER, rotation=90, fontsize=7)\n", + " ax.set_yticklabels(REGIME_ORDER, fontsize=7)\n", + " ax.set_xlabel(\"evaluation regime\")\n", + " ax.set_ylabel(\"training regime\")\n", + " shown = f\"log10({label})\" if is_log else label\n", + " ax.set_title(f\"{task}\\n{shown} (mean over seeds)\")\n", + " for i in range(data.shape[0]):\n", + " for j in range(data.shape[1]):\n", + " ax.text(j, i, f\"{data[i, j]:.2f}\", ha=\"center\", va=\"center\", fontsize=6)\n", + " fig.colorbar(im, ax=ax, fraction=0.046, label=shown)\n", + "fig.suptitle(\"Full 12x12 transfer matrices (the official heatmaps show only the diagonal)\")\n", + "fig.tight_layout()\n", + "plt.show()" + ] + }, + { + "cell_type": "markdown", + "id": "cell-07", + "metadata": {}, + "source": [ + "Read the two panels by comparing how each varies. Community detection varies\n", + "mostly left-to-right (along the evaluation axis); triangle counting varies\n", + "mostly top-to-bottom (along the training axis).\n", + "\n", + "## 2. Is transfer across homophily symmetric?\n", + "\n", + "Collapsing onto the homophily axis isolates the structural property the\n", + "challenge cares most about. Symmetric transfer would give a symmetric matrix." + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "id": "cell-08", + "metadata": { + "execution": { + "iopub.execute_input": "2026-07-26T18:10:53.112690Z", + "iopub.status.busy": "2026-07-26T18:10:53.112608Z", + "iopub.status.idle": "2026-07-26T18:10:53.121257Z", + "shell.execute_reply": "2026-07-26T18:10:53.120802Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "community_detection — accuracy (rows = trained on, cols = evaluated on)\n", + "eval_homophily lo mid hi\n", + "train_homophily \n", + "lo 0.318 0.357 0.410\n", + "mid 0.283 0.398 0.559\n", + "hi 0.219 0.384 0.621\n", + " trained hi -> evaluated lo : 0.219 (native lo = 0.318)\n", + " trained lo -> evaluated hi : 0.410 (native hi = 0.621)\n", + " best generalist (row mean over all eval regimes): trained on h_mid\n", + "\n", + "triangle_counting — log10(MSE / triangles) (rows = trained on, cols = evaluated on)\n", + "eval_homophily lo mid hi\n", + "train_homophily \n", + "lo -0.506 0.006 0.430\n", + "mid 0.344 0.276 0.452\n", + "hi 1.186 0.751 0.627\n", + " trained hi -> evaluated lo : 1.186 (native lo = -0.506)\n", + " trained lo -> evaluated hi : 0.430 (native hi = 0.627)\n", + " best generalist (row mean over all eval regimes): trained on h_lo\n" + ] + } + ], + "source": [ + "for task, (metric, label, direction, is_log) in TASK_METRIC.items():\n", + " sub = df[df.task == task]\n", + " shown = f\"log10({label})\" if is_log else label\n", + " hom = (\n", + " sub.pivot_table(index=\"train_homophily\", columns=\"eval_homophily\",\n", + " values=\"value\", aggfunc=\"mean\")\n", + " .reindex(index=HOM_ORDER, columns=HOM_ORDER)\n", + " )\n", + " print(f\"\\n{task} — {shown} (rows = trained on, cols = evaluated on)\")\n", + " print(hom.round(3).to_string())\n", + " print(f\" trained hi -> evaluated lo : {hom.loc['hi', 'lo']:.3f} \"\n", + " f\"(native lo = {hom.loc['lo', 'lo']:.3f})\")\n", + " print(f\" trained lo -> evaluated hi : {hom.loc['lo', 'hi']:.3f} \"\n", + " f\"(native hi = {hom.loc['hi', 'hi']:.3f})\")\n", + " best = hom.mean(axis=1)\n", + " best = best.idxmax() if direction == \"max\" else best.idxmin()\n", + " print(f\" best generalist (row mean over all eval regimes): trained on h_{best}\")" + ] + }, + { + "cell_type": "markdown", + "id": "cell-09", + "metadata": {}, + "source": [ + "Transfer is clearly asymmetric, and the asymmetry runs in opposite directions\n", + "for the two tasks. For community detection the mid-homophily model is the best\n", + "generalist, and a high-homophily model degrades badly on heterophilic graphs.\n", + "For triangle counting the low-homophily model is the safest, because it is\n", + "calibrated to the sparse-triangle end of the scale.\n", + "\n", + "## 3. Which structural axis dominates?\n", + "\n", + "Variance explained by each axis of the *evaluation* regime alone." + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "id": "cell-10", + "metadata": { + "execution": { + "iopub.execute_input": "2026-07-26T18:10:53.122575Z", + "iopub.status.busy": "2026-07-26T18:10:53.122505Z", + "iopub.status.idle": "2026-07-26T18:10:53.132719Z", + "shell.execute_reply": "2026-07-26T18:10:53.132222Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "axis degree homophily powerlaw\n", + "task side \n", + "community_detection eval 0.026 0.722 0.001\n", + " train 0.001 0.035 0.000\n", + "triangle_counting eval 0.003 0.008 0.000\n", + " train 0.147 0.178 0.046\n", + "\n", + "Mean value by regime level (log10 scale for triangle counting):\n", + " community_detection eval_homophily {'hi': 0.53, 'lo': 0.273, 'mid': 0.38}\n", + " community_detection eval_degree {'hi': 0.414, 'lo': 0.374}\n", + " community_detection eval_powerlaw {'hi': 0.398, 'lo': 0.391}\n", + " triangle_counting train_homophily {'hi': 0.855, 'lo': -0.024, 'mid': 0.357}\n", + " triangle_counting train_degree {'hi': 0.723, 'lo': 0.069}\n", + " triangle_counting train_powerlaw {'hi': 0.212, 'lo': 0.58}\n" + ] + } + ], + "source": [ + "axis_rows = []\n", + "for task, sub in df.groupby(\"task\"):\n", + " for prefix in (\"eval\", \"train\"):\n", + " for axis in (\"homophily\", \"degree\", \"powerlaw\"):\n", + " axis_rows.append(\n", + " dict(task=task, side=prefix, axis=axis,\n", + " variance_explained=variance_explained(sub, f\"{prefix}_{axis}\"))\n", + " )\n", + "axis_table = (\n", + " pd.DataFrame(axis_rows)\n", + " .pivot(index=[\"task\", \"side\"], columns=\"axis\", values=\"variance_explained\")\n", + " .round(3)\n", + ")\n", + "print(axis_table.to_string())\n", + "\n", + "print(\"\\nMean value by regime level (log10 scale for triangle counting):\")\n", + "for task, (metric, label, direction, is_log) in TASK_METRIC.items():\n", + " sub = df[df.task == task]\n", + " side = \"eval\" if task == \"community_detection\" else \"train\"\n", + " for axis in (\"homophily\", \"degree\", \"powerlaw\"):\n", + " means = sub.groupby(f\"{side}_{axis}\").value.mean().round(3).to_dict()\n", + " print(f\" {task:20s} {side}_{axis:10s} {means}\")" + ] + }, + { + "cell_type": "markdown", + "id": "cell-11", + "metadata": {}, + "source": [ + "Homophily dominates in both cases — on the evaluation side for community\n", + "detection and on the training side for triangle counting. Average degree and the\n", + "power-law exponent are secondary throughout.\n", + "\n", + "## 4. Seed stability\n", + "\n", + "Three seeds per (task, regime). Small dispersion is what licenses reading the\n", + "differences above as real rather than as noise." + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "id": "cell-12", + "metadata": { + "execution": { + "iopub.execute_input": "2026-07-26T18:10:53.133957Z", + "iopub.status.busy": "2026-07-26T18:10:53.133885Z", + "iopub.status.idle": "2026-07-26T18:10:53.139578Z", + "shell.execute_reply": "2026-07-26T18:10:53.139029Z" + } + }, + "outputs": [ + { + "data": { + "text/html": [ + "
\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", + "
metricmax_seed_stdmedian_seed_stdspread_across_regimes
task
community_detectionaccuracy0.00350.00140.3978
triangle_countingMSE / triangles0.05490.00653.6515
\n", + "
" + ], + "text/plain": [ + " metric max_seed_std median_seed_std \\\n", + "task \n", + "community_detection accuracy 0.0035 0.0014 \n", + "triangle_counting MSE / triangles 0.0549 0.0065 \n", + "\n", + " spread_across_regimes \n", + "task \n", + "community_detection 0.3978 \n", + "triangle_counting 3.6515 " + ] + }, + "execution_count": 6, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "stability = []\n", + "for task, sub in df[df.in_distribution].groupby(\"task\"):\n", + " per_regime = sub.groupby(\"train\").score\n", + " stability.append(\n", + " dict(task=task, metric=TASK_METRIC[task][1],\n", + " max_seed_std=per_regime.std().max(),\n", + " median_seed_std=per_regime.std().median(),\n", + " spread_across_regimes=per_regime.mean().max() - per_regime.mean().min())\n", + " )\n", + "pd.DataFrame(stability).set_index(\"task\").round(4)" + ] + }, + { + "cell_type": "markdown", + "id": "cell-13", + "metadata": {}, + "source": [ + "## Takeaways\n", + "\n", + "1. **The two tasks fail differently.** For community detection, GREAD's score is\n", + " set by the regime it is *evaluated* on (variance explained 0.77)\n", + " and barely by the one it was *trained* on (0.04). The apparent\n", + " OOD penalty is target difficulty, not failed transfer. For triangle counting\n", + " the ordering reverses on a log scale (train 0.43 vs eval\n", + " 0.11): that task does exhibit a real distribution-shift failure.\n", + "2. **Counting fails by miscalibrated scale.** Triangle counts are extensive and\n", + " their magnitude shifts by orders of magnitude across regimes, so a model\n", + " calibrated on one regime mispredicts the scale on another. The worst cells of\n", + " the transfer matrix are exactly the largest scale jumps.\n", + "3. **Homophily transfer is asymmetric**, in opposite directions per task: the\n", + " best community-detection generalist here is the model trained on\n", + " mid homophily, whereas the safest triangle counter is the\n", + " one trained on low homophily, which is calibrated to the sparse-triangle end\n", + " of the scale.\n", + "4. **Methodological note.** Any OOD summary of a regression task on this grid\n", + " should be read on a log scale; on the raw scale a few extreme cells dominate\n", + " the statistics and invert the conclusion.\n", + "\n", + "All numbers derive from the committed `results.json`. Nothing here re-trains or\n", + "modifies the official evaluation pipeline." + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.11.13" + } + }, + "nbformat": 4, + "nbformat_minor": 5 +} diff --git a/2026_tdl_challenge/outputs/2026-07-25_13-46-02/results.json b/2026_tdl_challenge/outputs/2026-07-25_13-46-02/results.json new file mode 100644 index 000000000..85e4e1c25 --- /dev/null +++ b/2026_tdl_challenge/outputs/2026-07-25_13-46-02/results.json @@ -0,0 +1,5272 @@ +{ + "metadata": { + "study_id": "2026-07-25_13-46-02", + "model_config": "graph/gread", + "generated_at_utc": "2026-07-25T17:01:06.057701+00:00", + "n_runs": 72, + "train_seeds": [ + 42, + 43, + 44 + ], + "heatmap_note": "Cells show mean \u00b1 std over train_seeds (in-distribution test)." + }, + "results": [ + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_1.5-2__s42", + "train_seed": 42, + "homophily": "h_lo", + "avg_degree": "d_lo", + "power_law": "pl_lo", + "run_slug": "h_lo__d_lo__pl_lo", + "test_loss": 2.2027204036712646, + "test_best_rerun_accuracy": 0.3146917223930359, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3126671314239502, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3287111222743988, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.32661011815071106, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.34807854890823364, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.34383833408355713, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.35598593950271606, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.3577813506126404, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.38031935691833496, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3824203610420227, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.39800596237182617, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.40927496552467346, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__00__h_lo__d_lo__pl_lo__s42" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_1.5-2__s43", + "train_seed": 43, + "homophily": "h_lo", + "avg_degree": "d_lo", + "power_law": "pl_lo", + "run_slug": "h_lo__d_lo__pl_lo", + "test_loss": 2.1980178356170654, + "test_best_rerun_accuracy": 0.315952330827713, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3117503225803375, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3279089331626892, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.3265337347984314, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3481549322605133, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3421957492828369, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3549163341522217, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.35526013374328613, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3750859498977661, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3789823651313782, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3924669623374939, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.40297195315361023, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__00__h_lo__d_lo__pl_lo__s43" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_1.5-2__s44", + "train_seed": 44, + "homophily": "h_lo", + "avg_degree": "d_lo", + "power_law": "pl_lo", + "run_slug": "h_lo__d_lo__pl_lo", + "test_loss": 2.211678981781006, + "test_best_rerun_accuracy": 0.3149591386318207, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.30919092893600464, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.32416534423828125, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.32252272963523865, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.34425854682922363, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.341508150100708, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3537321388721466, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.35392314195632935, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3761173486709595, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3786003589630127, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3949499726295471, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.4067155718803406, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__00__h_lo__d_lo__pl_lo__s44" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_4-5__s42", + "train_seed": 42, + "homophily": "h_lo", + "avg_degree": "d_lo", + "power_law": "pl_hi", + "run_slug": "h_lo__d_lo__pl_hi", + "test_loss": 2.19852876663208, + "test_best_rerun_accuracy": 0.31450071930885315, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.31545573472976685, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3269539177417755, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.3251967430114746, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3446023464202881, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3418901264667511, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.35040873289108276, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.3525097370147705, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.37527695298194885, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.37325236201286316, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3888379633426666, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.39930474758148193, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__01__h_lo__d_lo__pl_hi__s42" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_4-5__s43", + "train_seed": 43, + "homophily": "h_lo", + "avg_degree": "d_lo", + "power_law": "pl_hi", + "run_slug": "h_lo__d_lo__pl_hi", + "test_loss": 2.2020461559295654, + "test_best_rerun_accuracy": 0.3129345178604126, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.31438612937927246, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.32512032985687256, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.32443273067474365, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3420047461986542, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3402857482433319, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.34536632895469666, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.34700894355773926, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3654213547706604, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.36725494265556335, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3788677453994751, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.3884941637516022, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__01__h_lo__d_lo__pl_hi__s43" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_4-5__s44", + "train_seed": 44, + "homophily": "h_lo", + "avg_degree": "d_lo", + "power_law": "pl_hi", + "run_slug": "h_lo__d_lo__pl_hi", + "test_loss": 2.215632200241089, + "test_best_rerun_accuracy": 0.31167393922805786, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.31423333287239075, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.321605920791626, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.32145312428474426, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.33856675028800964, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.33528152108192444, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.34658873081207275, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.34498435258865356, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3647337555885315, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.36469554901123047, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.37974634766578674, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.3862403631210327, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__01__h_lo__d_lo__pl_hi__s44" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_1.5-2__s42", + "train_seed": 42, + "homophily": "h_lo", + "avg_degree": "d_hi", + "power_law": "pl_lo", + "run_slug": "h_lo__d_hi__pl_lo", + "test_loss": 2.15183687210083, + "test_best_rerun_accuracy": 0.32852011919021606, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3089617192745209, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3028497099876404, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.3258843421936035, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.35934755206108093, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.35392314195632935, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.37584996223449707, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.3787531554698944, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.41007715463638306, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.4181373715400696, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.44556498527526855, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.4679119884967804, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__02__h_lo__d_hi__pl_lo__s42" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_1.5-2__s43", + "train_seed": 43, + "homophily": "h_lo", + "avg_degree": "d_hi", + "power_law": "pl_lo", + "run_slug": "h_lo__d_hi__pl_lo", + "test_loss": 2.1434247493743896, + "test_best_rerun_accuracy": 0.3331805467605591, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.31236153841018677, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3061731159687042, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.3290931284427643, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3573993444442749, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3535793423652649, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3772633373737335, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.37894415855407715, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.40667736530303955, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.4155779778957367, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.4422033727169037, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.4629077911376953, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__02__h_lo__d_hi__pl_lo__s43" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_1.5-2__s44", + "train_seed": 44, + "homophily": "h_lo", + "avg_degree": "d_hi", + "power_law": "pl_lo", + "run_slug": "h_lo__d_hi__pl_lo", + "test_loss": 2.158991575241089, + "test_best_rerun_accuracy": 0.32771793007850647, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.30865612626075745, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.30120712518692017, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.325464129447937, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.35770493745803833, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3555275499820709, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3780273497104645, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.378447562456131, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.41294217109680176, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.4188249707221985, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.44690197706222534, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.47127360105514526, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__02__h_lo__d_hi__pl_lo__s44" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_4-5__s42", + "train_seed": 42, + "homophily": "h_lo", + "avg_degree": "d_hi", + "power_law": "pl_hi", + "run_slug": "h_lo__d_hi__pl_hi", + "test_loss": 2.164921998977661, + "test_best_rerun_accuracy": 0.32554054260253906, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.30854153633117676, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3030407130718231, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.32767972350120544, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3594621419906616, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.35365575551986694, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3731759488582611, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.37856215238571167, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.41160517930984497, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.4196653664112091, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.44816258549690247, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.47096797823905945, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__03__h_lo__d_hi__pl_hi__s42" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_4-5__s43", + "train_seed": 43, + "homophily": "h_lo", + "avg_degree": "d_hi", + "power_law": "pl_hi", + "run_slug": "h_lo__d_hi__pl_hi", + "test_loss": 2.1560213565826416, + "test_best_rerun_accuracy": 0.3271067440509796, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.31098631024360657, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.30808311700820923, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3285965323448181, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.35812515020370483, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.35254794359207153, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.37363433837890625, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.3764611482620239, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.4082435667514801, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.414622962474823, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.43914738297462463, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.45943158864974976, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__03__h_lo__d_hi__pl_hi__s43" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_4-5__s44", + "train_seed": 44, + "homophily": "h_lo", + "avg_degree": "d_hi", + "power_law": "pl_hi", + "run_slug": "h_lo__d_hi__pl_hi", + "test_loss": 2.1702499389648438, + "test_best_rerun_accuracy": 0.32649552822113037, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.30800673365592957, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.302620530128479, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3253113329410553, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.35911834239959717, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.35365575551986694, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.37393996119499207, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.3790205419063568, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.41603636741638184, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.41970357298851013, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.45056918263435364, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.47345098853111267, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__03__h_lo__d_hi__pl_hi__s44" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_1.5-2__s42", + "train_seed": 42, + "homophily": "h_mid", + "avg_degree": "d_lo", + "power_law": "pl_lo", + "run_slug": "h_mid__d_lo__pl_lo", + "test_loss": 1.9939359426498413, + "test_best_rerun_accuracy": 0.38845595717430115, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.28497210144996643, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.2793949246406555, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.30689892172813416, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.30074870586395264, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3811979591846466, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.4097333550453186, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.4174497723579407, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.5027503967285156, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.516808032989502, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.5733058452606201, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.6190312504768372, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__04__h_mid__d_lo__pl_lo__s42" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_1.5-2__s43", + "train_seed": 43, + "homophily": "h_mid", + "avg_degree": "d_lo", + "power_law": "pl_lo", + "run_slug": "h_mid__d_lo__pl_lo", + "test_loss": 1.9884096384048462, + "test_best_rerun_accuracy": 0.3901367485523224, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.2869967222213745, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.2819543182849884, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.30487433075904846, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.29983192682266235, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3789059519767761, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.40831995010375977, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.41767895221710205, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.5052334070205688, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.5202460289001465, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.5740698575973511, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.6203300356864929, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__04__h_mid__d_lo__pl_lo__s43" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_1.5-2__s44", + "train_seed": 44, + "homophily": "h_mid", + "avg_degree": "d_lo", + "power_law": "pl_lo", + "run_slug": "h_mid__d_lo__pl_lo", + "test_loss": 1.9952137470245361, + "test_best_rerun_accuracy": 0.3864695429801941, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.2871495187282562, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.27836352586746216, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.30632591247558594, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.2990679144859314, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3788677453994751, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.40977156162261963, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.4180227816104889, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.5043166279792786, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.5172663927078247, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.5742226243019104, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.6211704611778259, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__04__h_mid__d_lo__pl_lo__s44" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_4-5__s42", + "train_seed": 42, + "homophily": "h_mid", + "avg_degree": "d_lo", + "power_law": "pl_hi", + "run_slug": "h_mid__d_lo__pl_hi", + "test_loss": 2.0197396278381348, + "test_best_rerun_accuracy": 0.37963175773620605, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.2844373285770416, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.2800443172454834, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3004431128501892, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.3001757264137268, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.38249674439430237, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3996867537498474, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.4174497723579407, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.4986630082130432, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.5124531984329224, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.5634502172470093, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.6109328269958496, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__05__h_mid__d_lo__pl_hi__s42" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_4-5__s43", + "train_seed": 43, + "homophily": "h_mid", + "avg_degree": "d_lo", + "power_law": "pl_hi", + "run_slug": "h_mid__d_lo__pl_hi", + "test_loss": 2.0160462856292725, + "test_best_rerun_accuracy": 0.37542974948883057, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.2864237129688263, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.282756507396698, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3004431128501892, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.29994651675224304, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3821147382259369, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3976621627807617, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.41431736946105957, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.4986630082130432, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.5134463906288147, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.5636412501335144, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.6115058660507202, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__05__h_mid__d_lo__pl_hi__s43" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_4-5__s44", + "train_seed": 44, + "homophily": "h_mid", + "avg_degree": "d_lo", + "power_law": "pl_hi", + "run_slug": "h_mid__d_lo__pl_hi", + "test_loss": 2.011624574661255, + "test_best_rerun_accuracy": 0.37676674127578735, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.28894492983818054, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.2828711271286011, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3039957284927368, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.3015127182006836, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.382267564535141, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3988845646381378, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.4143555760383606, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.5001146197319031, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.5137901902198792, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.5611964464187622, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.6087936162948608, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__05__h_mid__d_lo__pl_hi__s44" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_1.5-2__s42", + "train_seed": 42, + "homophily": "h_mid", + "avg_degree": "d_hi", + "power_law": "pl_lo", + "run_slug": "h_mid__d_hi__pl_lo", + "test_loss": 1.9068946838378906, + "test_best_rerun_accuracy": 0.4208495616912842, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.26747649908065796, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.2562839090824127, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.29287952184677124, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.2860417068004608, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3847505450248718, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.37118953466415405, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.4337993860244751, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.5080602169036865, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.5256704092025757, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.5920238494873047, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.6447016596794128, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__06__h_mid__d_hi__pl_lo__s42" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_1.5-2__s43", + "train_seed": 43, + "homophily": "h_mid", + "avg_degree": "d_hi", + "power_law": "pl_lo", + "run_slug": "h_mid__d_hi__pl_lo", + "test_loss": 1.9076277017593384, + "test_best_rerun_accuracy": 0.4181755781173706, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.2659102976322174, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.25605469942092896, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.2917335033416748, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.28539231419563293, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3804721534252167, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.37172433733940125, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.43116357922554016, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.5079838037490845, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.520628035068512, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.5925968289375305, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.6452746391296387, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__06__h_mid__d_hi__pl_lo__s43" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_1.5-2__s44", + "train_seed": 44, + "homophily": "h_mid", + "avg_degree": "d_hi", + "power_law": "pl_lo", + "run_slug": "h_mid__d_hi__pl_lo", + "test_loss": 1.9092298746109009, + "test_best_rerun_accuracy": 0.4204675555229187, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.2661777138710022, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.2574681043624878, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.2918481230735779, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.28611811995506287, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.37963175773620605, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.36885935068130493, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.43185117840766907, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.5057681798934937, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.5177248120307922, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.5874398350715637, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.6438612341880798, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__06__h_mid__d_hi__pl_lo__s44" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_4-5__s42", + "train_seed": 42, + "homophily": "h_mid", + "avg_degree": "d_hi", + "power_law": "pl_hi", + "run_slug": "h_mid__d_hi__pl_hi", + "test_loss": 1.8688710927963257, + "test_best_rerun_accuracy": 0.4370845854282379, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.2569715082645416, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.24967530369758606, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.2807319164276123, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.27687370777130127, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.37825655937194824, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3698907494544983, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.4148903787136078, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.5055007934570312, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.5232256054878235, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.5895026326179504, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.6494002342224121, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__07__h_mid__d_hi__pl_hi__s42" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_4-5__s43", + "train_seed": 43, + "homophily": "h_mid", + "avg_degree": "d_hi", + "power_law": "pl_hi", + "run_slug": "h_mid__d_hi__pl_hi", + "test_loss": 1.8615974187850952, + "test_best_rerun_accuracy": 0.43505996465682983, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.2612881064414978, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.2546412944793701, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.2845519185066223, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.28050270676612854, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3750477433204651, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3707311451435089, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.4139735698699951, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.5062648057937622, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.522843599319458, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.5922530293464661, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.6482542753219604, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__07__h_mid__d_hi__pl_hi__s43" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_4-5__s44", + "train_seed": 44, + "homophily": "h_mid", + "avg_degree": "d_hi", + "power_law": "pl_hi", + "run_slug": "h_mid__d_hi__pl_hi", + "test_loss": 1.8566935062408447, + "test_best_rerun_accuracy": 0.434486985206604, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.26243409514427185, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.2577736973762512, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.28581252694129944, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.2823745012283325, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.378447562456131, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.37149515748023987, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.41523417830467224, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.5106959939002991, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.5280770063400269, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.5913744568824768, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.6498204469680786, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__07__h_mid__d_hi__pl_hi__s44" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_1.5-2__s42", + "train_seed": 42, + "homophily": "h_hi", + "avg_degree": "d_lo", + "power_law": "pl_lo", + "run_slug": "h_hi__d_lo__pl_lo", + "test_loss": 1.5320980548858643, + "test_best_rerun_accuracy": 0.5589044094085693, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.2361142933368683, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.22996409237384796, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.23332569003105164, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.22285889089107513, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3816945552825928, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3701963424682617, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.39090076088905334, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.4064863622188568, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.5736878514289856, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.6494766473770142, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.6999006867408752, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__08__h_hi__d_lo__pl_lo__s42" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_1.5-2__s43", + "train_seed": 43, + "homophily": "h_hi", + "avg_degree": "d_lo", + "power_law": "pl_lo", + "run_slug": "h_hi__d_lo__pl_lo", + "test_loss": 1.5363527536392212, + "test_best_rerun_accuracy": 0.5583696365356445, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.2374894917011261, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.22618229687213898, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.2348918914794922, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.2233554869890213, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3791733384132385, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3693177402019501, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.38627853989601135, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.4054931700229645, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.5715868473052979, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.6474138498306274, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.6968064904212952, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__08__h_hi__d_lo__pl_lo__s43" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_1.5-2__s44", + "train_seed": 44, + "homophily": "h_hi", + "avg_degree": "d_lo", + "power_law": "pl_lo", + "run_slug": "h_hi__d_lo__pl_lo", + "test_loss": 1.545002818107605, + "test_best_rerun_accuracy": 0.5550844073295593, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.23928490281105042, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.2279776930809021, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.23454809188842773, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.2241576910018921, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3764229416847229, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3692031502723694, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.38750094175338745, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.4048437476158142, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.5676904320716858, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.644090473651886, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.6955077052116394, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__08__h_hi__d_lo__pl_lo__s44" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_4-5__s42", + "train_seed": 42, + "homophily": "h_hi", + "avg_degree": "d_lo", + "power_law": "pl_hi", + "run_slug": "h_hi__d_lo__pl_hi", + "test_loss": 1.4838958978652954, + "test_best_rerun_accuracy": 0.5757888555526733, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.23183588683605194, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.2233554869890213, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.23046068847179413, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.21770188212394714, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.377645343542099, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.36668193340301514, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3863167464733124, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.40316295623779297, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.5545114278793335, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.6455420851707458, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.6988692879676819, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__09__h_hi__d_lo__pl_hi__s42" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_4-5__s43", + "train_seed": 43, + "homophily": "h_hi", + "avg_degree": "d_lo", + "power_law": "pl_hi", + "run_slug": "h_hi__d_lo__pl_hi", + "test_loss": 1.4814717769622803, + "test_best_rerun_accuracy": 0.5724654197692871, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.23351669311523438, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.2237756848335266, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.2314920872449875, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.21605928242206573, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.375773549079895, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3675605356693268, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.38517075777053833, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.4035067558288574, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.5566506385803223, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.6483688354492188, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.7006646990776062, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__09__h_hi__d_lo__pl_hi__s43" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_4-5__s44", + "train_seed": 44, + "homophily": "h_hi", + "avg_degree": "d_lo", + "power_law": "pl_hi", + "run_slug": "h_hi__d_lo__pl_hi", + "test_loss": 1.4923393726348877, + "test_best_rerun_accuracy": 0.5687600374221802, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.23638169467449188, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.2240430861711502, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.23279088735580444, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.21892428398132324, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.37783634662628174, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3678279519081116, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.38547635078430176, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.4063717722892761, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.5565360188484192, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.644166886806488, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.6990220546722412, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__09__h_hi__d_lo__pl_hi__s44" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_1.5-2__s42", + "train_seed": 42, + "homophily": "h_hi", + "avg_degree": "d_hi", + "power_law": "pl_lo", + "run_slug": "h_hi__d_hi__pl_lo", + "test_loss": 1.2372453212738037, + "test_best_rerun_accuracy": 0.6587210893630981, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.219153493642807, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.210176482796669, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.22014668583869934, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.20246008038520813, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3768049478530884, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.36060813069343567, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.40205517411231995, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.41225457191467285, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.5612727999687195, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.5721980333328247, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.7097944617271423, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__10__h_hi__d_hi__pl_lo__s42" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_1.5-2__s43", + "train_seed": 43, + "homophily": "h_hi", + "avg_degree": "d_hi", + "power_law": "pl_lo", + "run_slug": "h_hi__d_hi__pl_lo", + "test_loss": 1.237139105796814, + "test_best_rerun_accuracy": 0.6592176556587219, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.21976469457149506, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.20639468729496002, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.2190006822347641, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.20177248120307922, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3734433352947235, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3631293475627899, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.39796775579452515, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.4101153612136841, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.5592482089996338, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.572350800037384, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.7077316641807556, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__10__h_hi__d_hi__pl_lo__s43" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_1.5-2__s44", + "train_seed": 44, + "homophily": "h_hi", + "avg_degree": "d_hi", + "power_law": "pl_lo", + "run_slug": "h_hi__d_hi__pl_lo", + "test_loss": 1.240233063697815, + "test_best_rerun_accuracy": 0.6589120626449585, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.22045229375362396, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.20849567651748657, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.22094888985157013, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.20184887945652008, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3726411461830139, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3610283434391022, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.3950263559818268, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.4074031710624695, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.5573764443397522, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.5696386098861694, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.7117044925689697, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__10__h_hi__d_hi__pl_lo__s44" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_4-5__s42", + "train_seed": 42, + "homophily": "h_hi", + "avg_degree": "d_hi", + "power_law": "pl_hi", + "run_slug": "h_hi__d_hi__pl_hi", + "test_loss": 1.0778677463531494, + "test_best_rerun_accuracy": 0.7097944617271423, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.21006187796592712, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.2036442756652832, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.21036748588085175, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.1941324770450592, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.3679807484149933, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3512873351573944, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.39132094383239746, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.4114905595779419, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.5508823990821838, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.5668882131576538, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.651654064655304, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__11__h_hi__d_hi__pl_hi__s42" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_4-5__s43", + "train_seed": 43, + "homophily": "h_hi", + "avg_degree": "d_hi", + "power_law": "pl_hi", + "run_slug": "h_hi__d_hi__pl_hi", + "test_loss": 1.0654218196868896, + "test_best_rerun_accuracy": 0.7102910876274109, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.21136067807674408, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.20058828592300415, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.21216288208961487, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.1924898773431778, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.36641454696655273, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.3533119559288025, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.39338377118110657, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.4098479747772217, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.5567651987075806, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.570135235786438, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.655512273311615, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__11__h_hi__d_hi__pl_hi__s43" + }, + { + "experiment": "community_detection", + "wandb_project": "challenge_community_detection", + "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_4-5__s44", + "train_seed": 44, + "homophily": "h_hi", + "avg_degree": "d_hi", + "power_law": "pl_hi", + "run_slug": "h_hi__d_hi__pl_hi", + "test_loss": 1.0614569187164307, + "test_best_rerun_accuracy": 0.7123920917510986, + "test_best_rerun_mse": null, + "test_triangles_total_structural": null, + "test_mse_by_total_triangles": null, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.21170447766780853, + "test_best_rerun_mse": null + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.20203988254070282, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.210176482796669, + "test_best_rerun_mse": null + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.19550767540931702, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.368171751499176, + "test_best_rerun_mse": null + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.35610052943229675, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.39296355843544006, + "test_best_rerun_mse": null + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": 0.4139353632926941, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": 0.5557720065116882, + "test_best_rerun_mse": null + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": 0.5720834136009216, + "test_best_rerun_mse": null + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": 0.6558942794799805, + "test_best_rerun_mse": null + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__11__h_hi__d_hi__pl_hi__s44" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_1.5-2__s42", + "train_seed": 42, + "homophily": "h_lo", + "avg_degree": "d_lo", + "power_law": "pl_lo", + "run_slug": "h_lo__d_lo__pl_lo", + "test_loss": 87.79856872558594, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 84.7706298828125, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 0.061073940837761166, + "ood_test": { + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 26.20136070251465, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 0.13790189843428763 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 10183.357421875, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.7723441351441032 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 391.2452087402344, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 0.1318655910819799 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 7205.154296875, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 0.962612464512358 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 163.98965454101562, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 0.14161455487134336 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 170734.9375, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 3.964679024242987 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 15777.7197265625, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 1.1062768003479526 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 45563.41015625, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 2.3355072098134193 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 3589.44384765625, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.6381233506944445 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 752714.5625, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 8.514581660124657 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 261834.453125, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 4.357153963440001 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__00__h_lo__d_lo__pl_lo__s42" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_1.5-2__s43", + "train_seed": 43, + "homophily": "h_lo", + "avg_degree": "d_lo", + "power_law": "pl_lo", + "run_slug": "h_lo__d_lo__pl_lo", + "test_loss": 94.0346908569336, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 88.72521209716797, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 0.0639230634705821, + "ood_test": { + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 39.16325378417969, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 0.20612238833778782 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 10529.6162109375, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.7986057042804323 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 440.15911865234375, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 0.1483515735262365 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 7313.99267578125, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 0.9771533300976953 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 176.3068084716797, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 0.15225112994100146 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 171964.015625, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 3.9932197572218096 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 15931.330078125, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 1.117047404159655 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 45739.92578125, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 2.3445551171895023 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 3617.634521484375, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.6431350260416666 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 754751.875, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 8.537627399522641 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 262293.9375, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 4.364800184713694 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__00__h_lo__d_lo__pl_lo__s43" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_1.5-2__s44", + "train_seed": 44, + "homophily": "h_lo", + "avg_degree": "d_lo", + "power_law": "pl_lo", + "run_slug": "h_lo__d_lo__pl_lo", + "test_loss": 91.22958374023438, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 93.32726287841797, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 0.06723866201615128, + "ood_test": { + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 32.397438049316406, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 0.1705128318385074 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 10556.314453125, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.8006305994027304 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 368.49139404296875, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 0.12419662758441818 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 7264.04150390625, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 0.9704798268411824 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 163.54283142089844, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 0.14122869725466186 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 171505.5, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 3.9825724503065207 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 15698.0771484375, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 1.1006925500236642 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 45547.19140625, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 2.3346758627428366 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 3584.070068359375, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.6371680121527777 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 753381.3125, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 8.522123824983314 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 261664.484375, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 4.354325535004077 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__00__h_lo__d_lo__pl_lo__s44" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_4-5__s42", + "train_seed": 42, + "homophily": "h_lo", + "avg_degree": "d_lo", + "power_law": "pl_hi", + "run_slug": "h_lo__d_lo__pl_hi", + "test_loss": 3.94545316696167, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 3.997734785079956, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 0.021040709395157665, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 214.29408264160156, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 0.15439054945360342 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 12880.9287109375, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.9769380895667426 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 327.1950378417969, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 0.11027807139932487 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 8220.9306640625, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 1.09832073000167 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 182.99462890625, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 0.15802644983268566 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 177818.703125, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 4.12917293156697 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 15884.2216796875, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 1.1137443331711891 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 47834.33203125, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 2.4519110170306013 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 4011.482177734375, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.7131523871527777 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 764179.0, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 8.644265466104091 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 265889.375, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 4.424631404656116 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__01__h_lo__d_lo__pl_hi__s42" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_4-5__s43", + "train_seed": 43, + "homophily": "h_lo", + "avg_degree": "d_lo", + "power_law": "pl_hi", + "run_slug": "h_lo__d_lo__pl_hi", + "test_loss": 2.637273073196411, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 2.6410555839538574, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 0.013900292547125565, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 217.00003051757812, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 0.15634007962361537 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 12980.216796875, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.984468471511187 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 341.5283203125, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 0.11510897213094035 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 8230.470703125, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 1.0995952843186372 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 179.60137939453125, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 0.15509618255140867 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 178184.984375, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 4.137678440809029 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 15936.5751953125, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 1.1174151728588206 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 47872.2578125, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 2.453855031652058 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 4002.09375, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.7114833333333334 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 764736.4375, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 8.650571106184179 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 266129.59375, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 4.428628854442281 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__01__h_lo__d_lo__pl_hi__s43" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_4-5__s44", + "train_seed": 44, + "homophily": "h_lo", + "avg_degree": "d_lo", + "power_law": "pl_hi", + "run_slug": "h_lo__d_lo__pl_hi", + "test_loss": 3.575599193572998, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 3.5890095233917236, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 0.01888952380732486, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 219.34445190429688, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 0.15802914402326865 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 12829.2109375, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.9730156190747061 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 297.767822265625, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 0.10035989965137344 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 8212.6416015625, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 1.0972133068219774 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 180.3816375732422, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 0.15576998063319705 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 177597.78125, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 4.124042849015419 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 15731.7119140625, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 1.103050898475845 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 47818.60546875, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 2.4511048987005997 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 3994.02001953125, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.7100480034722222 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 763774.0, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 8.639684173614018 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 265265.75, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 4.414253740036277 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__01__h_lo__d_lo__pl_hi__s44" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_1.5-2__s42", + "train_seed": 42, + "homophily": "h_lo", + "avg_degree": "d_hi", + "power_law": "pl_lo", + "run_slug": "h_lo__d_hi__pl_lo", + "test_loss": 3781.752685546875, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 3639.166748046875, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.2760080961734452, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 4683.3271484375, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 3.3741550060788903 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 3266.08447265625, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 17.18991827713816 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 1367.878662109375, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 0.4610308938690175 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 5900.775390625, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 0.7883467455744823 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 3303.512451171875, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 2.8527741374541233 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 133401.6875, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 3.0977542146572543 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 10316.59375, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 0.7233623439910251 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 33247.765625, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 1.704227055461582 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 4507.5302734375, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.8013387152777778 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 672685.25, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 7.609303417304843 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 221016.734375, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 3.677911476794302 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__02__h_lo__d_hi__pl_lo__s42" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_1.5-2__s43", + "train_seed": 43, + "homophily": "h_lo", + "avg_degree": "d_hi", + "power_law": "pl_lo", + "run_slug": "h_lo__d_hi__pl_lo", + "test_loss": 4730.625, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 4473.701171875, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.33930232627038304, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 7426.0400390625, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 5.350172938805835 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 8192.68359375, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 43.11938733552632 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 4050.074951171875, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 1.3650404284367628 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 7471.56494140625, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 0.998205068992151 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 7668.984375, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 6.622611722797927 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 130463.4375, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 3.029524370704068 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 8364.19140625, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 0.5864669335471884 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 31084.86328125, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 1.5933601558895893 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 6533.189453125, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 1.1614559027777778 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 657630.0, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 7.439000938882165 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 206809.140625, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 3.4414847091175345 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__02__h_lo__d_hi__pl_lo__s43" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_1.5-2__s44", + "train_seed": 44, + "homophily": "h_lo", + "avg_degree": "d_hi", + "power_law": "pl_lo", + "run_slug": "h_lo__d_hi__pl_lo", + "test_loss": 2995.67431640625, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 3030.6328125, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.22985459328782706, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 1809.7470703125, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 1.3038523561329252 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 856.2131958007812, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 4.506385241056743 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 1417.048095703125, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 0.4776029982147371 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 5545.8681640625, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 0.7409309504425518 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 862.7684936523438, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 0.7450505126531466 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 134738.71875, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 3.1288017543655955 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 11481.5048828125, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 0.8050417110371968 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 38070.41015625, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 1.9514280668537598 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 3524.151611328125, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.6265158420138889 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 689925.4375, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 7.80432154451772 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 228730.859375, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 3.806281253640191 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__02__h_lo__d_hi__pl_lo__s44" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_4-5__s42", + "train_seed": 42, + "homophily": "h_lo", + "avg_degree": "d_hi", + "power_law": "pl_hi", + "run_slug": "h_lo__d_hi__pl_hi", + "test_loss": 141.59347534179688, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 137.14686584472656, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 0.04622408690418826, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 162.5874786376953, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 0.11713795290900239 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 202.92503356933594, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 1.0680264924701892 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 10650.7060546875, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.8077896135523321 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 7001.46728515625, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 0.9353997708959586 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 209.96044921875, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 0.181312995871114 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 169328.984375, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 3.9320310323007615 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 13579.9501953125, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 0.9521771277038634 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 44206.00390625, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 2.2659287460274746 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 3260.388427734375, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.579624609375 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 746189.25, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 8.440768412836668 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 254947.46875, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 4.2425485289467995 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__03__h_lo__d_hi__pl_hi__s42" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_4-5__s43", + "train_seed": 43, + "homophily": "h_lo", + "avg_degree": "d_hi", + "power_law": "pl_hi", + "run_slug": "h_lo__d_hi__pl_hi", + "test_loss": 136.92222595214844, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 133.60507202148438, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 0.04503035794455153, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 155.73580932617188, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 0.11220159173355322 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 176.49093627929688, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 0.9288996646278783 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 10851.1845703125, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.822994658347554 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 7126.55322265625, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 0.952111318992151 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 188.96495056152344, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 0.1631821680151325 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 170130.109375, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 3.950634157881293 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 13553.0517578125, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 0.9502911062833053 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 44547.57421875, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 2.2834370915346764 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 3278.650390625, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.5828711805555555 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 748030.75, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 8.461599153874868 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 255422.21875, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 4.2504487835521605 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__03__h_lo__d_hi__pl_hi__s43" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_4-5__s44", + "train_seed": 44, + "homophily": "h_lo", + "avg_degree": "d_hi", + "power_law": "pl_hi", + "run_slug": "h_lo__d_hi__pl_hi", + "test_loss": 137.71493530273438, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 136.60125732421875, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 0.04604019458180612, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 157.6785125732422, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 0.11360123384239351 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 61.32759094238281, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 0.32277679443359375 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 11010.056640625, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.8350441138130451 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 7674.73095703125, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 1.025348157252004 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 155.83151245117188, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 0.13456952715990664 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 171492.6875, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 3.9822749280141183 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 13858.89453125, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 0.9717356984469219 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 46144.7890625, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 2.3653077585985955 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 3617.7861328125, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.6431619791666666 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 751923.75, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 8.505636120946122 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 256966.0, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 4.276138651756444 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__03__h_lo__d_hi__pl_hi__s44" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_1.5-2__s42", + "train_seed": 42, + "homophily": "h_mid", + "avg_degree": "d_lo", + "power_law": "pl_lo", + "run_slug": "h_mid__d_lo__pl_lo", + "test_loss": 5257.2451171875, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 5374.14990234375, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 0.7179892989103206, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 1289.4329833984375, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 0.9289862992784131 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 1900.9085693359375, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 10.004781943873356 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 6492.744140625, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.49243414035836175 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 1523.965576171875, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 0.5136385494344035 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 1454.740234375, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 1.256252361291019 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 150097.625, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 3.485454788222181 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 9604.4755859375, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 0.6734311867856893 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 37404.9296875, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 1.9173166070787842 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 2736.078857421875, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.4864140190972222 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 706084.8125, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 7.987113700892504 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 230611.890625, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 3.837583256369294 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__04__h_mid__d_lo__pl_lo__s42" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_1.5-2__s43", + "train_seed": 43, + "homophily": "h_mid", + "avg_degree": "d_lo", + "power_law": "pl_lo", + "run_slug": "h_mid__d_lo__pl_lo", + "test_loss": 5255.0380859375, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 5427.216796875, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 0.7250790643787575, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 1388.0482177734375, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 1.000034739029854 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 2033.54296875, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 10.702857730263158 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 7138.91845703125, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.5414424313258438 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 936.6683349609375, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 0.3156954280286274 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 1641.92431640625, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 1.4178966462920986 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 152554.234375, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 3.542500333805499 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 10032.2578125, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 0.703425733592764 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 37066.734375, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 1.899981258649854 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 2791.48876953125, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.49626467013888886 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 708865.9375, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 8.018573323303508 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 233214.046875, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 3.8808854088662574 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__04__h_mid__d_lo__pl_lo__s43" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_1.5-2__s44", + "train_seed": 44, + "homophily": "h_mid", + "avg_degree": "d_lo", + "power_law": "pl_lo", + "run_slug": "h_mid__d_lo__pl_lo", + "test_loss": 5430.181640625, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 5587.08349609375, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 0.7464373408274884, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 1241.96435546875, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 0.894786999617255 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 1809.4840087890625, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 9.523600046258224 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 6959.564453125, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.5278395489666288 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 1121.3714599609375, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 0.3779479137043942 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 1441.8082275390625, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 1.2450848251632665 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 152518.625, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 3.5416734395318596 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 10085.66015625, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 0.7071701133256205 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 37927.12890625, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 1.9440837001512123 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 2825.195556640625, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.5022569878472222 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 710137.125, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 8.032952784407769 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 233200.609375, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 3.8806617971311135 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__04__h_mid__d_lo__pl_lo__s44" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_4-5__s42", + "train_seed": 42, + "homophily": "h_mid", + "avg_degree": "d_lo", + "power_law": "pl_hi", + "run_slug": "h_mid__d_lo__pl_hi", + "test_loss": 140.5848846435547, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 148.2747802734375, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 0.12804385170417745, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 157.74961853027344, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 0.11365246291806444 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 28.901439666748047, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 0.1521128403513055 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 12167.857421875, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.9228560805365946 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 231.3026580810547, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 0.07795842874319335 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 7747.02490234375, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 1.0350066669797928 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 175088.234375, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 4.065768028399591 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 15111.373046875, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 1.0595549745389847 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 46514.1015625, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 2.384238124070942 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 3693.9208984375, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.6566970486111111 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 758421.4375, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 8.579136878838954 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 262386.5625, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 4.366341545604313 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__05__h_mid__d_lo__pl_hi__s42" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_4-5__s43", + "train_seed": 43, + "homophily": "h_mid", + "avg_degree": "d_lo", + "power_law": "pl_hi", + "run_slug": "h_mid__d_lo__pl_hi", + "test_loss": 139.52407836914062, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 146.0721893310547, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 0.12614178698709386, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 170.1674041748047, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 0.12259899436225122 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 35.388675689697266, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 0.18625618784051193 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 12316.30859375, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.9341151758627227 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 228.41705322265625, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 0.07698586222536442 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 7770.365234375, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 1.0381249478122911 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 175380.28125, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 4.072549722506038 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 14979.7880859375, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 1.0503287116770088 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 46556.9296875, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 2.386433424957712 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 3661.864501953125, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.6509981336805556 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 758924.5, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 8.584827437982874 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 262227.5625, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 4.363695646747542 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__05__h_mid__d_lo__pl_hi__s43" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_4-5__s44", + "train_seed": 44, + "homophily": "h_mid", + "avg_degree": "d_lo", + "power_law": "pl_hi", + "run_slug": "h_mid__d_lo__pl_hi", + "test_loss": 151.0240478515625, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 158.71817016601562, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 0.13706232311400313, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 164.77513122558594, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 0.11871407148817431 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 40.075828552246094, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 0.2109254134328742 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 12168.087890625, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.9228735601535836 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 229.33416748046875, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 0.07729496713194094 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 7734.0068359375, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 1.0332674463510354 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 175163.96875, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 4.067526675413339 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 15078.30859375, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 1.0572366143423082 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 46495.76953125, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 2.3832984535983393 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 3675.234619140625, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.6533750434027777 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 758374.6875, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 8.578608050631766 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 261976.296875, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 4.359514367313997 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__05__h_mid__d_lo__pl_hi__s44" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_1.5-2__s42", + "train_seed": 42, + "homophily": "h_mid", + "avg_degree": "d_hi", + "power_law": "pl_lo", + "run_slug": "h_mid__d_hi__pl_lo", + "test_loss": 114037.5, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 79800.609375, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 1.8530700672255247, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 68935.5, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 49.66534582132565 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 76215.3828125, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 401.13359375 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 35490.3125, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 2.691718809252939 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 67818.5234375, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 22.85760816902595 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 52983.8671875, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 7.078672970941883 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 72574.1484375, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 62.67197619818653 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 42939.69921875, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 3.010776834858365 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 44125.953125, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 2.261825471577221 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 59294.69921875, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 10.54127986111111 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 479049.65625, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 5.41892985814961 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 124289.984375, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 2.068293884063036 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__06__h_mid__d_hi__pl_lo__s42" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_1.5-2__s43", + "train_seed": 43, + "homophily": "h_mid", + "avg_degree": "d_hi", + "power_law": "pl_lo", + "run_slug": "h_mid__d_hi__pl_lo", + "test_loss": 112811.703125, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 79222.84375, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 1.839653625998514, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 72659.921875, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 52.34864688400577 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 80542.1015625, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 423.9057976973684 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 37689.73046875, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 2.8585309418847173 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 71708.2265625, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 24.168596751769464 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 56212.0390625, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 7.509958458583834 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 76693.90625, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 66.22962543177893 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 45696.79296875, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 3.2040943043577337 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 46085.1484375, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 2.362250675970065 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 62861.3203125, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 11.175345833333333 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 474396.375, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 5.366292716310532 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 122311.703125, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 2.03537355640424 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__06__h_mid__d_hi__pl_lo__s43" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_1.5-2__s44", + "train_seed": 44, + "homophily": "h_mid", + "avg_degree": "d_hi", + "power_law": "pl_lo", + "run_slug": "h_mid__d_hi__pl_lo", + "test_loss": 113326.296875, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 79597.296875, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 1.8483488964099943, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 70258.875, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 50.61878602305475 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 76866.59375, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 404.5610197368421 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 36926.33203125, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 2.8006319325938565 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 68707.9765625, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 23.157390145770137 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 54116.28125, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 7.2299640948563795 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 73382.203125, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 63.36977817357513 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 43582.765625, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 3.055866331860889 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 44945.671875, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 2.303842937874827 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 60275.953125, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 10.715725 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 477642.8125, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 5.403015876158049 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 123579.5625, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 2.0564718436423544 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__06__h_mid__d_hi__pl_lo__s44" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_4-5__s42", + "train_seed": 42, + "homophily": "h_mid", + "avg_degree": "d_hi", + "power_law": "pl_hi", + "run_slug": "h_mid__d_hi__pl_hi", + "test_loss": 8595.5, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 8167.642578125, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 0.5726856386288739, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 5147.3974609375, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 3.708499611626441 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 6535.10107421875, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 34.39526881167763 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 5018.2138671875, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.3806002174582859 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 4475.4990234375, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 1.5084256904069768 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 6020.1455078125, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 0.8042946570223781 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 5679.240234375, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 4.904352534002591 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 133994.984375, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 3.1115313109557867 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 31297.95703125, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 1.6042829991926804 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 4724.83642578125, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.8399709201388889 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 665115.4375, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 7.5236749601257875 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 208794.03125, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 3.4745150225483834 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__07__h_mid__d_hi__pl_hi__s42" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_4-5__s43", + "train_seed": 43, + "homophily": "h_mid", + "avg_degree": "d_hi", + "power_law": "pl_hi", + "run_slug": "h_mid__d_hi__pl_hi", + "test_loss": 8443.4091796875, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 8042.169921875, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 0.5638879485258028, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 5105.32861328125, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 3.6781906435743874 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 6542.2548828125, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 34.432920435855266 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 5097.06494140625, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.38658057955299585 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 4294.68798828125, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 1.4474849977355073 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 6060.990234375, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 0.8097515343186372 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 5671.640625, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 4.897789831606218 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 134758.1875, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 3.1292538431172208 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 31428.107421875, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 1.6109542991375776 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 4679.52490234375, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.8319155381944444 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 666766.875, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 7.542355745845729 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 209098.015625, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 3.4795735880218994 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__07__h_mid__d_hi__pl_hi__s43" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_4-5__s44", + "train_seed": 44, + "homophily": "h_mid", + "avg_degree": "d_hi", + "power_law": "pl_hi", + "run_slug": "h_mid__d_hi__pl_hi", + "test_loss": 8486.0244140625, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 8143.67822265625, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 0.5710053444577373, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 5631.97265625, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 4.0576171875 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 7004.40576171875, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 36.865293482730266 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 5033.87353515625, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.3817879055863671 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 4876.11328125, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 1.6434490331142568 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 6416.95166015625, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 0.8573081710295591 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 6146.17236328125, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 5.307575443248057 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 133150.296875, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 3.0919166095810886 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 31501.736328125, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 1.6147283985916756 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 5078.74072265625, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.9028872395833333 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 662870.625, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 7.498282015316223 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 206839.375, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 3.441987835521608 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__07__h_mid__d_hi__pl_hi__s44" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_1.5-2__s42", + "train_seed": 42, + "homophily": "h_hi", + "avg_degree": "d_lo", + "power_law": "pl_lo", + "run_slug": "h_hi__d_lo__pl_lo", + "test_loss": 26912.43359375, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 27320.7109375, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 1.4004157536265314, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 13399.8154296875, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 9.654045698622118 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 16025.248046875, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 84.34341077302632 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 6593.76513671875, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.5000959527280053 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 16185.9501953125, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 5.455325310182845 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 9942.73046875, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 1.3283541040414162 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 14064.2119140625, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 12.145260720261227 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 111503.5625, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 2.589252333735835 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 10522.2880859375, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 0.7377848889312508 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 10269.8876953125, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 1.8257578125 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 609100.375, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 6.8900419103424095 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 178438.59375, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 2.9693740327492386 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__08__h_hi__d_lo__pl_lo__s42" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_1.5-2__s43", + "train_seed": 43, + "homophily": "h_hi", + "avg_degree": "d_lo", + "power_law": "pl_lo", + "run_slug": "h_hi__d_lo__pl_lo", + "test_loss": 27183.927734375, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 27604.82421875, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 1.4149789440130196, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 12871.2861328125, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 9.273260902602665 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 15470.7685546875, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 81.42509765625 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 6133.82421875, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.4652123032802427 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 14489.2333984375, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 4.8834625542425005 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 9831.2646484375, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 1.3134622108800935 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 13759.1298828125, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 11.88180473472582 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 113593.4296875, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 2.6377816665312093 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 9971.8212890625, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 0.6991881425510097 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 10068.2685546875, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 1.7899144097222222 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 613865.5, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 6.943944210038121 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 180074.6875, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 2.996600061571231 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__08__h_hi__d_lo__pl_lo__s43" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_1.5-2__s44", + "train_seed": 44, + "homophily": "h_hi", + "avg_degree": "d_lo", + "power_law": "pl_lo", + "run_slug": "h_hi__d_lo__pl_lo", + "test_loss": 27479.0625, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 27932.48828125, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 1.4317744774847507, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 13285.484375, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 9.571674621757925 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 15702.0595703125, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 82.64241879111842 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 6248.439453125, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.4739051538206295 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 14116.6982421875, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 4.757903013881867 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 10264.9541015625, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 1.3714033535821644 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 14139.0400390625, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 12.209879135632557 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 113889.1171875, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 2.6446479005085455 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 10075.3671875, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 0.7064484074814191 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 10568.3359375, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 1.8788152777777778 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 613494.0, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 6.939741863963893 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 180730.046875, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 3.0075058139051136 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__08__h_hi__d_lo__pl_lo__s44" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_4-5__s42", + "train_seed": 42, + "homophily": "h_hi", + "avg_degree": "d_lo", + "power_law": "pl_hi", + "run_slug": "h_hi__d_lo__pl_hi", + "test_loss": 2638.443115234375, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 2797.404052734375, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.49731627604166667, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 456.7854309082031, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 0.3290961317782443 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 750.0567626953125, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 3.947667172080592 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 8805.228515625, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.6678216545790672 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 311.92974853515625, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 0.1051330463549566 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 6033.4267578125, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 0.8060690391199065 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 582.322265625, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 0.5028689685880829 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 161415.3125, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 3.7482656627345348 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 11794.5849609375, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 0.8269937569020824 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 40758.4453125, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 2.089212430801169 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 729070.9375, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 8.247128915308304 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 244943.390625, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 4.0760719322550045 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__09__h_hi__d_lo__pl_hi__s42" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_4-5__s43", + "train_seed": 43, + "homophily": "h_hi", + "avg_degree": "d_lo", + "power_law": "pl_hi", + "run_slug": "h_hi__d_lo__pl_hi", + "test_loss": 2535.148193359375, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 2684.48974609375, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.4772426215277778, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 563.5260620117188, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 0.4059986037548406 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 919.5280151367188, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 4.83962113229852 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 8234.5439453125, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.6245387899364808 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 562.0038452148438, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 0.18941821544147075 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 5981.49169921875, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 0.7991304875375751 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 677.7760620117188, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 0.5852988445697053 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 159199.953125, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 3.696822244217908 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 10987.5, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 0.7704038704249053 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 40434.38671875, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 2.0726017078656005 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 725228.9375, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 8.2036688517358 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 241114.109375, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 4.012349348093788 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__09__h_hi__d_lo__pl_hi__s43" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_4-5__s44", + "train_seed": 44, + "homophily": "h_hi", + "avg_degree": "d_lo", + "power_law": "pl_hi", + "run_slug": "h_hi__d_lo__pl_hi", + "test_loss": 2634.956298828125, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 2796.9013671875, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 0.4972269097222222, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 542.5733032226562, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 0.3909029562122884 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 865.4780883789062, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 4.555147833573191 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 8386.9013671875, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 0.6360941499573379 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 465.0816650390625, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 0.1567514880482179 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 6042.7861328125, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 0.8073194566215764 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 675.635498046875, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 0.583450343736507 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 159965.765625, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 3.7146053693340146 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 11366.46484375, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 0.7969755184230823 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 40628.6328125, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 2.082558450586909 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 726139.1875, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 8.213965448005158 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 242234.234375, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 4.0309892063135475 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__09__h_hi__d_lo__pl_hi__s44" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_1.5-2__s42", + "train_seed": 42, + "homophily": "h_hi", + "avg_degree": "d_hi", + "power_law": "pl_lo", + "run_slug": "h_hi__d_hi__pl_lo", + "test_loss": 504694.40625, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 324194.34375, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 3.667232376163705, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 372864.625, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 268.6344560518732 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 400687.8125, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 2108.8832236842104 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 278948.46875, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 21.15650123246113 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 395364.125, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 133.2538338388945 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 324730.53125, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 43.384172511690046 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 387715.84375, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 334.81506368739207 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 165992.984375, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 3.854564935328813 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 311005.09375, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 21.80655544453793 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 254248.140625, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 13.032351254549182 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 347621.375, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 61.79935555555556 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 162152.5, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 2.698359209891335 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__10__h_hi__d_hi__pl_lo__s42" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_1.5-2__s43", + "train_seed": 43, + "homophily": "h_hi", + "avg_degree": "d_hi", + "power_law": "pl_lo", + "run_slug": "h_hi__d_hi__pl_lo", + "test_loss": 504446.0625, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 325006.71875, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 3.676421826747961, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 372370.15625, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 268.27821055475505 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 420939.65625, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 2215.471875 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 319796.6875, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 24.254583807356845 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 497963.03125, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 167.83384942703066 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 324205.96875, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 43.31409068136273 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 396425.96875, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 342.3367605785838 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 188871.515625, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 4.385833076932008 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 386584.84375, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 27.105934914458 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 258376.34375, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 13.243956315034087 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 355812.21875, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 63.25550555555556 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 194173.546875, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 3.2312173942888522 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__10__h_hi__d_hi__pl_lo__s43" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_1.5-2__s44", + "train_seed": 44, + "homophily": "h_hi", + "avg_degree": "d_hi", + "power_law": "pl_lo", + "run_slug": "h_hi__d_hi__pl_lo", + "test_loss": 503276.6875, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 323960.75, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 3.6645900026017215, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 389638.25, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 280.7192002881844 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 418526.71875, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 2202.7722039473683 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 291043.6875, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 22.073848122866895 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 411130.34375, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 138.5676925345467 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 340389.4375, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 45.47621075484302 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 405894.15625, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 350.51308829879105 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 173506.9375, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 4.029048335036225 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 325409.6875, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 22.816553603982612 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 267784.8125, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 13.726219309036855 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 364238.8125, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 64.75356666666667 + }, + "h_hi__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 167508.921875, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 2.787494747724361 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__10__h_hi__d_hi__pl_lo__s44" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_4-5__s42", + "train_seed": 42, + "homophily": "h_hi", + "avg_degree": "d_hi", + "power_law": "pl_hi", + "run_slug": "h_hi__d_hi__pl_hi", + "test_loss": 112238.046875, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 108224.453125, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 1.8009494138252375, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 122463.5703125, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 88.23023797730548 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 132866.03125, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 699.2949013157895 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 77175.2265625, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 5.853259504171407 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 128721.5625, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 43.3844160768453 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 98648.359375, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 13.179473530394121 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 126825.3359375, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 109.52101549006909 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 80075.2265625, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 1.859447022164685 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 87367.6953125, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 6.125907678621512 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 74841.546875, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 3.8362574645035625 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 107863.3828125, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 19.1757125 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 417606.28125, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 4.723892642218024 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__11__h_hi__d_hi__pl_hi__s42" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_4-5__s43", + "train_seed": 43, + "homophily": "h_hi", + "avg_degree": "d_hi", + "power_law": "pl_hi", + "run_slug": "h_hi__d_hi__pl_hi", + "test_loss": 111056.9140625, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 107066.0703125, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 1.781672912194432, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 126469.625, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 91.11644452449568 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 136628.953125, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 719.0997532894737 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 77999.96875, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 5.91581105422829 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 127268.8125, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 42.89478008088979 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 102083.4765625, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 13.63840702237809 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 130883.796875, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 113.02573132556131 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 80019.90625, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 1.8581624152424299 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 86924.9609375, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 6.094864741095218 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 76914.8359375, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 3.942530931236865 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 111583.9765625, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 19.837151388888888 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 416493.125, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 4.711300804271349 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__11__h_hi__d_hi__pl_hi__s43" + }, + { + "experiment": "triangle_counting", + "wandb_project": "challenge_triangle_counting", + "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_4-5__s44", + "train_seed": 44, + "homophily": "h_hi", + "avg_degree": "d_hi", + "power_law": "pl_hi", + "run_slug": "h_hi__d_hi__pl_hi", + "test_loss": 110714.2109375, + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 106493.765625, + "test_triangles_total_structural": 60093.0, + "test_mse_by_total_triangles": 1.7721492623932904, + "ood_test": { + "h_lo__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 129045.0390625, + "test_triangles_total_structural": 1388.0, + "test_mse_by_total_triangles": 92.97193016030259 + }, + "h_lo__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 137913.546875, + "test_triangles_total_structural": 190.0, + "test_mse_by_total_triangles": 725.8607730263158 + }, + "h_lo__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 80733.359375, + "test_triangles_total_structural": 13185.0, + "test_mse_by_total_triangles": 6.123121681835419 + }, + "h_lo__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 128834.28125, + "test_triangles_total_structural": 2967.0, + "test_mse_by_total_triangles": 43.42240689248399 + }, + "h_mid__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 104496.15625, + "test_triangles_total_structural": 7485.0, + "test_mse_by_total_triangles": 13.960742317969272 + }, + "h_mid__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 132674.8125, + "test_triangles_total_structural": 1158.0, + "test_mse_by_total_triangles": 114.57237694300518 + }, + "h_mid__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 81772.6328125, + "test_triangles_total_structural": 43064.0, + "test_mse_by_total_triangles": 1.8988629205949286 + }, + "h_mid__d_hi__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 88149.8359375, + "test_triangles_total_structural": 14262.0, + "test_mse_by_total_triangles": 6.180748558231665 + }, + "h_hi__d_lo__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 79145.4453125, + "test_triangles_total_structural": 19509.0, + "test_mse_by_total_triangles": 4.056868384463581 + }, + "h_hi__d_lo__pl_hi": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 113563.1484375, + "test_triangles_total_structural": 5625.0, + "test_mse_by_total_triangles": 20.189004166666667 + }, + "h_hi__d_hi__pl_lo": { + "test_best_rerun_accuracy": null, + "test_best_rerun_mse": 416349.90625, + "test_triangles_total_structural": 88403.0, + "test_mse_by_total_triangles": 4.70968073764465 + } + }, + "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__11__h_hi__d_hi__pl_hi__s44" + } + ] +} diff --git a/2026_tdl_challenge/run_evaluation.ipynb b/2026_tdl_challenge/run_evaluation.ipynb index 8542dbaab..5ab3b385d 100644 --- a/2026_tdl_challenge/run_evaluation.ipynb +++ b/2026_tdl_challenge/run_evaluation.ipynb @@ -104,7 +104,7 @@ "outputs": [], "source": [ "# Your model configuration (e.g., \"graph/gcn\", \"graph/gin\", \"graph/gat\")\n", - "MODEL_CONFIG = \"graph/gin\"" + "MODEL_CONFIG = \"graph/gread\"" ] }, { diff --git a/configs/model/graph/gread.yaml b/configs/model/graph/gread.yaml new file mode 100644 index 000000000..a059f4b8d --- /dev/null +++ b/configs/model/graph/gread.yaml @@ -0,0 +1,47 @@ +_target_: topobench.model.TBModel + +model_name: gread +model_domain: graph + +feature_encoder: + _target_: topobench.nn.encoders.${model.feature_encoder.encoder_name} + encoder_name: AllCellFeatureEncoder + in_channels: ${infer_in_channels:${dataset},${oc.select:transforms,null}} + out_channels: 64 + proj_dropout: 0.0 + +backbone: + _target_: topobench.nn.backbones.GREAD + in_channels: ${model.feature_encoder.out_channels} + hidden_channels: ${model.feature_encoder.out_channels} + reaction_term: bspm # GREAD-BS. Options: bspm, fisher, allen-cahn, zeldovich, st, fb, fb3, none + time: 3.0 + step_size: 1.0 + beta_diag: false + add_source: false + data_norm: rw + self_loop_weight: 1.0 + # Terminal ReLU pinned explicitly for the submitted benchmark runs + # (the constructor default is false, matching the reference CLI default). + xn_activation: true + input_dropout: 0.0 + dropout: 0.0 + +backbone_wrapper: + _target_: topobench.nn.wrappers.GNNWrapper + _partial_: true + wrapper_name: GNNWrapper + out_channels: ${model.feature_encoder.out_channels} + num_cell_dimensions: ${infer_num_cell_dimensions:${oc.select:model.feature_encoder.selected_dimensions,null},${model.feature_encoder.in_channels}} + +readout: + _target_: topobench.nn.readouts.${model.readout.readout_name} + readout_name: NoReadOut # Use in case readout is not needed Options: PropagateSignalDown + num_cell_dimensions: ${infer_num_cell_dimensions:${oc.select:model.feature_encoder.selected_dimensions,null},${model.feature_encoder.in_channels}} # The highest order of cell dimensions to consider + hidden_dim: ${model.feature_encoder.out_channels} + out_channels: ${dataset.parameters.num_classes} + task_level: ${define_task_level:${dataset.parameters.task_level},${dataset.split_params.learning_setting}} # Handles the edge case of node-inductive task + pooling_type: sum + +# compile model for faster training with pytorch 2.0 +compile: false diff --git a/configs/model/graph/gread_allen_cahn.yaml b/configs/model/graph/gread_allen_cahn.yaml new file mode 100644 index 000000000..7ce21327c --- /dev/null +++ b/configs/model/graph/gread_allen_cahn.yaml @@ -0,0 +1,47 @@ +_target_: topobench.model.TBModel + +model_name: gread_allen_cahn +model_domain: graph + +feature_encoder: + _target_: topobench.nn.encoders.${model.feature_encoder.encoder_name} + encoder_name: AllCellFeatureEncoder + in_channels: ${infer_in_channels:${dataset},${oc.select:transforms,null}} + out_channels: 64 + proj_dropout: 0.0 + +backbone: + _target_: topobench.nn.backbones.GREAD + in_channels: ${model.feature_encoder.out_channels} + hidden_channels: ${model.feature_encoder.out_channels} + reaction_term: allen-cahn + time: 3.0 + step_size: 1.0 + beta_diag: false + add_source: false + data_norm: rw + self_loop_weight: 1.0 + # Terminal ReLU pinned explicitly for the submitted benchmark runs + # (the constructor default is false, matching the reference CLI default). + xn_activation: true + input_dropout: 0.0 + dropout: 0.0 + +backbone_wrapper: + _target_: topobench.nn.wrappers.GNNWrapper + _partial_: true + wrapper_name: GNNWrapper + out_channels: ${model.feature_encoder.out_channels} + num_cell_dimensions: ${infer_num_cell_dimensions:${oc.select:model.feature_encoder.selected_dimensions,null},${model.feature_encoder.in_channels}} + +readout: + _target_: topobench.nn.readouts.${model.readout.readout_name} + readout_name: NoReadOut # Use in case readout is not needed Options: PropagateSignalDown + num_cell_dimensions: ${infer_num_cell_dimensions:${oc.select:model.feature_encoder.selected_dimensions,null},${model.feature_encoder.in_channels}} # The highest order of cell dimensions to consider + hidden_dim: ${model.feature_encoder.out_channels} + out_channels: ${dataset.parameters.num_classes} + task_level: ${define_task_level:${dataset.parameters.task_level},${dataset.split_params.learning_setting}} # Handles the edge case of node-inductive task + pooling_type: sum + +# compile model for faster training with pytorch 2.0 +compile: false diff --git a/configs/model/graph/gread_fisher.yaml b/configs/model/graph/gread_fisher.yaml new file mode 100644 index 000000000..1cd3f040d --- /dev/null +++ b/configs/model/graph/gread_fisher.yaml @@ -0,0 +1,47 @@ +_target_: topobench.model.TBModel + +model_name: gread_fisher +model_domain: graph + +feature_encoder: + _target_: topobench.nn.encoders.${model.feature_encoder.encoder_name} + encoder_name: AllCellFeatureEncoder + in_channels: ${infer_in_channels:${dataset},${oc.select:transforms,null}} + out_channels: 64 + proj_dropout: 0.0 + +backbone: + _target_: topobench.nn.backbones.GREAD + in_channels: ${model.feature_encoder.out_channels} + hidden_channels: ${model.feature_encoder.out_channels} + reaction_term: fisher + time: 3.0 + step_size: 1.0 + beta_diag: false + add_source: false + data_norm: rw + self_loop_weight: 1.0 + # Terminal ReLU pinned explicitly for the submitted benchmark runs + # (the constructor default is false, matching the reference CLI default). + xn_activation: true + input_dropout: 0.0 + dropout: 0.0 + +backbone_wrapper: + _target_: topobench.nn.wrappers.GNNWrapper + _partial_: true + wrapper_name: GNNWrapper + out_channels: ${model.feature_encoder.out_channels} + num_cell_dimensions: ${infer_num_cell_dimensions:${oc.select:model.feature_encoder.selected_dimensions,null},${model.feature_encoder.in_channels}} + +readout: + _target_: topobench.nn.readouts.${model.readout.readout_name} + readout_name: NoReadOut # Use in case readout is not needed Options: PropagateSignalDown + num_cell_dimensions: ${infer_num_cell_dimensions:${oc.select:model.feature_encoder.selected_dimensions,null},${model.feature_encoder.in_channels}} # The highest order of cell dimensions to consider + hidden_dim: ${model.feature_encoder.out_channels} + out_channels: ${dataset.parameters.num_classes} + task_level: ${define_task_level:${dataset.parameters.task_level},${dataset.split_params.learning_setting}} # Handles the edge case of node-inductive task + pooling_type: sum + +# compile model for faster training with pytorch 2.0 +compile: false diff --git a/configs/model/graph/gread_source_term.yaml b/configs/model/graph/gread_source_term.yaml new file mode 100644 index 000000000..5afb51593 --- /dev/null +++ b/configs/model/graph/gread_source_term.yaml @@ -0,0 +1,47 @@ +_target_: topobench.model.TBModel + +model_name: gread_source_term +model_domain: graph + +feature_encoder: + _target_: topobench.nn.encoders.${model.feature_encoder.encoder_name} + encoder_name: AllCellFeatureEncoder + in_channels: ${infer_in_channels:${dataset},${oc.select:transforms,null}} + out_channels: 64 + proj_dropout: 0.0 + +backbone: + _target_: topobench.nn.backbones.GREAD + in_channels: ${model.feature_encoder.out_channels} + hidden_channels: ${model.feature_encoder.out_channels} + reaction_term: st + time: 3.0 + step_size: 1.0 + beta_diag: false + add_source: false + data_norm: rw + self_loop_weight: 1.0 + # Terminal ReLU pinned explicitly for the submitted benchmark runs + # (the constructor default is false, matching the reference CLI default). + xn_activation: true + input_dropout: 0.0 + dropout: 0.0 + +backbone_wrapper: + _target_: topobench.nn.wrappers.GNNWrapper + _partial_: true + wrapper_name: GNNWrapper + out_channels: ${model.feature_encoder.out_channels} + num_cell_dimensions: ${infer_num_cell_dimensions:${oc.select:model.feature_encoder.selected_dimensions,null},${model.feature_encoder.in_channels}} + +readout: + _target_: topobench.nn.readouts.${model.readout.readout_name} + readout_name: NoReadOut # Use in case readout is not needed Options: PropagateSignalDown + num_cell_dimensions: ${infer_num_cell_dimensions:${oc.select:model.feature_encoder.selected_dimensions,null},${model.feature_encoder.in_channels}} # The highest order of cell dimensions to consider + hidden_dim: ${model.feature_encoder.out_channels} + out_channels: ${dataset.parameters.num_classes} + task_level: ${define_task_level:${dataset.parameters.task_level},${dataset.split_params.learning_setting}} # Handles the edge case of node-inductive task + pooling_type: sum + +# compile model for faster training with pytorch 2.0 +compile: false diff --git a/configs/model/graph/gread_zeldovich.yaml b/configs/model/graph/gread_zeldovich.yaml new file mode 100644 index 000000000..bdb2f4f7f --- /dev/null +++ b/configs/model/graph/gread_zeldovich.yaml @@ -0,0 +1,47 @@ +_target_: topobench.model.TBModel + +model_name: gread_zeldovich +model_domain: graph + +feature_encoder: + _target_: topobench.nn.encoders.${model.feature_encoder.encoder_name} + encoder_name: AllCellFeatureEncoder + in_channels: ${infer_in_channels:${dataset},${oc.select:transforms,null}} + out_channels: 64 + proj_dropout: 0.0 + +backbone: + _target_: topobench.nn.backbones.GREAD + in_channels: ${model.feature_encoder.out_channels} + hidden_channels: ${model.feature_encoder.out_channels} + reaction_term: zeldovich + time: 3.0 + step_size: 1.0 + beta_diag: false + add_source: false + data_norm: rw + self_loop_weight: 1.0 + # Terminal ReLU pinned explicitly for the submitted benchmark runs + # (the constructor default is false, matching the reference CLI default). + xn_activation: true + input_dropout: 0.0 + dropout: 0.0 + +backbone_wrapper: + _target_: topobench.nn.wrappers.GNNWrapper + _partial_: true + wrapper_name: GNNWrapper + out_channels: ${model.feature_encoder.out_channels} + num_cell_dimensions: ${infer_num_cell_dimensions:${oc.select:model.feature_encoder.selected_dimensions,null},${model.feature_encoder.in_channels}} + +readout: + _target_: topobench.nn.readouts.${model.readout.readout_name} + readout_name: NoReadOut # Use in case readout is not needed Options: PropagateSignalDown + num_cell_dimensions: ${infer_num_cell_dimensions:${oc.select:model.feature_encoder.selected_dimensions,null},${model.feature_encoder.in_channels}} # The highest order of cell dimensions to consider + hidden_dim: ${model.feature_encoder.out_channels} + out_channels: ${dataset.parameters.num_classes} + task_level: ${define_task_level:${dataset.parameters.task_level},${dataset.split_params.learning_setting}} # Handles the edge case of node-inductive task + pooling_type: sum + +# compile model for faster training with pytorch 2.0 +compile: false diff --git a/test/nn/backbones/graph/test_gread.py b/test/nn/backbones/graph/test_gread.py new file mode 100644 index 000000000..cfc4ae8cd --- /dev/null +++ b/test/nn/backbones/graph/test_gread.py @@ -0,0 +1,564 @@ +"""Unit tests for the GREAD backbone.""" + +import pytest +import torch +import torch_geometric + +from topobench.nn.backbones.graph import GREAD +from topobench.nn.wrappers.graph import GNNWrapper + + +class TestGREAD: + """Unit tests for the GREAD backbone.""" + + def setup_method(self): + """Set up test fixtures.""" + torch.manual_seed(0) + self.num_nodes = 8 + self.in_channels = 12 + self.hidden_channels = 16 + self.x = torch.randn(self.num_nodes, self.in_channels) + self.edge_index = torch.randint( + 0, self.num_nodes, (2, self.num_nodes * 2) + ) + + @pytest.mark.parametrize( + "reaction_term", + [ + "bspm", + "fisher", + "allen-cahn", + "zeldovich", + "st", + "fb", + "fb3", + "none", + ], + ) + def test_forward_reaction_terms(self, reaction_term): + """Test the forward pass for every reaction term. + + Parameters + ---------- + reaction_term : str + Reaction term to test. + """ + model = GREAD( + self.in_channels, + self.hidden_channels, + reaction_term=reaction_term, + ) + out = model(self.x, self.edge_index) + assert out.shape == (self.num_nodes, self.hidden_channels) + assert torch.isfinite(out).all() + + def test_invalid_reaction_term(self): + """Test that an invalid reaction term raises a ValueError.""" + with pytest.raises(ValueError): + GREAD(self.in_channels, self.hidden_channels, reaction_term="foo") + + def test_invalid_data_norm(self): + """Test that an invalid data normalization raises a ValueError.""" + with pytest.raises(ValueError): + GREAD(self.in_channels, self.hidden_channels, data_norm="foo") + + def test_gcn_norm(self): + """Test the forward pass with symmetric (GCN) normalization.""" + model = GREAD(self.in_channels, self.hidden_channels, data_norm="gcn") + out = model(self.x, self.edge_index) + assert out.shape == (self.num_nodes, self.hidden_channels) + + def test_beta_diag_and_source(self): + """Test the diagonal beta parameterization and the source term.""" + model = GREAD( + self.in_channels, + self.hidden_channels, + beta_diag=True, + add_source=True, + ) + assert model.b_w.shape == (self.hidden_channels,) + out = model(self.x, self.edge_index) + out.sum().backward() + assert model.b_w.grad is not None + assert model.source_train.grad is not None + + def test_gradient_flow(self): + """Test that gradients flow through alpha, beta and the encoder.""" + model = GREAD(self.in_channels, self.hidden_channels) + out = model(self.x, self.edge_index) + out.sum().backward() + assert model.alpha_train.grad is not None + assert model.beta_train.grad is not None + assert model.encoder.weight.grad is not None + + def test_euler_step_pure_diffusion(self): + """Test one Euler step of pure diffusion against a manual computation. + + With ``reaction_term="none"``, one Euler step computes + ``H(t + tau) = H(t) + tau * sigmoid(alpha) * (A - I) H(t)``. + """ + model = GREAD( + self.in_channels, + self.hidden_channels, + reaction_term="none", + time=1.0, + step_size=1.0, + xn_activation=True, + ) + model.eval() + edge_index, edge_weight = model.normalize_adjacency( + self.edge_index, None, self.num_nodes + ) + with torch.no_grad(): + h0 = model.encoder(self.x) + ah0 = model.spmm(edge_index, edge_weight, self.num_nodes, h0) + alpha = torch.sigmoid(model.alpha_train) + expected = torch.relu(h0 + 1.0 * alpha * (ah0 - h0)) + out = model(self.x, self.edge_index) + assert torch.allclose(out, expected, atol=1e-6) + + def test_rw_normalization_column_stochastic(self): + """Test that the rw-normalized adjacency is column-normalized. + + Matches ``get_rw_adj(..., norm_dim=1)`` of the reference + implementation, where the degree is computed over the target + (column) index. + """ + model = GREAD(self.in_channels, self.hidden_channels) + edge_index, edge_weight = model.normalize_adjacency( + self.edge_index, None, self.num_nodes + ) + col_sums = torch.zeros(self.num_nodes) + col_sums.scatter_add_(0, edge_index[1], edge_weight) + assert torch.allclose(col_sums, torch.ones(self.num_nodes), atol=1e-6) + + def test_beta_diag_weights_reaction_per_channel(self): + """Test the ``beta_diag`` update against a manual Euler step. + + With ``beta_diag=True`` the reaction is scaled by the trainable + diagonal :math:`\\beta_W` — used directly, not through a sigmoid — + instead of the scalar gate :math:`\\beta = \\sigma(\\beta_{train})`, + matching the "(VC)" variants of the paper. + """ + model = GREAD( + self.in_channels, + self.hidden_channels, + reaction_term="fisher", + time=1.0, + step_size=1.0, + beta_diag=True, + xn_activation=False, + ) + model.eval() + edge_index, edge_weight = model.normalize_adjacency( + self.edge_index, None, self.num_nodes + ) + with torch.no_grad(): + h0 = model.encoder(self.x) + ax = model.spmm(edge_index, edge_weight, self.num_nodes, h0) + diffusion = ax - h0 + reaction = -(h0 - 1.0) * h0 + alpha = torch.sigmoid(model.alpha_train) + expected = h0 + (alpha * diffusion + reaction * model.b_w) + out = model(self.x, self.edge_index) + assert torch.allclose(out, expected, atol=1e-6) + + # The per-channel gate must not collapse to the scalar beta path. + scalar_beta = torch.sigmoid(model.beta_train) + scalar_expected = h0 + (alpha * diffusion + scalar_beta * reaction) + assert not torch.allclose(out, scalar_expected, atol=1e-4) + + def test_self_loop_weight(self): + """Test that ``self_loop_weight`` enters the normalized adjacency. + + On the single-edge graph ``0 -> 1`` with self-loop fill value + :math:`w`, the column-stochastic normalization :math:`A D^{-1}` + (degrees over the target index) gives :math:`A_{0,0} = 1`, + :math:`A_{0,1} = 1 / (1 + w)` and :math:`A_{1,1} = w / (1 + w)`. + """ + edge_index = torch.tensor([[0], [1]]) + for weight in (0.5, 2.0): + model = GREAD( + self.in_channels, + self.hidden_channels, + self_loop_weight=weight, + ) + out_index, out_weight = model.normalize_adjacency( + edge_index, None, 2 + ) + dense = torch.zeros(2, 2) + dense[out_index[0], out_index[1]] = out_weight + expected = torch.tensor( + [ + [1.0, 1.0 / (1.0 + weight)], + [0.0, weight / (1.0 + weight)], + ] + ) + assert torch.allclose(dense, expected, atol=1e-6) + + # A non-positive weight must leave the edge set untouched. + model = GREAD( + self.in_channels, self.hidden_channels, self_loop_weight=0.0 + ) + out_index, out_weight = model.normalize_adjacency(edge_index, None, 2) + assert out_index.shape == (2, 1) + assert torch.allclose(out_weight, torch.tensor([1.0]), atol=1e-6) + + def test_dropout_is_training_only(self): + """Test that dropout perturbs training passes but not eval passes.""" + model = GREAD( + self.in_channels, + self.hidden_channels, + input_dropout=0.9, + dropout=0.9, + ) + model.train() + torch.manual_seed(1) + first = model(self.x, self.edge_index) + torch.manual_seed(2) + second = model(self.x, self.edge_index) + assert not torch.allclose(first, second) + + model.eval() + with torch.no_grad(): + repeatable = model(self.x, self.edge_index) + assert torch.allclose( + repeatable, model(self.x, self.edge_index), atol=1e-6 + ) + # In eval mode the result must not depend on the dropout rates. + model.input_dropout = 0.0 + model.dropout = 0.0 + assert torch.allclose( + repeatable, model(self.x, self.edge_index), atol=1e-6 + ) + + def test_euler_step_schedule(self): + """Test the Euler step schedule for divisible and partial horizons.""" + model = GREAD( + self.in_channels, self.hidden_channels, time=3.0, step_size=1.0 + ) + assert model.step_sizes == [1.0, 1.0, 1.0] + model = GREAD( + self.in_channels, self.hidden_channels, time=0.5, step_size=1.0 + ) + assert model.step_sizes == [0.5] + model = GREAD( + self.in_channels, self.hidden_channels, time=2.5, step_size=1.0 + ) + assert model.step_sizes == [1.0, 1.0, 0.5] + + def test_tiny_horizon_step_schedule(self): + """Test that a tiny positive horizon still yields one step.""" + model = GREAD( + self.in_channels, self.hidden_channels, time=1e-12, step_size=1.0 + ) + assert model.step_sizes == [1e-12] + assert sum(model.step_sizes) == 1e-12 + + def test_near_multiple_horizon_step_schedule(self): + """Test that the step sizes sum exactly to a near-multiple horizon.""" + time = 1.0000000005 + model = GREAD( + self.in_channels, self.hidden_channels, time=time, step_size=1.0 + ) + assert len(model.step_sizes) == 2 + assert sum(model.step_sizes) == time + + def test_default_xn_activation_disabled(self): + """Test that the terminal ReLU is disabled by default. + + Matches the ``XN_activation`` CLI default of the reference + implementation. + """ + model = GREAD(self.in_channels, self.hidden_channels) + assert model.xn_activation is False + model.eval() + with torch.no_grad(): + out = model(self.x, self.edge_index) + assert (out < 0).any() + + @pytest.mark.parametrize( + "time, step_size", + [(0.0, 1.0), (-1.0, 1.0), (1.0, 0.0), (1.0, -0.5)], + ) + def test_nonpositive_time_or_step_size_rejected(self, time, step_size): + """Test that non-positive time or step size raises a ValueError. + + Parameters + ---------- + time : float + Terminal integration time. + step_size : float + Euler step size. + """ + with pytest.raises(ValueError): + GREAD( + self.in_channels, + self.hidden_channels, + time=time, + step_size=step_size, + ) + + def test_partial_final_step_numerical(self): + """Test integration with a partial final step against manual Euler. + + With ``reaction_term="none"`` and ``time=2.5``, ``step_size=1.0``, + the model must take steps of 1.0, 1.0 and 0.5: + ``H_{k+1} = H_k + tau_k * sigmoid(alpha) * (A - I) H_k``. + """ + model = GREAD( + self.in_channels, + self.hidden_channels, + reaction_term="none", + time=2.5, + step_size=1.0, + xn_activation=True, + ) + model.eval() + edge_index, edge_weight = model.normalize_adjacency( + self.edge_index, None, self.num_nodes + ) + with torch.no_grad(): + h = model.encoder(self.x) + alpha = torch.sigmoid(model.alpha_train) + for tau in (1.0, 1.0, 0.5): + ah = model.spmm(edge_index, edge_weight, self.num_nodes, h) + h = h + tau * alpha * (ah - h) + expected = torch.relu(h) + out = model(self.x, self.edge_index) + assert torch.allclose(out, expected, atol=1e-6) + + def test_short_horizon_numerical(self): + """Test that ``time < step_size`` takes a single step of size time. + + With ``time=0.5``, ``step_size=1.0`` a single Euler step of size + 0.5 must be taken: + ``H(0.5) = H(0) + 0.5 * sigmoid(alpha) * (A - I) H(0)``. + """ + model = GREAD( + self.in_channels, + self.hidden_channels, + reaction_term="none", + time=0.5, + step_size=1.0, + xn_activation=True, + ) + model.eval() + edge_index, edge_weight = model.normalize_adjacency( + self.edge_index, None, self.num_nodes + ) + with torch.no_grad(): + h0 = model.encoder(self.x) + ah0 = model.spmm(edge_index, edge_weight, self.num_nodes, h0) + alpha = torch.sigmoid(model.alpha_train) + expected = torch.relu(h0 + 0.5 * alpha * (ah0 - h0)) + out = model(self.x, self.edge_index) + assert torch.allclose(out, expected, atol=1e-6) + + def test_zero_initialized_source_contributes_nothing(self): + """Test that a zero-initialized source term contributes zero. + + With ``add_source=True`` and the source coefficient at its zero + initialization, the output must equal that of an identical model + without the source term. + """ + torch.manual_seed(1) + with_source = GREAD( + self.in_channels, + self.hidden_channels, + reaction_term="none", + add_source=True, + ) + torch.manual_seed(1) + without_source = GREAD( + self.in_channels, + self.hidden_channels, + reaction_term="none", + add_source=False, + ) + with_source.eval() + without_source.eval() + assert with_source.source_train.item() == 0.0 + with torch.no_grad(): + out_with = with_source(self.x, self.edge_index) + out_without = without_source(self.x, self.edge_index) + assert torch.allclose(out_with, out_without, atol=1e-7) + + def test_source_term_no_sigmoid(self): + """Test the source-term contribution numerically without sigmoid. + + With a source coefficient ``s``, each Euler step must add + ``tau * s * H(0)`` (not ``tau * sigmoid(s) * H(0)``). + """ + model = GREAD( + self.in_channels, + self.hidden_channels, + reaction_term="none", + time=1.0, + step_size=1.0, + add_source=True, + xn_activation=False, + ) + model.eval() + with torch.no_grad(): + model.source_train.fill_(0.3) + edge_index, edge_weight = model.normalize_adjacency( + self.edge_index, None, self.num_nodes + ) + with torch.no_grad(): + h0 = model.encoder(self.x) + ah0 = model.spmm(edge_index, edge_weight, self.num_nodes, h0) + alpha = torch.sigmoid(model.alpha_train) + expected = h0 + 1.0 * (alpha * (ah0 - h0) + 0.3 * h0) + out = model(self.x, self.edge_index) + assert torch.allclose(out, expected, atol=1e-6) + + def test_xn_activation_toggle(self): + """Test the terminal ReLU toggle numerically. + + With ``xn_activation=False`` the raw terminal state is returned; + with ``xn_activation=True`` its ReLU is returned. + """ + torch.manual_seed(2) + relu_model = GREAD( + self.in_channels, + self.hidden_channels, + xn_activation=True, + ) + torch.manual_seed(2) + raw_model = GREAD( + self.in_channels, + self.hidden_channels, + xn_activation=False, + ) + relu_model.eval() + raw_model.eval() + with torch.no_grad(): + out_relu = relu_model(self.x, self.edge_index) + out_raw = raw_model(self.x, self.edge_index) + assert (out_raw < 0).any() + assert torch.allclose(out_relu, torch.relu(out_raw), atol=1e-7) + + @pytest.mark.parametrize( + "reaction_term, reaction_fn", + [ + ("fisher", lambda h, h0: -(h - 1) * h), + ("allen-cahn", lambda h, h0: -(h**2 - 1) * h), + ("zeldovich", lambda h, h0: -(h**2 - h) * h), + ("st", lambda h, h0: h0), + ], + ) + def test_pointwise_reactions_numerical(self, reaction_term, reaction_fn): + """Test pointwise reaction terms against manual Euler computations. + + One Euler step computes ``H1 = H0 + tau * (sigmoid(alpha) * + (A - I) H0 + sigmoid(beta) * r(H0))``. + + Parameters + ---------- + reaction_term : str + Reaction term to test. + reaction_fn : callable + Manual implementation of the reaction term. + """ + model = GREAD( + self.in_channels, + self.hidden_channels, + reaction_term=reaction_term, + time=1.0, + step_size=1.0, + xn_activation=False, + ) + model.eval() + edge_index, edge_weight = model.normalize_adjacency( + self.edge_index, None, self.num_nodes + ) + with torch.no_grad(): + h0 = model.encoder(self.x) + ah0 = model.spmm(edge_index, edge_weight, self.num_nodes, h0) + alpha = torch.sigmoid(model.alpha_train) + beta = torch.sigmoid(model.beta_train) + expected = h0 + 1.0 * ( + alpha * (ah0 - h0) + beta * reaction_fn(h0, h0) + ) + out = model(self.x, self.edge_index) + assert torch.allclose(out, expected, atol=1e-6) + + @pytest.mark.parametrize("reaction_term", ["bspm", "fb", "fb3"]) + def test_graph_reactions_numerical(self, reaction_term): + """Test graph-coupled reaction terms against manual computations. + + Covers ``bspm`` (``-A(A - I)H``), ``fb`` (``H + AH``) and + ``fb3`` (``fb`` plus an extra ``H`` inside the beta gate). + + Parameters + ---------- + reaction_term : str + Reaction term to test. + """ + model = GREAD( + self.in_channels, + self.hidden_channels, + reaction_term=reaction_term, + time=1.0, + step_size=1.0, + xn_activation=False, + ) + model.eval() + edge_index, edge_weight = model.normalize_adjacency( + self.edge_index, None, self.num_nodes + ) + with torch.no_grad(): + h0 = model.encoder(self.x) + ah0 = model.spmm(edge_index, edge_weight, self.num_nodes, h0) + diffusion = ah0 - h0 + alpha = torch.sigmoid(model.alpha_train) + beta = torch.sigmoid(model.beta_train) + if reaction_term == "bspm": + reaction = -model.spmm( + edge_index, edge_weight, self.num_nodes, diffusion + ) + f = alpha * diffusion + beta * reaction + else: + reaction = h0 + model.spmm( + edge_index, edge_weight, self.num_nodes, h0 + ) + if reaction_term == "fb3": + f = alpha * diffusion + beta * (reaction + h0) + else: + f = alpha * diffusion + beta * reaction + expected = h0 + 1.0 * f + out = model(self.x, self.edge_index) + assert torch.allclose(out, expected, atol=1e-6) + + def test_edge_weight_passthrough(self): + """Test the forward pass with explicit edge weights.""" + model = GREAD(self.in_channels, self.hidden_channels) + edge_weight = torch.rand(self.edge_index.size(1)) + out = model(self.x, self.edge_index, edge_weight=edge_weight) + assert out.shape == (self.num_nodes, self.hidden_channels) + + def test_with_gnn_wrapper(self, random_graph_input): + """Test GREAD within the GNNWrapper. + + Parameters + ---------- + random_graph_input : tuple + Fixture with random graph inputs. + """ + x, _, _, edges_1, _ = random_graph_input + batch = torch_geometric.data.Data( + x_0=x, + y=x, + x=x, + edge_index=edges_1, + batch_0=torch.zeros(x.shape[0], dtype=torch.long), + ) + model = GREAD(x.shape[1], x.shape[1]) + wrapper = GNNWrapper( + model, + out_channels=x.shape[1], + num_cell_dimensions=1, + ) + model_out = wrapper(batch) + assert model_out["x_0"].shape == x.shape diff --git a/test/pipeline/test_pipeline.py b/test/pipeline/test_pipeline.py index a61165ae9..1b0966cd3 100644 --- a/test/pipeline/test_pipeline.py +++ b/test/pipeline/test_pipeline.py @@ -7,7 +7,13 @@ DATASET = "graph/MUTAG" # ADD YOUR DATASET HERE -MODELS = ["graph/gcn", "cell/topotune", "simplicial/topotune"] # ADD ONE OR SEVERAL MODELS +MODELS = [ + "graph/gread", + "graph/gread_fisher", + "graph/gread_allen_cahn", + "graph/gread_zeldovich", + "graph/gread_source_term", +] # ADD ONE OR SEVERAL MODELS class TestPipeline: diff --git a/topobench/nn/backbones/graph/gread.py b/topobench/nn/backbones/graph/gread.py new file mode 100644 index 000000000..8f2c12a26 --- /dev/null +++ b/topobench/nn/backbones/graph/gread.py @@ -0,0 +1,427 @@ +"""GREAD: Graph Neural Reaction-Diffusion Networks. + +Implementation of the paper "GREAD: Graph Neural Reaction-Diffusion +Networks" (Choi et al., ICML 2023, https://arxiv.org/abs/2211.14208), +adapted from the official reference implementation +https://github.com/jeongwhanchoi/GREAD. +""" + +import torch +import torch.nn.functional as F +from torch import nn +from torch_geometric.utils import ( + add_remaining_self_loops, + scatter, +) + + +class GREAD(nn.Module): + r"""GREAD: Graph Neural Reaction-Diffusion Network backbone. + + GREAD (Choi et al., ICML 2023, https://arxiv.org/abs/2211.14208) + evolves node representations :math:`\mathbf{H}(t)` with the + reaction-diffusion equation (Section 3.1, Eqs. (6)-(8) of the paper): + + .. math:: + \mathbf{f}(\mathbf{H}(t)) := \frac{d\mathbf{H}(t)}{dt} = + -\alpha \, \mathbf{L}\mathbf{H}(t) + + \beta \, \mathbf{r}(\mathbf{H}(t)), + + with initial value :math:`\mathbf{H}(0) = \mathbf{e}(\mathbf{X})` + (Eq. (6)), where :math:`\mathbf{e}` is an encoding layer, + :math:`\mathbf{L} = \mathbf{I} - \mathbf{A}` the Laplacian of a + normalized adjacency matrix :math:`\mathbf{A}`, and + :math:`\mathbf{r}` one of the reaction terms of Eq. (10) of the + paper: + + - ``"bspm"``: blurring-sharpening + :math:`(\mathbf{A} - \mathbf{A}^2)\mathbf{H}(t)` + (GREAD-BS, Eq. (14)). + - ``"fisher"``: Fisher + :math:`\mathbf{H}(t)\odot(1 - \mathbf{H}(t))` (GREAD-F). + - ``"allen-cahn"``: Allen-Cahn + :math:`\mathbf{H}(t)\odot(1 - \mathbf{H}(t)^{\circ 2})` + (GREAD-AC). + - ``"zeldovich"``: Zeldovich + :math:`\mathbf{H}(t)\odot(\mathbf{H}(t) - + \mathbf{H}(t)^{\circ 2})` (GREAD-Z). + - ``"st"``: source term :math:`\mathbf{H}(0)` (GREAD-ST). + - ``"fb"``: filter bank (high-pass) reaction (GREAD-FB). Eq. (10) + of the paper writes it as :math:`\mathbf{L}\mathbf{H}(t)`, while + the reference implementation computes + :math:`(\mathbf{I} + \mathbf{A})\mathbf{H}(t)`; this + implementation follows the reference code. + - ``"fb3"``: filter bank with identity channel (GREAD-FB*), the + ``"fb"`` reaction plus :math:`\mathbf{H}(t)`, again following the + reference code. + - ``"none"``: no reaction, i.e. pure diffusion. + + The ODE is integrated with the explicit Euler method, the base + solver in which Eq. (5) of the paper is written and one of the two + fixed-grid solvers (``euler`` / ``rk4``) selected per dataset by + the reference implementation + (https://github.com/jeongwhanchoi/GREAD, + ``src/gread_params.py``). Integration reaches the terminal time + ``time`` up to floating-point rounding: full steps of size + ``step_size`` are taken and, when ``time`` is not an integer + multiple of ``step_size``, a final shorter step covers the remaining + duration. + + The scalar gates are parameterized as + :math:`\alpha = \sigma(\alpha_{\text{train}})` and + :math:`\beta = \sigma(\beta_{\text{train}})` following + ``ODEFuncGread.forward`` of the reference code; with + ``beta_diag=True`` the reaction is instead weighted by a trainable + diagonal matrix :math:`\beta_W` (the "(VC)" variants in the paper). + + Parameters + ---------- + in_channels : int + Number of input features. + hidden_channels : int + Number of hidden units. + reaction_term : str, optional + Reaction term, one of ``"bspm"``, ``"fisher"``, ``"allen-cahn"``, + ``"zeldovich"``, ``"st"``, ``"fb"``, ``"fb3"``, ``"none"`` + (default: ``"bspm"``). + time : float, optional + Terminal integration time :math:`T`; must be positive + (default: 3.0). + step_size : float, optional + Step size :math:`\tau` of the explicit Euler solver; must be + positive (default: 1.0). + beta_diag : bool, optional + If True, use a trainable diagonal matrix ("vector channel") + instead of the scalar :math:`\beta` gate (default: False). + add_source : bool, optional + If True, add a learnable source term + :math:`s\,\mathbf{H}(0)` to the dynamics, with :math:`s` + initialized at zero so the source contributes nothing until + trained (default: False). + data_norm : str, optional + Adjacency normalization: ``"rw"`` for the column-stochastic + random-walk normalization :math:`\mathbf{A}\mathbf{D}^{-1}` + (degrees taken over the target index, matching + ``get_rw_adj(..., norm_dim=1)`` of the reference + implementation) or ``"gcn"`` for the symmetric normalization + :math:`\mathbf{D}^{-1/2}\mathbf{A}\mathbf{D}^{-1/2}` + (default: ``"rw"``). + self_loop_weight : float, optional + Weight of the self loops added to the adjacency matrix before + normalization (default: 1.0). + xn_activation : bool, optional + If True, apply a terminal ReLU to the integrated + representations :math:`\mathbf{H}(T)` before the output + dropout (``XN_activation`` in the reference implementation, + whose CLI default is ``False``; it is enabled only for selected + dataset configurations). Default: ``False``. + input_dropout : float, optional + Dropout rate applied to the input features (default: 0.0). + dropout : float, optional + Dropout rate applied to the final representation + (default: 0.0). + **kwargs + Additional arguments (ignored). + """ + + _REACTION_TERMS = ( + "bspm", + "fisher", + "allen-cahn", + "zeldovich", + "st", + "fb", + "fb3", + "none", + ) + + def __init__( + self, + in_channels, + hidden_channels, + reaction_term="bspm", + time=3.0, + step_size=1.0, + beta_diag=False, + add_source=False, + data_norm="rw", + self_loop_weight=1.0, + xn_activation=False, + input_dropout=0.0, + dropout=0.0, + **kwargs, + ): + super().__init__() + if reaction_term not in self._REACTION_TERMS: + raise ValueError( + f"Unknown reaction term '{reaction_term}'. " + f"Expected one of {self._REACTION_TERMS}." + ) + if data_norm not in ("rw", "gcn"): + raise ValueError( + f"Unknown data normalization '{data_norm}'. " + "Expected 'rw' or 'gcn'." + ) + if time <= 0: + raise ValueError( + f"Terminal time must be positive, got time={time}." + ) + if step_size <= 0: + raise ValueError( + f"Step size must be positive, got step_size={step_size}." + ) + self.in_channels = in_channels + self.hidden_channels = hidden_channels + self.out_channels = hidden_channels + self.reaction_term = reaction_term + self.time = time + self.step_size = step_size + self.step_sizes = self._euler_step_sizes(time, step_size) + self.beta_diag = beta_diag + self.add_source = add_source + self.data_norm = data_norm + self.self_loop_weight = self_loop_weight + self.xn_activation = xn_activation + self.input_dropout = input_dropout + self.dropout = dropout + + # Encoding layer E (m1 in the reference implementation). + self.encoder = nn.Linear(in_channels, hidden_channels) + + # Scalar gates alpha and beta (sigmoid-activated, initialized at + # zero as in ODEFunc.__init__ of the reference implementation). + self.alpha_train = nn.Parameter(torch.tensor(0.0)) + self.beta_train = nn.Parameter(torch.tensor(0.0)) + if beta_diag: + self.b_w = nn.Parameter(torch.empty(hidden_channels)) + if add_source: + self.source_train = nn.Parameter(torch.tensor(0.0)) + + self.reset_parameters() + + @staticmethod + def _euler_step_sizes(time, step_size): + """Compute the explicit Euler step schedule. + + Full steps of ``step_size`` are taken while they fit within the + horizon; a final shorter step covers any remaining duration so + that the step sizes sum to ``time`` up to floating-point + rounding. At least one step is always taken for a positive + horizon. + + Parameters + ---------- + time : float + Terminal integration time. + step_size : float + Nominal step size. + + Returns + ------- + list of float + The sizes of the successive Euler steps; they sum to + ``time`` up to floating-point rounding. + """ + n_full = int(time / step_size) + remainder = time - n_full * step_size + sizes = [step_size] * n_full + if remainder > 0: + sizes.append(remainder) + return sizes + + def reset_parameters(self): + """Reset the learnable parameters.""" + self.encoder.reset_parameters() + with torch.no_grad(): + self.alpha_train.zero_() + self.beta_train.zero_() + if self.beta_diag: + nn.init.uniform_(self.b_w, a=-1.0, b=1.0) + if self.add_source: + with torch.no_grad(): + self.source_train.zero_() + + def normalize_adjacency(self, edge_index, edge_weight, num_nodes): + r"""Normalize the adjacency matrix. + + Adds weighted self loops and normalizes the adjacency matrix, + following ``get_rw_adj`` (``data_norm="rw"``) and + ``gcn_norm_fill_val`` (``data_norm="gcn"``) of the reference + implementation (``src/utils.py``). + + Parameters + ---------- + edge_index : torch.Tensor + Edge index of shape ``[2, num_edges]``. + edge_weight : torch.Tensor or None + Optional edge weights of shape ``[num_edges]``. + num_nodes : int + Number of nodes. + + Returns + ------- + tuple of (torch.Tensor, torch.Tensor) + Normalized edge index and edge weights. + """ + if edge_weight is None: + edge_weight = torch.ones( + edge_index.size(1), + dtype=torch.float, + device=edge_index.device, + ) + if self.self_loop_weight > 0: + edge_index, edge_weight = add_remaining_self_loops( + edge_index, + edge_weight, + fill_value=self.self_loop_weight, + num_nodes=num_nodes, + ) + row, col = edge_index[0], edge_index[1] + if self.data_norm == "rw": + # A D^{-1}: degrees over the target (column) index, i.e. a + # column-stochastic operator (norm_dim=1 in the reference + # implementation). + deg = scatter( + edge_weight, col, dim=0, dim_size=num_nodes, reduce="sum" + ) + deg_inv = deg.pow(-1) + deg_inv.masked_fill_(deg_inv == float("inf"), 0) + edge_weight = deg_inv[col] * edge_weight + else: + deg = scatter( + edge_weight, col, dim=0, dim_size=num_nodes, reduce="sum" + ) + deg_inv_sqrt = deg.pow(-0.5) + deg_inv_sqrt.masked_fill_(deg_inv_sqrt == float("inf"), 0) + edge_weight = deg_inv_sqrt[row] * edge_weight * deg_inv_sqrt[col] + return edge_index, edge_weight + + @staticmethod + def spmm(edge_index, edge_weight, num_nodes, x): + r"""Multiply the sparse adjacency matrix with a dense matrix. + + Computes :math:`\mathbf{A}\mathbf{X}`, the equivalent of + ``ODEFuncGread.sparse_multiply`` of the reference implementation. + + Parameters + ---------- + edge_index : torch.Tensor + Edge index of shape ``[2, num_edges]``. + edge_weight : torch.Tensor + Edge weights of shape ``[num_edges]``. + num_nodes : int + Number of nodes. + x : torch.Tensor + Dense feature matrix of shape ``[num_nodes, channels]``. + + Returns + ------- + torch.Tensor + The product :math:`\mathbf{A}\mathbf{X}`. + """ + row, col = edge_index[0], edge_index[1] + return scatter( + edge_weight.unsqueeze(-1) * x[col], + row, + dim=0, + dim_size=num_nodes, + reduce="sum", + ) + + def reaction(self, x, diffusion, x0, edge_index, edge_weight, num_nodes): + r"""Compute the reaction term :math:`\mathbf{r}(\mathbf{H}, \mathbf{A})`. + + Implements the reaction terms of Eq. (10) of the paper, matching + ``ODEFuncGread.forward`` of the reference implementation. + + Parameters + ---------- + x : torch.Tensor + Current node representations :math:`\mathbf{H}(t)`. + diffusion : torch.Tensor + Diffusion term :math:`(\mathbf{A} - \mathbf{I})\mathbf{H}(t)`. + x0 : torch.Tensor + Initial node representations :math:`\mathbf{H}(0)`. + edge_index : torch.Tensor + Edge index of shape ``[2, num_edges]``. + edge_weight : torch.Tensor + Normalized edge weights of shape ``[num_edges]``. + num_nodes : int + Number of nodes. + + Returns + ------- + torch.Tensor + The reaction term. + """ + if self.reaction_term == "bspm": + return -self.spmm(edge_index, edge_weight, num_nodes, diffusion) + if self.reaction_term == "fisher": + return -(x - 1) * x + if self.reaction_term == "allen-cahn": + return -(x**2 - 1) * x + if self.reaction_term == "zeldovich": + return -(x**2 - x) * x + if self.reaction_term == "st": + return x0 + if self.reaction_term in ("fb", "fb3"): + ax = -self.spmm(edge_index, edge_weight, num_nodes, x) + return x - ax + return torch.zeros_like(x) + + def forward(self, x, edge_index, batch=None, edge_weight=None): + r"""Forward pass. + + Encodes the input features, integrates the reaction-diffusion + equation with the explicit Euler method and returns the terminal + node representations :math:`\mathbf{H}(T)`. + + Parameters + ---------- + x : torch.Tensor + Input node features of shape ``[num_nodes, in_channels]``. + edge_index : torch.Tensor + Edge index of shape ``[2, num_edges]``. + batch : torch.Tensor, optional + Batch assignment vector (unused, present for API + compatibility). + edge_weight : torch.Tensor, optional + Optional edge weights of shape ``[num_edges]``. + + Returns + ------- + torch.Tensor + Node representations of shape + ``[num_nodes, hidden_channels]``. + """ + num_nodes = x.size(0) + edge_index, edge_weight = self.normalize_adjacency( + edge_index, edge_weight, num_nodes + ) + + x = F.dropout(x, self.input_dropout, training=self.training) + x = self.encoder(x) + x0 = x + + alpha = torch.sigmoid(self.alpha_train) + beta = torch.sigmoid(self.beta_train) + + for tau in self.step_sizes: + ax = self.spmm(edge_index, edge_weight, num_nodes, x) + diffusion = ax - x + reaction = self.reaction( + x, diffusion, x0, edge_index, edge_weight, num_nodes + ) + if self.beta_diag: + f = alpha * diffusion + reaction * self.b_w + elif self.reaction_term == "fb3": + f = alpha * diffusion + beta * (reaction + x) + else: + f = alpha * diffusion + beta * reaction + if self.add_source: + f = f + self.source_train * x0 + x = x + tau * f + + if self.xn_activation: + x = F.relu(x) + return F.dropout(x, self.dropout, training=self.training) From 4493d86a8bf1467850743743acee8beec4341ca9 Mon Sep 17 00:00:00 2001 From: Ace Taffy Date: Sun, 2 Aug 2026 05:09:49 +0900 Subject: [PATCH 2/2] Switch GREAD benchmark to beta_diag=true and update the official results Every tuned dataset configuration in the reference implementation (src/gread_params.py) combines reaction_term bspm with beta_diag: True (the paper's "(VC)" per-channel reaction gate); the originally submitted config benchmarked the non-default scalar gate. This commit flips graph/gread.yaml to the reference setting and replaces the committed official grid (72 runs, 12 regimes x 2 tasks x 3 seeds, seeds 42/43/44, produced by the unmodified run_evaluation.ipynb harness path) with the re-run: - community detection mean accuracy: 0.4504 -> 0.4816 (better in all 12 regimes) - triangle counting mean MSE/triangles: 0.9206 -> 0.8773 (better in 10 of 12 regimes) The supplementary analysis notebook is re-executed against the new study (outputs/2026-08-02_gread-vc); its prose is updated where the conclusions changed. Co-Authored-By: Claude Fable 5 --- 2026_tdl_challenge/analysis_gread.ipynb | 194 +- .../results.json | 3026 ++++++++--------- configs/model/graph/gread.yaml | 5 +- 3 files changed, 1616 insertions(+), 1609 deletions(-) rename 2026_tdl_challenge/outputs/{2026-07-25_13-46-02 => 2026-08-02_gread-vc}/results.json (56%) diff --git a/2026_tdl_challenge/analysis_gread.ipynb b/2026_tdl_challenge/analysis_gread.ipynb index 5ba560785..23b0ac358 100644 --- a/2026_tdl_challenge/analysis_gread.ipynb +++ b/2026_tdl_challenge/analysis_gread.ipynb @@ -7,7 +7,8 @@ "source": [ "# GREAD on GraphUniverse — supplementary transfer analysis\n", "\n", - "Model: GREAD (Choi et al., ICML 2023, [arXiv:2211.14208](https://arxiv.org/abs/2211.14208)), config `graph/gread`.\n", + "Model: GREAD (Choi et al., ICML 2023, [arXiv:2211.14208](https://arxiv.org/abs/2211.14208)), config `graph/gread` (`beta_diag: true`, the per-channel reaction gate\n", + "used by every tuned setting of the reference implementation).\n", "\n", "This notebook is **supplementary** to the official `run_evaluation.ipynb`. It\n", "re-trains nothing: it reads the committed `results.json` from the official grid\n", @@ -37,10 +38,10 @@ "id": "cell-02", "metadata": { "execution": { - "iopub.execute_input": "2026-07-26T18:10:52.074950Z", - "iopub.status.busy": "2026-07-26T18:10:52.074874Z", - "iopub.status.idle": "2026-07-26T18:10:52.686136Z", - "shell.execute_reply": "2026-07-26T18:10:52.685739Z" + "iopub.execute_input": "2026-08-01T20:07:36.727803Z", + "iopub.status.busy": "2026-08-01T20:07:36.727715Z", + "iopub.status.idle": "2026-08-01T20:07:37.346509Z", + "shell.execute_reply": "2026-08-01T20:07:37.345700Z" } }, "outputs": [ @@ -50,7 +51,7 @@ "text": [ "864 measurements (72 in-distribution, 792 OOD)\n", "\n", - "triangle-counting MSE/triangles spans 0.0139 to 2215 -> analysed as log10\n" + "triangle-counting MSE/triangles spans 0.01082 to 2164 -> analysed as log10\n" ] }, { @@ -97,14 +98,14 @@ " h_lo__d_lo__pl_lo\n", " 42\n", " True\n", - " 0.314692\n", + " 0.338834\n", " lo\n", " lo\n", " lo\n", " lo\n", " lo\n", " lo\n", - " 0.314692\n", + " 0.338834\n", " \n", " \n", " 1\n", @@ -113,14 +114,14 @@ " h_lo__d_lo__pl_hi\n", " 42\n", " False\n", - " 0.312667\n", + " 0.330354\n", " lo\n", " lo\n", " lo\n", " lo\n", " lo\n", " hi\n", - " 0.312667\n", + " 0.330354\n", " \n", " \n", " 2\n", @@ -129,14 +130,14 @@ " h_lo__d_hi__pl_lo\n", " 42\n", " False\n", - " 0.328711\n", + " 0.371724\n", " lo\n", " lo\n", " lo\n", " lo\n", " hi\n", " lo\n", - " 0.328711\n", + " 0.371724\n", " \n", " \n", " 3\n", @@ -145,14 +146,14 @@ " h_lo__d_hi__pl_hi\n", " 42\n", " False\n", - " 0.326610\n", + " 0.367255\n", " lo\n", " lo\n", " lo\n", " lo\n", " hi\n", " hi\n", - " 0.326610\n", + " 0.367255\n", " \n", " \n", " 4\n", @@ -161,14 +162,14 @@ " h_mid__d_lo__pl_lo\n", " 42\n", " False\n", - " 0.348079\n", + " 0.350371\n", " lo\n", " lo\n", " lo\n", " mid\n", " lo\n", " lo\n", - " 0.348079\n", + " 0.350371\n", " \n", " \n", "\n", @@ -183,18 +184,18 @@ "4 community_detection h_lo__d_lo__pl_lo h_mid__d_lo__pl_lo 42 \n", "\n", " in_distribution score train_homophily train_degree train_powerlaw \\\n", - "0 True 0.314692 lo lo lo \n", - "1 False 0.312667 lo lo lo \n", - "2 False 0.328711 lo lo lo \n", - "3 False 0.326610 lo lo lo \n", - "4 False 0.348079 lo lo lo \n", + "0 True 0.338834 lo lo lo \n", + "1 False 0.330354 lo lo lo \n", + "2 False 0.371724 lo lo lo \n", + "3 False 0.367255 lo lo lo \n", + "4 False 0.350371 lo lo lo \n", "\n", " eval_homophily eval_degree eval_powerlaw value \n", - "0 lo lo lo 0.314692 \n", - "1 lo lo hi 0.312667 \n", - "2 lo hi lo 0.328711 \n", - "3 lo hi hi 0.326610 \n", - "4 mid lo lo 0.348079 " + "0 lo lo lo 0.338834 \n", + "1 lo lo hi 0.330354 \n", + "2 lo hi lo 0.371724 \n", + "3 lo hi hi 0.367255 \n", + "4 mid lo lo 0.350371 " ] }, "execution_count": 1, @@ -210,7 +211,7 @@ "import numpy as np\n", "import pandas as pd\n", "\n", - "STUDY_ID = \"2026-07-25_13-46-02\"\n", + "STUDY_ID = \"2026-08-02_gread-vc\"\n", "RESULTS = Path(\"outputs\") / STUDY_ID / \"results.json\"\n", "if not RESULTS.exists(): # allow running from the repository root\n", " RESULTS = Path(\"2026_tdl_challenge\") / RESULTS\n", @@ -296,10 +297,10 @@ "id": "cell-04", "metadata": { "execution": { - "iopub.execute_input": "2026-07-26T18:10:52.687651Z", - "iopub.status.busy": "2026-07-26T18:10:52.687518Z", - "iopub.status.idle": "2026-07-26T18:10:52.694178Z", - "shell.execute_reply": "2026-07-26T18:10:52.693786Z" + "iopub.execute_input": "2026-08-01T20:07:37.349490Z", + "iopub.status.busy": "2026-08-01T20:07:37.349051Z", + "iopub.status.idle": "2026-08-01T20:07:37.367538Z", + "shell.execute_reply": "2026-08-01T20:07:37.366693Z" } }, "outputs": [ @@ -339,14 +340,14 @@ " \n", " community_detection\n", " accuracy\n", - " 0.767\n", - " 0.038\n", + " 0.850\n", + " 0.030\n", " \n", " \n", " triangle_counting\n", " log10(MSE / triangles)\n", - " 0.105\n", - " 0.427\n", + " 0.116\n", + " 0.407\n", " \n", " \n", "\n", @@ -355,8 +356,8 @@ "text/plain": [ " analysed by_eval_regime by_train_regime\n", "task \n", - "community_detection accuracy 0.767 0.038\n", - "triangle_counting log10(MSE / triangles) 0.105 0.427" + "community_detection accuracy 0.850 0.030\n", + "triangle_counting log10(MSE / triangles) 0.116 0.407" ] }, "execution_count": 2, @@ -415,16 +416,16 @@ "id": "cell-06", "metadata": { "execution": { - "iopub.execute_input": "2026-07-26T18:10:52.695582Z", - "iopub.status.busy": "2026-07-26T18:10:52.695502Z", - "iopub.status.idle": "2026-07-26T18:10:53.111369Z", - "shell.execute_reply": "2026-07-26T18:10:53.110836Z" + "iopub.execute_input": "2026-08-01T20:07:37.369619Z", + "iopub.status.busy": "2026-08-01T20:07:37.369504Z", + "iopub.status.idle": "2026-08-01T20:07:37.801557Z", + "shell.execute_reply": "2026-08-01T20:07:37.800784Z" } }, "outputs": [ { "data": { - "image/png": "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", + "image/png": "iVBORw0KGgoAAAANSUhEUgAABokAAALmCAYAAACerH0sAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjExLjEsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvctoD+AAAAAlwSFlzAAAPYQAAD2EBqD+naQABAABJREFUeJzs3QV4FEcbB/B/lAju7u7urqW4FileqECBUlpatAUKpRT64S1Q3LVQWiha3N1dEzwh7rnveSdcepbkLsldEvL/9dmSzO7dzspdduYdsdNoNBoQERERERERERERERFRqmKf1BkgIiIiIiIiIiIiIiIi22OQiIiIiIiIiIiIiIiIKBVikIiIiIiIiIiIiIiIiCgVYpCIiIiIiIiIiIiIiIgoFWKQiIiIiIiIiIiIiIiIKBVikIiIiIiIiIiIiIiIiCgVYpCIiIiIiIiIiIiIiIgoFWKQiIiIiIiIiIiIiIiIKBVikIiIiCiZefr0KR4/fhxnGqVMERERePToEe7cuYPULrnf176+vnjw4IHZ23t4eKglpdxzPj4+uHv3rroOQta/fv3aZtfQ3Ncm1/OaGqXGa2HL76mkPr8vX77E/fv3U9T3dGIwdYxyHuRvABEREVFqYKfRaDRJnQkiIqKUzNPTE4GBgSbXZc+eHenTp7fo/Vq3bo0nT57gwoULsabF5dmzZ/D390e+fPmQJk2aGLeLjIxU20peHR0dLcprUu5fzru8Nm/evEgpJk2ahKlTpyJbtmxwcnJ6ZwJF8b0W8bmvbalJkybInz8/li5datax1q1bV/175MgRJBdTpkzBDz/8gCxZssDZ2Vndc/J91bVrV+zbtw+5c+dGgwYNsGTJEtjZ2WHUqFH48ccfbXINzX1tUp7XlPg9Y03J8R63tsT+nkrO3yGffPIJ1q1bhzdv3qSY7+nEYOoY+/btq9L27t2bpHkjIiIisoXEqwkiIiJKpXr37o0DBw6gUKFCRuskINClSxeb5eX58+dYuXIlNmzYgMuXLyM4OBinT59G1apVjbbdtGkTFi5ciKNHjyJDhgzw8vJCzZo18dNPP6F69erJfv9y3qUi68yZM0gJzp49i/Hjx2Pt2rXo1q0b3iXxvRYSoLC3T54d2//8808cOnQIt27dSrH33cWLFzFmzBije+7XX3/F/v37VS8iuQZaRYoUQdasWd+Za5gYUtL1ppQhpd1T7/pnPCby97p48eLqb0GbNm2SOjtEREREVpX6nvaIiIisIF26dKqFvuFiywCRkGCV9MqZN28eJkyYEOu23bt3R65cuXDz5k3VslkWOY769evj3LlzKXL/yZm2QlCOj6JIkHD79u3J8nRIsLJFixYmg78phQRoTd1zp06dQtGiRfUCREK+s0aOHPnOXEMiSrjU+hkvXLgwmjVrhunTpyd1VoiIiIisjj2JiIiIbODhw4dwcXFBjhw5jMb/lyHZEqsiWnoLaHsMyFBSsZkzZ44aWkYrc+bMWLx4MXLmzKmCPL///rtKlxbPr169Uulp06bVew+Z5yQ8PFxVplhr/6bIfoOCghASEhI9ZJu0dNbmQzunQ548eaJ7OMkIu/Lecs5lLhbtazJlyqQWQ7rvIa+V93B1dVW9nmIi8xdIvgyvs8xtoM2nvK8M9yXvI8PO6ZK8yTFJ8MzBwSHG/BgekymG+ZfgnZxjw6H/JN3d3V0F6AyZc64Sci1kHgi5f2RIQlOkd1lYWJjR+TT3nMV1XWJy/fp1NdzTmjVrLDpWXebeM+YegykyZJWcQxm6UIZrNLzntL2gtPeclsy9IfkzNdyhXF8Zms7ca2HqGpr7GYuPxDqviX1vW/MzFtO+5PvY1HvG97639HXmXovY7lP52+Ln56f3N1DuVWk0kDFjRr2ebXLMoaGhagjIuMS2z/h+v2vJEGTy/qbyIefuxYsXMQ6zmty+QyQv8tmW7+OYXpcYn3HD/cT27JOY1y6h30U9evRQPb/kb0KpUqXMfh0RERFRiiNzEhEREVH8NWnSRJMhQ4ZYtylQoIDmgw8+MEofMGCAJkuWLHpprVq10lSoUCHOtLhMnTpV5h3UnD592uzXhISEaBwcHNT+tB4/fqzJli2bpn79+pqwsLDo9J07d2rs7e01o0ePtur+TWncuLHG1dVVkyZNGk2RIkXUUr58+ej1derUUcvx48c1JUuW1GTKlEnTqVMnte7bb7+Nfk3+/Pk1Tk5Oapu//vpLbx/a9zh79qymVKlSmuzZs6vj7dChg8bf319v282bN2uKFi2qcXd31+TOnVvtb9SoUdHbVatWTV1nOR+FChVS+x43blz06xcvXqwpWLCgxsXFRZM5c2a1/PTTTybzY+qYTNFuf+zYMU2xYsU0WbNmVedrxowZav3hw4c1xYsXV+mOjo6aESNGGL2HOecqIdcipvtazoe8j+QrR44cmly5cmnmz59vtE1c5yyu6xKTmTNnqmvl4eGhl27usZpzz5h7DKaEh4drxo8fr45HvnskP3JfrV27NnobU/ecdjE8Bu0i28r5seRamLqG5n7GzP1es8Z5Tcx729qfMd19HTlyRN3T8p0s56Bjx44ab2/vRLnvzXmdJdfCnPt0zpw56r7z9PSMTvvll19UWosWLfTer0yZMprWrVvHegzm7NPS4zC8T7/88kt1nZ4+fWq0/549e2rSp0+vCQgIMJm/5PIdEhoaqhk8eLDKh5wr+VyvWbNG8/HHHxs9zyTkMy5/0z/55BONs7OzypvcVzHtxxrXLqHfRQ8fPlT3ovYzTURERPSuYpCIiIgoEYJE6dKl09y+fVtvkcqFlBYkWrFihXqNbgBD7NmzR1XCjBw5Uv3+4MEDVeHTqFEjVbFj7f3HdN6rVKlicp1UIElFkFQaPX/+XBMZGanZv3+/yW2lUmnYsGGqYvTu3bt67yEVUN27d9e8ePFCpR08eFBV9n7//ffR2z169EhVPI0ZMyY6iObr66sq6i5fvmx0Pl6+fKm3f9nOzs5OM3v27OhzKRVYUuk3bdq0eB2T7vY9evSIrkResGCByoNU0rVv317z+vXr6IpGSZfAX2xiOlfxvRam7uspU6ao8/Hjjz+qCkYhlbFDhw616JyZe11M6dy5s6p4NCWuYzXnnjH3GGIi50KObdWqVep8ynmSz6Zcw/Xr18d5z8V0DIZBInOuhTnfTTHdN5YEiax9XhNyb9viM6b7OdJeT/lulcCdfA9rxfe+N/d1llwLc+7TGzduqN+XL18e/br3339f/X2RYEpwcLBKkyCSbDdr1qxE+WxYchyG9+mdO3fUPTZx4kS97eS6SGBj0KBBseYxOXyHfPrpp+r8btu2Tf0eGBio6du3r6ZGjRpmBYnMvW/lXLi5uWl27Nihfg8KCtIMHDjQ5H6sce3MzWdsxygBWfncEREREb3LGCQiIiJKIKnwkQCKYat8aTGckoJE2h5D0updW/GiSypf5P2kVa/0UpAWwc+ePbPZ/i2taJNrcuvWrRhfL5VVEsiTgJ5Ugsr2Utmm+x5S6SSVp7qkJbv0DtDatWuXOk5p4R8bUxX2b968URVovXr1Mtp++PDhmowZM0ZX2JpzTIbnQPIv51UrIiJCBT+kVbcE+rSkQk5akvfu3Tte5yq+18LwvpYKdang7NatW4zHZe45M/e6mFKrVi1N2bJlTa6L61jNuWcsue6G5DMn+5DvDl1yDaWnhe5+EhIkMudaxPXdFNd9Y0mQyFrnNaH3tq0+Y9rPkW7Ftvjtt9/UdTt06JD6Pb73vbmvM/daWHKf5suXTwXahAQGpAJ/+vTpKgCyd+9elS5BJMnftWvXYsybJfs09zhiuk/fe+89Td68efUaSUgwVfJ48uRJTWySw3eI9NiV7XRJ7yf5+2tJkCi2+1YCyrIfwx508hr5jOjux1rXzpx8xnWMsv/q1aubXEdERET0rrBP6uHuiIiI3gUyL4TML6C7xDUnT3Ii8w61bt1ajd0v87AYzpUjxo0bh/feew/du3fH+fPnsX79eovnukjI/i0l8ycUK1bMKP3KlSto2rSpumZVq1ZFixYt0L59ezW3wYMHD/S2lXkiDOfLkfeUeSW0KlWqpOZC+PDDDzFz5kxcvHhRvZc5jh07pubfqFy5stq3zCNz7949tRQsWFCdlxs3bsR5TDGR/OfNmzf6d5mTQd5X0goUKBCdbmdnp+aGkLmz4nuuYmNuvo8fP47g4GC0bds2wecsIddF5sqQOWTiw5x7xtLrruvUqVNqfhCZUF2XXEO5VjIPkczzklDmXIuYJNZ9Y83zmlh5tOVnTN7XcO6aJk2aqH9lDi0R3/vekteZcy0suU9lm71796r9yX0XEBCg8lG+fHns3r1bbbNnzx41B01s88JY+tkw5zhiMnjwYDU30Z9//hk9j85vv/2GsmXLonr16kgIa3+HnDx5EhEREWjcuLFeupubG2rWrGlWHs25b+V6yH4aNmyo91qZn7FGjRo2uXaJ8TmXvwXyN4GIiIjoXeaY1BkgIiJKDaSiwxSpQElqMmm4BH+uXbuGjRs3GlXS6B6DVOzs2rVLVX5WqVLFpvu3lEzgbWrybHl/qWCSCiKpdBRhYWGq4srwepgKVklFmlSeS4WWo6OjmlhbKjZ/+ukn/Pzzz/jyyy/VZOv9+vXDxIkT1fvGRCbyFvK6uXPnGq0vUqSImqg9tmOKjan8ywTfMaVLxWJ8z1VszM23doJx3cnq43vOEnJdMmfOjGfPniE+zLlnLL3uunx9fdW/WbJkMVqnPW+yTWznMLGuhSmJed9Y67wmZh5t+RmT+9KQ9j6Q71ER3/vekteZcy0suU/lHCxZsgQXLlxQwaBy5cohZ86cKl1+nzZtmgoiyd+J2Fj62TDnOGLy/vvvq2DMggULVNBB/i5KoOaXX35BQln7O0R7r8R2P8XG3Ps2tv0Yplnj2iXW59zb21vdj0RERETvMgaJiIiIbEAqRHQrB7UeP36cpOdfWiK3atUKZ8+eVT2D2rVrF+O20jNq0qRJ6NixI3bs2KFaUkvFnq32bykHBwejNGmtLJX/v/76a3SFkZDW19ISPL6kdfvSpUuj30t+njx5sqoU/v7772N8nTYPUgnas2fPeB2TtSTmuTI339r9xNbC25JzFt/rIhWs586dU8cpPUMSm6XX3VTATXoxGJJW9JJfqfBPrDxa2vPHWp+xxDyvBw8eTLI8JuRcmbrm2jTdQGx87/v4vi6h96n0hpJGCNJrSIJCzZs3V+ny74wZM9TfHjlPcTUgsNVnQ8h7ffLJJ/j2229Vz+H58+fD2dlZ9YCyhYR8h2gDHrHdT4lx32r34+HhYfQenp6eVr92ifFdJNs9ffoUderUsWjfRERERCkNh5sjIiKygZIlS6pW0tKCVTdAdPTo0SQ7/9qhpKT1+Nq1a1XwJyZSodOjRw/UqlVLBXNkOCKpQExIkMiS/ZsiQyNJS2FLaIcQk+CULmkNHl+611RIq2UJpklL58uXL8f62tq1a6vKKzmPpoZ2MnxvW7LkXMXnWpgi95ecDxm2yfB8SOtwS85ZQq5L/fr1VSt4U8M1JcaxJuS6y3BQmTJlwqpVq/TSZTikLVu2oEGDBkibNi1scS1s9RlL7POaFPd2Ypyr58+f499//9VLW716tQqwaHvZxPe+T8jnJaH3qeyjQoUK2LBhA86cORMdDKpXrx7SpEmDb775Rh2jdmi9xNhnYhgwYIAKDEmgaOfOnaqRgzk975L6O0Q+2xkzZlR/d3XJM4l22MLEuG9lP3Ks69at00uXoMvhw4etfu0S47tIhquTPMj+iYiIiN5l7ElERERkA0OHDlUVMh9//LHqgSMta6XlsfSiMaz0S4iQkJDo3kmvX7+ObpkrFULSm0PmxNC2ju3SpYtqoT116lRUrFhRtYbWkqFYtPNsSIVw165d1c9SiSfDuMgxSIBryJAhatg5qeCz1v5jIvM+bN++Hdu2bUPp0qXV+xvO12FI8ilDGY0dO1ZVXknlpAxxJwErqYyMD6kAW758Ofr06aPyIZWZmzdvxsuXL/HBBx/E+lqpYJRKsTZt2qjhi+R85s+fXwXlpLLun3/+Ua2hk4Il5yo+1yKm87Fs2TIVPJSK4hEjRiB37txqfpRZs2ap3j3mnrOEXBf5XMoxSs8GeW1iH2tCrrsMqSTnom/fvmqR7xSZw2X8+PHqs/q///3Porwk5FrY6jOW2Oc1Ke7txDhXkhcZDk4q2UuUKKGGONMODaed8yu+931CPi+JcZ/KPTZ9+nR13BKk1f4dqFu3rhpqTuYnimsOPFt9NrQkICR/G1euXBkdNDJHcvgOkXM9cOBAfP755+jVq5e6znIvyXdfXPMpmnvfyn7kPWU/w4YNU72sZF4haWQied6/f79Vr11ifBfJ3wDZVs4LERER0buMQSIiIqIEkta82vkBYiJz+WzdulVVgkglnARWFi1apFq0Gk60LBWxhkNcmUoz5fbt22p+BN1hs0aOHKl+Tp8+fXTFrlS+XL9+Xa1fvHixWnRJxYrkV0gwS1qwS5BL8qG1cOFCFQCSVtRS2eXk5GSV/ccWeJN8yXwZMleBVDJJBbaIKcAkeZTKMxk6SSqOpPJHejBJoEpaNuvOcxDTe8icCZJv7TxTUsEm20oFq1xfSS9evLgK/um2PpZW0vI6w6HXZFJvaak/Z84cNbST9GCRyj6pKJVK4LjyE5OYtpf71dRcFZIuk3vH51zF51rEdF/LJOPnz59XlYKyXzmfEkTUbfVuzjkz97rEVPkrQcwVK1Zg+PDheuvic6yG94y5xxATOTaZF0zmIpEKWLlW8h0jlca6lc0x3XOG11pLttXtCWHOtTC8hpbcN+Z+ryX2ebX2vW2Nz5iQAL3ck2PGjFHbyHeq3AODBg2K3ia+9725r7PkWph7nwqphJceIxJAkeHttDp16qTm+uncuXOMeTc8DnP2aclxxHaffvbZZypIlC9fPrPn00sO3yEfffSRur/kPPXv3x9lypRRw7JJzzQ537oS8hmX/ch3yrx589T8VvK3XZ4ppJea4fxYiX3tEuO7SK6t3Hum5kAiIiIiepfYaUz1TyciIiIiSsUkiCm9F2SOFKmMJUpK0qNGmDMcGNmOdg4l6fFi6bxNqZkEHmUYuNOnTyO5kp5OLVq0wKVLl9R8XURERETvMs5JRERERERkQCoFZRgvGXKNiMgUmaNPeupJbxwyFhERYZR29+5dNVytBNeSM+lVJ38DGCAiIiKi1IA9iYiIiIiIiJIx9iRKXu7du6eGeuvWrZsaJk2GXyVj0vNt9uzZ6N27txqS79q1axg9ejRkMJOzZ8+qIeKIiIiIKOlxTiIiIiIiIqJkzNI5yci62rRpo+awkR5EP/30E093LMHNFy9eYMmSJbhz546ab0rm+JG5ChkgIiIiIko+2JOIiIiIiIiIiIiIiIgoFeKcRERERERERERERERERKkQg0RERERERERERERERESpEINEREREREREREREREREqRCDRERERERERERERERERKkQg0RERERERERERERERESpEINEREREREREREREREREqRCDRERERERERERERERERKkQg0RERERERERERERERESpEINEREREREREREREREREqRCDRERERERERERERERERKkQg0RERERERERERERERESpEINEREREREREREREREREqRCDRERERERERERERERERKkQg0RERKnA8uXL8eWXXyI52b59Oz766COEh4cjJUqO55SIiIiIiBLPuHHj8OuvvyarU5oc85Sc8PwQEVmOQSIiolTg8OHDWL16tV7a77//jq+//jrJ8nTu3DmVh/gEiWyV99j2Y+qcEhERERFR8vPNN99g4cKFFr9u48aN2Lt3L5KT5Jin5HQ9eX6IiCzHIBERUSrQt29fzJw5Uy/twIEDWLNmDVIiW+U9tv2YOqdERERERJT8rFu3Dvv377f4dZMnT8ann35qlTyRda4nrxkRkeUc4/EaIiJKYerWrZvUWXjn8JwSEREREb3bOnfunNRZIAvxmhERWY5BIiJK0SIjI/H333/j6NGjCAwMRNmyZdGjRw+4u7tHbyPp69evx4ULF2BnZ4fKlSvjgw8+QJo0aaK3kTGdHz16hClTpmDHjh3Yt28fMmXKhIEDByJXrlxqG9mPdOt3dXVFnz59ULx4cb28JMZ7TJ06FY6Ojvjqq6/00v/44w/1ntKl3t7e3mh/e/bswc6dO+Hk5IROnTqhevXqRvPnXLp0CTNmzFC/T5s2DSdOnIC3t7eaF0hr3rx5qnu+DKU2e/ZsvXMkDh48iJUrV2Ls2LEoWLCg2dfp1atXWLZsGR48eKCOWXrhxHZN5ViPHDmCgIAAtX2vXr2QOXPmOPOuzW9c72HO/RPXfgzPaXzvt7iuHRERERERxZ88y79+/RqnT5+Ofq7Pnz8/xo8fr36WeUalHNCzZ0/VQ0W2a9SoETp27Kjmt8mTJw8++eST6PebPn06bt68qX52cHBAtmzZ0LhxY7Xo0r6vlENkmOozZ86obaUcWKBAAaN8Xr9+HWvXrsWbN29Qs2ZNVYaYNWuWKl9I+Ssu/v7+2LBhAy5evKh+r1atmnoPKWdYSspQmzZtUu8l5VPJT4cOHVT5RsvT01Odr7t37yJt2rTqnL333nt6Za1BgwahdevWaN++vd77yzmUYce//fZbi89XXNfT1DWz9bUgIkppONwcEaVYL1++RK1atfDhhx8iKCgIhQoVUg97UikvD8ji3r176mFw4sSJKkCQIUMGjBkzBhUqVMCTJ0+i30sCN/JALQ+UmzdvRr58+fDXX3+hatWqePbsGUaOHKkq/vPmzau6tcsD9/379/XykxjvsW3bNvUaQ3JcMj+OPGgb7k8CDRKAkf1duXJFnRMJfMQ2f06ZMmXUQ7EELuShV7tIISd37twqGCXBIkNyHv/55x/1EG4uKUDJNZD8y3vLuZCCgp+fn9G28rBfu3Zt9O/fXwXDChcujC1btqBEiRIq6BJX3s19D3Pun7j2Y2pOIkvvN3OuHRERERERxZ88w8szfZYsWaKf6cuVKxe9Xp7ppXwmDbZk3lQpA9y4cSPG+W1Kly4d/T7ly5dXDeKkfDNs2DC97eR9ZfhqaYQmwRYJRkjgpVKlSnj8+LHetpIu73X8+HG1nTR269evnyobSIPBuJw/f141jJMGbFmzZkX27Nnx3XffoUaNGirQYQkpExUrVkwN2+bm5oYcOXKoIbjbtm0bvY00cpP9Sb6l8WBoaKgKIrVp0wZhYWFqGym7ShlQ3s+QlHml7Buf8xXX9TR1zWx5LYiIUiQNEVEK1bZtW026dOk09+7d00v38PDQBAcHq5/r16+vyZo1q+bly5d66zNkyKB57733otM6deqkcXNz0/z000/RaT4+Php3d3dNgwYNNJMnT45O9/PzU/v97LPP9PabGO9Ro0YNta2hMWPGaOQrOywszGh/U6dOjU6LjIzUVKlSRS26BgwYoMmRI4deWs+ePTV58uTRmFKqVClNrVq19NJu3Lih8jBu3DiNJerWravJly+fOhdaly9fVtdA3i8oKCg6vWPHjpqMGTNqHj16FJ0WERGhadKkiaZMmTLq+OLKu7nvYc79E9t+TJ1TS+83c64dERERERElTIECBTQffPCByXXyTJ82bVrNP//8E53m7++v/i1RooR6do/LunXrVNnmypUreu8r5Y1///03Ou3Vq1caV1dXzbBhw6LT3rx5o0mfPr2mXbt20WUVsXz5cpUvw/KBYZ6k7JI/f35NtWrVNCEhIdHpXl5emuzZs2sGDRqkMVdgYKAq/1SuXDn6HGhpy01SrsuUKZOmYcOGeuXTXbt2qXOgLffKOvldyrKGpMwrZV9d5p6vuK6nqWtmrWtBRPSuYE8iIkqRpDfK9u3bVfdy6QGiS3qrSMsiaRF06NAhDBgwQLWm0l3fu3dv7Nq1S/Um0QoODsbgwYOjf0+fPr0a+kveY+jQodHp0pVeWiuZahGVGO9hCdnf559/Hv27dP9v2bKl6jEj3ffjSyZnlZZTuj1v5s+fr95fWlGZS4ZUk5ZXMuSenAst6W1Tr149vW1fvHiBrVu3qveXnjVa0htI8nP16lXV2yY25r6HOfePpeJzv1nj2hERERERkWWkx0jz5s2jf9cdvtwUGVpchjeTIc2kTCFlEG2PHl3S26ZBgwbRv0vvF+ndo1sOlF41vr6+GD58uN5wbjLigQxTHhfp4SLlrm+++QbOzs7R6TL0ufSckdEsNBqJ18RN8uLh4aFGQzA8B9pyk2wjQ3LLEG4yFJ1WixYtVM8cGZ48vsw5X9Z874ReCyKilIpzEhFRiqQdA1q3W7khGRtZlCxZ0mhdqVKloreR4QS0lfnSnV6XVPZL9/p06dIZpWvzoCsx3sMSMteR4cO7HE9ERIQKSGjnQrKUjM08evRoLFiwAL/99psae3nFihVqnG3DoEpstNfAcO4lIcO/ybxBuuM+S+Hl7NmzqrClLcjIv8+fP1c/y/B8sV1zc9/Dy8tL/Rzbe1nK0vvNWteOiIiIiIgsY6q8YoqUK2SuIplXVAIH8pwvz/RSvpBgjOHQbkWLFjV6D3nml/KKYTlChnjTpR06O64GZNeuXVP/yv53796t8qgtB8kcqj4+PiqoYzg/q7XK2TKsdnyZc76s+d4JvRZERCkVg0RElCJpWyxJb4yYaOeOkfGRDYWEhKh/dSfxNNV7RB4GY0qXynxDCX0PafmlHcNZl7RmMsXFxcXk+wpT+TOX9PqRiVtXrVqlJhWVB30p8EgvmfhcJ+351mWYps13kSJF1DxOhmR8axkbOjbmvoe0jovr/rGUpfebta4dERERERFZRnfUg9hIEEbmpVm7di26desWnX7q1CmT28f0zK/7vK8tM5kqR0iatowQE+16mQdV5iLSJaNaSE8nc3vBJEY5W7te/pUlpvKtbq8nS85XfNniWhARpVQMEhFRiiQtm+Sh8sSJE2oYMVPKlCmjHuJMtTqSNKmwN9X6KSnlzZsXx44dM0rXHfYtscgDe2zDDsiwedKLSHoQLV26VA1XIJORWkImdZVrIPmXIdd0GQ7FIMEbuSbSEk8KMvHJu7nvIa3o4rp/YtvPu3K/ERERERGlBpY818fmwYMH6t9q1arppR8+fDje76nttXPx4kW9URtkNIc7d+7E2ctJ2zhOGsp98MEHSIgqVaqof6WcFFNvIm3DPSnjSHlPKzIyUpX7tOtluDYZaUPbQE8rICAAt2/fVuWnpL6eiX0tiIhSKobAiSjFtvSSIMbq1avVuMGGrbvkwVMCAd27d1djIp87dy56vQRhpCt+//794xxr2taaNm2Khw8f4sCBA9FpmzZtUvPdJLY8efLg9evXCAoKivEBuW7duvj+++/V+ZOeRaZaX8VGxnnu2rUrlixZghs3bkSny5jdMm62rgwZMqh5mxYuXKjG+DZskbZs2bI4827ue5hz/5hzjnSlxPuNiIiIiCg1kOf6J0+eJPh9tIENGW5Od4i2hAyxJnP5SEBi8uTJeiNITJkyxWjI8pjKkLVq1VJzEsnQ2rqkLCPlSXM1adJE9T6SMqDu0OjS20bmdNXmV4Zuk/zpzrk6a9YsNVyb7ly8kjcpb8ncsUICOxMnTlTltuRwPRP7WhARpVTsSUREKda0adNUN/j27dur1lMFCxZUc9LIXC7169dX28yfP18NkybBjoYNG6qHUgnAtG3bFjNnzkRy06tXLyxfvhytW7dWD6gydrSMhywBmh9++CFR9yU9e2bPno06deqgYsWKqhfMvHnz9IbGk0CKBD6EBDniQ95Tzre0tpM5jSQAI72SZN4jefg2vKaiVatWargEKXzIXEJSQGnZsiX69u0bZ97NfQ9z7h9zzpGulHa/ERERERGlBjJfqZS1pIyVL18+5M+fH+PHj7f4feQ5/+OPP1blpG3btqnhySQwIwESCYjEh4w4sGXLFlUGlJ45tWvXVu8pvYJkJAJ/f/9YXy89diSAI0ODy5xAkkeZE1caH0og5auvvjI7L9r3kvJnpUqVVDkoY8aMahSILl26qHKNHPOff/6p5maSoJmUnTw9PXHmzBlMmDBBzdWkJb/v3btXlaXq1auneuPItZAyWkKG/k6s65nY14KIKKWy01ijfyYRkQ3JA+nx48fVWMdly5ZVi6GrV6+qLuPy0CsPu4bDfkmvE5lsVB58dUkFv7R6Muy2/++//+Lp06fRAZTEeg9tN/1Dhw6p3kPy0F25cmXVlV8ezOXBX44htv3JxKXSe6VHjx5wc3NTaUeOHFE9dyRNl+RL1klgQ/YrARTtOMxCJjmVgI6cs4RMFip/amQ/MjyDBL1q1qypetvI0q9fv+hxq3VbvMk1lSCZDMEnhQrJhyV5N+c9zLl/YtpPTOc0IfebqWtHREREREQJJ0OcSSBDhg6TUQC0Q2mvWbNGlRe0DcV0SS8cKUNIDxtdly9fxpUrV9S6Ro0aqZ428j4SVNAOwRbT+8ZUPpTRC/bv36/KYNKbRwIpUjaRIdtktIO48iQkoCHlNpk/R3rESFkypgZucZGyiZRpZD4jaVQnx6JLjlmGpbt3754aMUGOPWfOnEbvI40E5bj8/PxU0KlAgQKqd5G8XoJOWpaer5iup6nzY61rQUT0rmCQiIiIYrRx40Y1XJz0nvnss894poiIiIiIiGxAGqlJUELKYT///DPPeRLitSCidx3nJCIiohj9/vvvqhWWdOUnIiIiIiKixCc9dp49exb9u/SykTmGpEeQDNNNtsNrQUSpEeckIiIiI1Igka770sV+yZIlRpN0ylBry5Yti/XMSQ+k5s2b8+wSERERERHFQoa+lrmEZG6dbNmyqWG5ZWjyhQsXoly5cgk+d0uXLsXRo0dj3Wb06NEoXLhwqr9O1r4WRETJEYebIyIiI+vXr1et1mTsaZn81NDNmzdx+PDhWM9cjRo1+BBNRERERERkhpCQEJw+fVrN45oxY0YVqJB/E4PMeXvr1q1Yt2nXrp0KipB1rwURUXLEIBEREREREREREREREVEqxDmJiIiIiIiIiIiIiIiIUiEGiYiIkimZuFTmBCIydO/ePezduxeRkZEWnRwZMuHOnTs8oURERERW8vz5c/WcFhAQkKLPsQw9Lcfx5s2bpM5KspQcr7O183T37l2cOnXK4tfxXqKY+Pj4qHv21atXNrkXiShmDBIRESVT/fr1w6JFi5I6G5QMrVixAs2aNVMFLkvs27dPvU7G2CYiIiKixHfgwAH1vHX//n2bNyCKa84ZIQ2GZG7Rhw8fxrrd7NmzVXnE1dU1Ou3169dqP7J4eXmZfJ0ct3abiIgIo/VhYWGqgvfo0aNqvhdTjZ5092NquX37Nsx16dIlnDt3Lsb1T58+Ve8ZGBiI5H6dkzpP3t7eqF27tjqnljB1LxGJy5cvq3v2xIkTNrkXiShmjrGsIyKiJLJ79278888/uH79Oq8BJZrPP/8cP//8M2bNmoWvv/6aZ5aIiIgohZJgy//+9z9s2LBBBTo8PDwwbNgwlWaKBIW6deumAkmlSpVSlatSybp69WpkyZJFb1sJAE2ePBlTp05FmjRp9Hqlt2zZUv08fvx4fP/990b7+fjjj7Fnzx71s5+fH9KmTasXLJgyZYr6uUiRInj8+LFq9NS/f3+MHTsWbm5uevuRfObOndtoH127dkWxYsXMOk/t2rVDly5dULlyZZPrpcwlAQwpd5UsWRLmypkzJ5o0aaJ3fO+6qlWrqusycuRIVV41R0z3EpGt70Uiih17EhERJUMzZ85E/fr1UaJEiaTOCr1D3N3d0b17d8yZM8dky04iIiIiShmCgoLw4sULLFiwIM5W+OHh4WjdujWCg4NVr6MjR46o3jgSMJLAkaElS5ao4E3Pnj1Nvp8EbqRnu0aj0UuXoI/0XM+TJ4/Ra+bPn6+CWJ999pkKaElPIglcSQMmWefp6Wn0mhEjRpjsSTRo0CAzzhBw8eJF1VtJAkWJrWHDhiovBQsWRGry0UcfqSDglStXzNo+rnuJyFb3IhHFjkEiIkp0MnyA9gFe5tQ5c+ZMrN33pRWcPMAfP348xjGvY9tGWpqdPXs2xnzoFl6kO/OxY8fUz/KwKnmTJTHzLfuQrv6mSEEutuEZxKNHj1RrGKnMN6SbfykYnjx50mhYCUmX8XllW8OCmy45fhl6QYabkPmPDJl7PnTzJMOYyb7Pnz9vcRBCxiOWcymFXGl1aLgPS89pXMcX270QEzk+eQiVfL58+TLG7eLat6XbPXnyRBXmpZVoQvMnFQHyfjt37oz1vYiIiIgocUn5QZ6T5bnOVFBE97lYgig3btyIcd6O9OnTY/r06apFfVy2bt2qnhEnTpyIDBkyqLQcOXJg9OjR6n1lX7oWL16MVq1aqX2Y0qtXLxV8+ffff/XSly9fjkyZMuH99983es2aNWvUvseNGwcHBweVZmdnhw8//FC9jzV65Gzbtg3ZsmVDrVq1TK6XgNm1a9fUz/L8rC37aJ+jJZglv0twTci2EgSTIfJMzf8jQ2Bp30O2k7KGqblWDN/36tWralt/f/8E3RNxkdceOnQoxmEJpeyoOxSgqbKkXFu5L+QeMUdM95LuOZD9SM82KdNLQFM3P9qyTVzDZcd2bNa8LpZ+zrX7kOtnSI5d1pmaQza244vtPo2JOdfanH1bup3hfZyQ/Fl6LxJRHDRERIls6dKlmiZNmqilYcOGmoIFC2rc3Nw0P/74o952kZGRmp9//lmTKVMmTc6cOTU1atTQZM2aVfPxxx9rQkNDzd5G0ho0aGCUjzFjxsiThCYsLCw6rVOnTpoiRYpo/v33X03x4sU1lSpV0tSvXz9R8/3rr7+q/R4/ftwoT4MGDdK4uLhovL29Yzx/S5YsUa+/cuWK0Trd/JcoUUJTrVo1TZo0aTTdu3fXhIeHa/bu3RudLvtp1KiRJjAw0Oh9JP8ZMmTQFChQQFO1alWNq6urpkePHhp/f3+Lr6M2T3K8pUqV0lSvXl2TLl06dX4fPHigiYucs6FDh2qcnZ01JUuWVK+TYxo1apQ6JjFnzhx1Ts6cOWP0+sGDBxudU3OOL7Z7wZRVq1ap61y4cGFNnTp11LV///33NU+ePLH43Jq7XXBwsKZnz54aBwcHTbly5TRFixbVjBw5UjNhwgR1PoKCgizOX0hIiDq/n3/+eZzXhoiIiIgss3btWvWcdvnyZaNn/CxZsmhy586tqVy5snr2bdWqlebFixdGz4jyrCbPiBUqVNA0b95cs2PHDvWef/75p8l9Pn78WK0fNmyYyfUDBgzQ2NnZGT2P3r9/X71u3Lhx0WmPHj1Saf/73/+M3mfnzp1qnRyj5K1379566+XZWp4x5flctvPz84teJ+UJKSNoy3Cx0e5n0aJFmoSQ89y/f/8Y1y9fvlxTunRpta9atWpFl32OHj2q1i9YsECtO3bsmKZu3bqaihUratKmTauewU1d5wsXLkS/hyzly5dXz/GdO3dWz/Va2vc9deqUKq9JPuW+SJ8+vbrWhsy9J2K69/bs2aPKEZkzZ1ZlV/lXyj43b96M3ubSpUvqXEh5QsoSci3LlCmjOXjwoFF+mjZtqsp9cYntXtI9t7I/KQ/J50P26enpqcowck0kXfKUN29ezfXr143ex5xjs9Z1MSWuz7kcg+zDsEwttm7dqtbp7suc44vtPjXF3Gttzr4t2U7uAym3x3UfW+NeJKK4MUhERDaxYsUK9cd/37590WlTpkxRafPmzYtOi4iIUEEWHx8fs7exNEiUPXt2zQcffKDx9fWNfghJzHxL4Usq/3v16qX3XrJOHtak0j82EmxycnJS72lIm/8+ffpEB38kOCOFvokTJ2q6dOmiCQgIUOmnT59W6YYP5dOnT1f5lwdYrdu3b6sHb8M8m3M+tHmSQqK2ICgF1WzZsqngVVy++OIL9ZC+ffv26LT169ervI8dO1b97uXlpR4oP/vsM73XykO9BOt0z6m5x2fJvSD7d3R0jM6PbgFWN3Bl7r7N3W7IkCGqQLh///7otF9++SW6MKt98Dc3f1ry0C4FLiIiIiJKXKYq6v/66y+VNmLECNXgTFy9elU16pGKcG3ali1b1HY//PBD9Gulwlq2SUiQqF69euq515CUN+Q5vGvXrtFpGzduVO+l+/xpKkgkz6Tu7u7Rz/9SgSvrzp07ZzJINH/+fJXWsWNH9XyqPWZTtPuR8yUV0IbLy5cvY3yt4TnZtm1brNtJwzjZzlQAQlv5LhXR165dU2k3btxQga6YAjKGZL0EDUaPHm30vhI8kPfTNuSSAIaUB+RnLUvuCVN5OnnypCpbSqBQGxCR8mqLFi00hQoVii5PSJlaGszpBvEkb3I/GPrqq6/Uft68eRPrscd2L2nPgTRqu3v3rkp7/fq1CiDI/SiLlI+ENAaUvLZu3VrvPcw9NmtcF1PM/ZzXrl1bNew01KZNGxVc0jaUNPf4YrtPTTHnWpu7b3O3++OPP1Qepc5CS4JakhfD+9ga9yIRxY1BIiKyGulFcujQoeiHeWmBI3/EhQQ4pFDRvn37GF9vzjbxCRJJWmyBoYTmW0gLOqncf/XqVXSatjeMqYdkXW3bttXkyJHD5Dpt/rUP0lpVqlRR6YatdeTcSCsoLXlwk+OR9zE0d+5cjb29vWq5Ze750M2T9iFeN8AhPY9iIwVHOU/dunUzWteuXTt1rrUPm9LLJmPGjHoP+2vWrNE7p5Ycnzn3gtbFixfVtitXroxxG3P3be52EriSczNw4ECj7aTllW6QyJz86WrWrJkmT548Zm1LREREROYzVVEvFZ5S+WtYyawtH0jPdiG9AKRFvmFjsa+//jpBQSJ5dpQKW1Okd897770X/fvs2bNjDH7oBomkZ4RUDmsbPfXt21ftR5gKEkkF+axZs9ToBLJOGs9JLwGpNDYM+mj3Iz0EdHuAaJcTJ05o4iLP1dJT39SoCpYGiWbOnGm0LrYgkZQBpSGfjPIg5SfpRSWNtAzfV/7Vpa1Il8pzLUvuCVN5kiCMlC0NAyZS6S7brl69Wv0uDfykIaI5pk2bFuM50xXbvaQ9B7/99pteuowmYdggU1u2l3KSbuDA3GOzxnUxxdzPuXbkkMOHD0dv8/TpU9XoTzdoZe7xxXafmmLOtTZ33+ZuJ+dGegQZ3scSUDO8j61xLxJR3BzjGo6OiMhSMq9L//791Vi6pUuXVmNPy3jTMv+KTGYqZC4fGb+5cePGMb6POdvEh7u7O8qVK2e1fAuZkHXOnDlqos6vvvpKpcmksoULF1aTnMZGxhJOkyZNjOszZ86s3kdXvnz51OSzxYsXN0q/cOFC9O+Sf19fX+TMmVON/a0d21f+leOUcYtlPOhcuXKZdT5081S0aFG9NJnEVeYwev36NbJkyWLyWGReIBlfum7dukbrJE3GEb9+/ToqVqyoJqaUscy3bNmCHj16qG3k/OqeU0uOL7Z7wVDJkiVRqlQpfPzxx2qc5ebNm6NBgwZqjHNLz62M027OdunSpVPnpmbNmkb5kXHVZX+W5E+Xi4tL9JjVRERERGRd8lwtz2fOzs566dpnYJljVZ7dZE6WDh06wN5ef/roGjVqJGj/Tk5OenO9GM6fopsv7TNibOURIc+ZMtfMsmXL0LVrV2zatAmTJk2KcXspRwwdOlQtMheQPO/KvCk//vgj/ve//6nnYsPn8hEjRqgyQHxIOULOuaurKxIqpjmNDEm5R8pPf//9tyqXZc+eXc2/JHO0yHO+IcPrKuUn7Ry12n0m9J6QuU/LlCmj5n3VLXfIIveFzJ0jZauOHTti4cKF6hjatm2r7kfDsqVuWULEVZ4w516qVq2aUfk1pnQpJ8kcPwUKFLDo2KxxXRLyOZfPy7Bhw1RZVrtO5vOSOX0ln1rmHp+l96k519rcfZu7nRx7+/btje7j2rVrY+bMmRbnz9J7kYjixiARESUqKXzIH395kHrx4oXeBKQSSNA+OEjwQMQ0Gaq52wh50DA1kaEUeEyRSVqtmW9tpX2TJk3w66+/YuTIkariXiaQnDx5siogxSZr1qxqcs2YSJDBkDx4x5QeFBQU/bt20s2DBw+anCxS8uzm5mb2+YgrT0J3/4a060ydT22adhsJBBUpUgS///67etCUCSylYCmFUe05Nff4YrsXTJEHfXnwlUDfzp07sXr1apWvFi1aqPzkzp3b7H2bu532uCVYZMgwzZz86fLy8lL3GRERERFZn1RgxvW8K8/X0kBI97lby9TzoCXkWfDq1atG6X5+fipv2gZUQvuMGFt5RKtv374qgDFt2jSV9549e5qVH2nkJYuUN/r164eqVati9OjR+PPPP5EYpEGWPGv/9ttvifJ+5pYZJAB25MgR1RBOyoNaEhTYv3+/0faGZSjD8lNi3BPSwFGCG1IONVS/fn3kzZtX/Tx//nwVsNi8eTPGjBmDly9fqkZocg7r1atnVJYQcZUnzLmXYjoHcZ0bS44tsa9LQj7n2oaK3bp1Uw0gZ82apa6lBIwkGCLlXUuPz9L71Jxrbe6+zdlOex+bU641N3+W3otEFDcGiYgoUd28eRMeHh74/vvv9R5mnz17pvdwqO0JI613YmLONtpWbNIazdD9+/dNbm8qSJOY+dbtTdSpUyf8888/WLp0qWqt1KdPnzhfV6JECfXAGFsPnPgqVKiQ+rd79+6qIBYTKUSacz4SKn/+/OpfU9dPm6bdRq6btKwaO3asurbSalEChFI4tfT4tOIK2Bk+3I8aNUot0uJs+/bt+PDDD1UQUK6XufvW3pdxbSc9w4QEw2J6D0vyp0veUwrjRERERGR90gsiruddeS6VCtWHDx8abWfqedAS0itDng2lgZJuJbm08hfVq1fXK4tonzd10015//33VeXsDz/8gHbt2sXYi13b2E63sZZW5cqVVc8OKY8lFmk0JY3eWrdunSjlAXPLDBJwkJ4kuudYSGPB+EiMe0J63UjDv71798a6nZSrpOwgi7ZXjAT9OnfujOfPn+ttK/eGlD10g4umWHIvxYe5x5bY1yUhn3Mt6SG3aNEirF+/Xp0nqV8YN25cvI7P0vvUnGtt7r7N3U7uY3PLtda4F4kobvr9/IiIEkhbMJBggq7p06er7sZa8iAhrUMWL15sFHSQAoQ81JuzjZBhCaRQofvAIN3Q9+zZkyT51pKCkjwMSauarVu3ql4dhq19TJEWROLUqVNIbBLIkNY3kn9pOWhIAlOWnI+EkiHq5PpJwEe3i7j0tlmxYoXq6p8nT57odAkIyUOj9I6RLvlyTnXXm3t88WmNqHttpeeOPKRK/rVD75m7b3O3K1asGMqWLavOje6+5d7evXu3xfnTevLkiQoANmrUKF7ngoiIiIgsI8MnHT16VPVkMGwxL8OhSbBFSM8aqdDWrUyVVvjSSzwhpBe+PMNLbwVd8jwqQ0pLbyDdgJIEc8wpi8h7fvPNN+q5csiQIbFuKxW+cg4MSeW4jFwQ03BS8R1qrk6dOmb1LtD2GtH29k8IKUMZlp8kYGVOA8OYJPSekPKTDAkmw30ZkoZlUh6RYdwMy7bSoKxly5Z49eqVUQ+akydPql4i0ggyNpbcS/FhzrFZ67ok5HMuJGgm5WD5TMqSMWNG1cA0PsdnCXOvtbn7Nnc7+Y6R+1g3iCZ5kTJ/fPJn6b1IRHFjTyIiSlTSCqxLly5qyAF5EJKgioz9KwUIwyGvpJJf5vapVKkSvvjiC7WtzD8jleLyx14KLOZsM3jwYDX/jzxAyzbyULFx40ZVGJIHsqTIt5AHlUGDBmH8+PHq9wEDBpiVFynQSODjr7/+Ug9DiW3lypVo1qyZmudHCnMSSJAAmzy8y/Btd+/eteh8JJQEfCQ/EhyTaymFHhmXXIYLlNZVumTfck5+/vln1WX9l19+idfxWUrGTJdeYHJPSeBGxj6WHmJy3VetWmXxvs3dTrrUy3ayyL0k3fnl/pN8yFCGluZP7NixQ92bhoUQIiIiIrKOCRMmqJb2TZs2VcMnSat3Ka9I7x4J1GgbaMl2Unktwyx/++23quJ47dq1qhJUhk8z7Clw/Phx9XwoladCGgdpW/TLc6Y2SCKjIUyZMkUFdOR5XuYBkWdF6W0ulbS6w2vJeilXybP/jBkz4jw2mTdIlrjIM740lJIyhjSmk9EKpCeHlOPk+KXcYUieZU31UJDtK1SoYHI/UoaQcyijD5hDKuslkCFzI8moBdLYSt47tl5RMZHnepkjVIY3kyGkJVggz95yzHK+48PSe8KQvObcuXMqP5I/qXCXxnkycoS8j8wlJcEKOWaZY0rOh5QFZb2U06QMqzuvk8xXKw00zTm/lt5LljLn2KSnmjWuS0I+51pybocPH66OwfA8W3J8lpB5j8y51ubu29ztpE5E7gO5j2U0DfnOkTKx1KscO3bM4vxZei8SUdwYJCKiRCeV0lKBLfPwyDwpUqkvLUx69eqlxpLVksKKVG7LUGwyRvCBAwdUBbe0MNEGWszZRh6+ZD8ynq9UoMvEiVLgkYcQeVjRnRyxfPnyagxga+dba+DAgephUQpobdq0Mev8OTo6qoDA7Nmz1SSOuhNfxpR/ebCX1xmSfMmDlmGX8IsXL6rjlYKFPMhKDyeZ6FL2aen5iClP0t1ezr92MsnYWphduXJFBUTkIVIKOlKYkAf5nDlzGm0vk3xKCyJ5QDQ1hIS5xxfbvWBIeo/JeZD7a8uWLarwKeNFX7hwQW+CXXP3be52UoCXyWql8CzbSu8i7WtkODptiylz8yfkYVzuRXN6tRERERGRZeT5VZ6BdYdsljKCNCaTHgPyHCeBHSlTSAMh3UpemftTtpNyjczPI0NPf/311/Dx8VHrDYdrkwZVMh+IkH3KdhLsEDIEnG5PGhmCuEqVKupZUhrSybO6BJnkWdyQNNyShmuSP91hwuT9ZD+mntFNzc+qWz6RkRWkt4Eclyxv3rxR7yeBK+llpBuo0u5HnpdlMVSzZs0Yg0RyfuW9JRBlDjkWCcJI2U7Ou5SdJk6cqCr05XlZ8mFYeR/TdZYynDYwINdaKswlCCEV3LqjJsT0vlI2kXTdRnmW3BOm8iSBGjn3UjaW10te5D2lnCiV89prKUMRSu8k6RUic8BIHmQoNMNGi1KWkOsjARZzxHQvxXQOpLGk4TyyQvIj6brlN3OPzRrXxRRzP+dact9LsEoCqJJHQ+YeX2z3qan3NOdam7tvc7eTNDkPch/LMct9LJ99CXrK95A2gGZu/uJzLxJR7Ow0pmZ7JyKiRCHBFan4//LLL1XvF3PJQ7/0LpECyqeffsqrQYlCHtSl1aG09pIgGRERERElf3PnzsXnn3+u5t+QHv62IBWyUlEvk8enJHKepPGe9D54lyXFPSFBDxk6W3rJSMO9d/1eouQrvvciEcWMcxIREVmRdLGWnjHSo8gS0gpJhlKTlkhEiUV6LMmEqAwQERERESVPhvOMyLyTCxcuVKMl2CoYIGR4MMmL4VwuKaHyWHqvvEuSyz0hI2hIg7PPPvssVdxLlHzF914kopixJxERkRXIhJUyhra0aunatasavoCIiIiIiCg2UvEpiwxP5eXlpeYykaGud+3apdIp9eE9QURE1sYgERGRFcj4wt7e3mrcZRkz2pzxgYmIiIiIKHV7/fq1CgydP39e9YopXbq0GpVAhqKm1In3BBERWRuDRERERERERERERERERKkQ5yQiIiIiIiIiIiIiIiJKhRgkIiIiIiIiIiIiIiIiSoUckzoDlLxERkbC09MT6dKlg52dXVJnh4iIiJIxjUYDPz8/5M6dG/b2Sd/2KDg4GKGhoVbfj7OzM1xcXKy+HyKixMbyHhEREVmCZb7UgUEi0iMBonz58vGsEBERkdkeP36MvHnzJnmAKItrWgQiwur7ypkzJ+7fv89AERGlOCzvERERUUou87mmywyEB1l9XzlTWZmPQSLSIz2IRJOpf8DRxT3Zn53gMOtXBCWmWsWyIiVJaefX2z8EKYlPUBhSktBwDVKS196BSClm96iIlMQOKaunqbtT0vdwsYRDCjq9/n5+qFquZPTzQ1KSHkQSIOqJPHC24ojKoYjE6mcean+ppcBARO8O7ff1yoJl4GbvgJQgd5UcSEkKdXkPKcmFGRuQklSeMBApye3F65CSFJu5ACmF3Y0jSEns85VCShJ68V+kJP51eyMlyYCUU38lI0cUKVM+2ZT5JEDkWLor4OBkvR1FhOHZtQ2pqszHIBHp0Q4xJwEiJ9fkHyQKd0xZQQwX97RISTShKev8Okda8Q+EFTghZQWJNOGRSEkcg1JOTXvadOmRkqS0IFFaZwaJrC05DVHrCns421nvmjukrHg5EZHJ72sJELmnkCBRWqeU9Yyf3s0VKYm7Y8qqFkrv7oaUJMXdv+mTvhLYXHYp7F6wT5ey6oNC3VJWxbhd+pRVpk6PYKQ0yanMZ+fkAjsHZ6u9vyaFPCMlppRVa0JERERERERERERERESJImU1GSEiIiIiioWDnZ1arMVBetKxNxEREREREVGSsLN3UIvVaNiTiIiIiIiIiIiIiIiIiFIB9iQiIiIioneGvZ30JrLi+8v/2JOIiIiIiIgoSbAnUeLjnERERERERERERERERESpEHsSEREREdE7wyZzEhEREREREVGSsLOz8pxEkZyTiIiIiIiIiIiIiIiIiFIB9iQiIiIioneGg5XnJEp9bcqIiIiIiIiSDzsHe9g5WLFkpkl9M/SkviMmIiIiIiIiIiIiIiIi9iQiIiIioncH5yQiIiIiIiJ6d9nbW3dOIo015ztKptiTiIiIiIiIiIiIiIiIKBXinERERERE9M7gnERERERERETvLjsr9yQCexIRERERERERERERERFRasDh5oiIiIiIiIiIiIiIiFIhDjdHRERERO8MBzs7tVjt/WG99yYiIiIiIqLYcbi5VNiT6MyZM2jdurVZ2165cgVNmza1Sj6GDx+OTZs22Xy/5u7DFnmITZD3S5ye/zWOzRiMNw+vG62/sv4X/Ptdd5yYNQy+Hnej0zWRESrt+aUjNs2vn8cdnJszDBcWfIWgV55G6y8tGo3jP3yIS4vGIPjNyxjTbOXuuWP4/Yse2DRlBIL8fIzW//ZZR8wb2Bpbp3+DkEB/BPn7Yk7/ltHLvqW/2DS/V/Zvx4ovu2P3/IkIDwvVWxfo441Fn7TGkiEdsG/Rj4iMiMCLezdUmna5sGuDTfN7cfsy/DGmN85uXACNRqO37tX969gwoiO2fNMdF7YtU2m+LzxUmixbv+1h07zKZ+biulnYN7k/7h7YYrTe4+y/2DW6K/ZO6oeHx/7WW3dt+2L1WlsKDw7EuaWTcOjHgXh64ZDR+rv71mPfuA9weNogvLh2Sm/d+RVT1HpbCvV9jesrxuHyb1/A75Hxd9n9HQtwbkY/XFn8FQI8o77Lrq8YH53m/+SWTfN75+Z1DOjaDh/36IxHD+6b3MbfzxcftGyMa5cuxJpmCyeOHMSH7d7DiI/74o2Xl8ltHj24h7YNa8L79etY02zhj00b0L5FY3z75TAEBweb3ObE0SNo1bh+9O/Xr15Bnw86oXHNKti/5x8b5hb4de5stG7aCD/9MMnou0xr0/q1+Kj3f99bh/49gC7tWqF+jcq4ed34nicisoW5c+di8uTJZm27ePFijB071ir5qFmzJp48eWLz/Zq7D1vkIS4XA/3w0cNr+OrJbZPrN3o/V+uHPr6Jvb6vY0yzlYCwcIw+dhG9dh/HQY/nRutfB4dgxOFz6L7rKHbc91Bpy6/fQ5sdB9F7z3GcfPbKpvn1ePUGXSb9ihbf/IKztx8arf9x3U5U/OR7fL/yz+i0kb9tQKVPvkebcbNx5UHUMdjKtDt30PPcOWx/9szk+jsBARh25Qq+unYNnm+fpfzCwzH19m30Pn8eix8aH6M17TtzGY2GfIde38+Bt59/nOufe/mgfK8vUbnv1+j/w3z4+AfaNL9/PvDAh3uO4YczVxEaEWG0fu3th+i++yjGnLiIwLBw3PT2Rbu/D0Uvm+4+skk+795/gLZduqNlhy64dv2G0fpRYyegfI26alm+eq1K27R1G+o3ex/tuvbA3XumyzHWcurqbTQd8h06fzsdT195m7X+wwn/Q7meX2DjvmOwte2796Neh57oP2IM/AP078GIiAiUbdQGFZu2x6fffI+QkKj6l1Wbt6NWm27o9skIvPIyPkZre+rlgw+mLEaLMXNw+tYDo/XTN+1BpcFTMF7nu+zu05doP/FXNB71P+y7cNMm+ZQy009TJqNlk4aYO8u43uzF8+cY2LcX3mvcAL8v/FWlHT18CHWrVULzBnWxZNFvsKWQkBAM/fIr1G/aAqvXma43W/j7EtRr0hyVa9XVS/9m7HhMmjrNRjmllCjZB4koZbixdR6ylqyKwk274dIq4y+doi0+RNVPpiJ/nTa4uuG/L96Hh/6AS4ZsCAsyfkCzppvrZyBvvQ7IWq4ubm+da7S+WMfPUX7AZGQoXA73/14SY5otRESEY+e8yWjYawgy5cqPw2uN/wh1Gj0DXcb8Ape06XH6z7VI4+qObt/NU0vJ2k2Qt1RFm+U34M1rnNi4CI0/+hphIcG4svcPvfWSx45jZqPNyGnwffkMt0/sQ+a8hVSaLDkKl0S+MlVtlt+Xd6/h4ZmDqN33Kzy/dQmeV/QDFRnzFELzL2eg/sfj8fDMv3j98BYiw0KRrXApld7o86mwpcen9yHQ6zkqdv8Cd/ZvQsBL/SBn9lJVUfvzn1Cp+whc37EMoYF+Kt3v6UMEvXmFoDcvbJpfCfI4uaVD6Q6f4uqmuQgP0a9oz1u9Bap/Ng0l2nyEy2tnRKergJFGg1C/NzbN78Ndi5G+UHnkqdcFd7caPyTmrt8FJXt9j5zVW+Pen1HfHYVaf6bSsldqhvs75ts0vxNHjUCXD/uieeu2+HH8Nya3WbpgDvLmL4CgoKBY06xNCjCTR3+Fz0Z8jUJFi+PX/003ud2KhfORLUcOhEeEx5pmbd5erzH752kYM/EHhIeFYf2qFSaPaeum9XBwcIh+aO/foysaN2uBhSvWoHrN2jbL75XLl7Bj21ZM/PEnXDx/Dgf37zPaxtfXB5cvXsSzp0+jCz2ffdQXvfoOwJKVa1GgUCG8C+zePuBaa2E/IiJKzUq4uGFCrsII1hhXWItm6TOr9QOz5sEyr6cIiYw0mWYrEvDJ7OKMoRWKY/q56wgxqGhfeeM+imVMh1FVSmPe5dsIi4hEm0J5MKteFXxStiimnLkGWxq3bCsaVSiBzzs0wZC5a4zW92tRBz/064Bn3v81HBzRuTnWjfkYfZrVxlcLN9o0v/3y5UOnXLngG276GW3G3bvokDMn6mbOjLn3owIBM+/eVX9LJxQvjg65ctksr+HhERgxaznG9u2Ewnly4KdV2+JcnyV9WmycPAKrJgyFi7MTFm83fr6yFglgLrl+FyMrlUJwRAS2vQ1iar0MCsbmu48xpkoZuDs5Yuv9JyiY3h3/q1tZLSUzpUeVbJltkteR345Fp/Zt0b9XT3w+0rhM4vn0GRbNnYWNK5eizfvvqbRxk6bgx0nf4fNPP8ZXY8bDlobMWIwvurdB7fIlMX7hWrPW//BpT/R6rz68/QJsmtfAoCB8O2Umfhg1HBnSpcWcJSv11ks5ZNOiWVg1b7oqKy3f+AeCgoPxw6xf8cv336J4kYKYs2QVbG3Cyh2oU6YIhrVrhM/nGwcz+jStiR/7t8czL9/otIlr/ka7mhUwrsf7GLVkq03yefjgvzh35jR+mDYdO3f8iYsXzuutX7LoVxQvURLTZv4Ph//9FzdvXEfFylWwdPU6TPtlFtasWI5HNgx2r1i9BsHBIZj2wyRMmzETL17qN2Dfs/8AZs6ehykTv8O6FUuj089duAAfX188f2Hb+iBrsrO3t/qS2qS+Iyar8H/+CIUad0XOCvXgmMYV4SH6FY8uGbMhbc6CyFqqenRaqP8bBPu8QoYCJW3eE0OW7BUbIm/ddgj2Nm5R5polF9xzFkTmklUR+bZS0lSaLbx55qECJ0Uq10a97h/jyY1LRttkzVsI2QsWQ4FyVRERFgZ7Bwdky19YLc/uXkfRavVslt/nd6+heO2myFuqEqq17wOPmxf11kvesuQrjGwFiyN38XKICA+Do3MalZYhZ16EBAaon23lxe1LKNm4A7IXK4dSzTrjxZ3LeusdnV1UoChroVLInK8IIsPDVPrTa2dxYN44PDz7L2zJ6+4VFGvaFVmKlEOB2i3hdf+q3nont7RIn6sgMhcpC7csOaB5WxCWgFLxZt1ha973rqDYe72QtURltfg/029FlCZdRqTLVRBZi1eGo2talSafr2cXDiFPtWY2z2/gi4fIXbczMpeuDUfXdAgP1g9gp0mfFW7Z8yNj8SrA2+GsXDLnVGlp85eCe+6iNsurtHqSoMR7bTugU4/eeGmiBee927eQIWMm5MydJ9Y0W3jm+QRFi5dA7QaN8fHQL3H5wlmjbY4c2IuadRvA1dU91jRbuHLpEpq1bIVqNWph4GdDVeHB0JoVy9Cr30faWwHXr1xGvgIF0OejQepY06ZLZ7P8Sv569OqDylWroc+Aj0zmd9H8efhs6PDo348ePoj33m+Nth06qvy6uLjYLL9ERJQyudg7IJ+zS4xDb2Z0cFLrS7q4oWgaN6SxtzeZZiuXXr/BgNJFUC1HFlTKlhkPfPUreKXfbcVsmVA+a0Zkd02DSGiQ2SUNCmdIi2rZs8DdKaohiK3c8niOz9o2Qusa5ZHWJQ0CgkP01ufIlB4Fc2TVS8udJSNK5MuJJpVLw9ayp0mDrM7OJtdFaDRqaZg1K9rlzInnIVHHctbHB0MLFUIRd3dkieG11vDw2UuUL5ofTaqWw6gP2+HUtbtxrnd0dECJAnlQulBelMifW623lRvevmiSNycqZs2EXiUKqnvZ8N7N7e6KslkyolzmjCotjYMDCqVPizzurqoXnfxsC1Lx3Kdnd3Tp2B5BQYEme7QP++objP5uErze9mqpXKE8uvXpj3ETf8Cjx6Z7cFrDG78AZE6fFq3qVMEX3Vrj6v3HZq0vkDMbcmbJBFu7dusuGtSqhvo1q+GrTwfg5Dnj+qCSRQujbIliqF21EkLDwhAZGYncObOjRqXyqFXFdg2GdV1//Ayft22I96uXRaZ07vAJ0K8jzJ4xHQrmyKL/Ig1Qv1xR1C5dGOldbVMuOXv6FPr0/0iVoXr27qt+11W7bn0sWfgb+n/YA/7+fjh7+jTc3d1RrHgJVKpcBYWLFkX42zoiWzh5+iyGDf4UtWpUR7s2rXHxkn791c5du/H1l8NRr05tFC9WTKXJ53H5qjUY2L+vzfJJKVOKCBIFBASgY8eOSJ8+Pdq1a4fwGFqp6JIvxWHDhiFr1qzIkSMHvv/+e4v2Ka2DP/roI2TIkAF16tTB07etbq29Xxler2HDhmqIPTneDh06wN/fsl42luRBKhh9fX31lvj5r5Dg4OwSXZGu69S8r/DvhG4o3mqA+v3+/o0o0sy2Q3UJqTS3c3D6L8HO9Mfg5NQ+uLjgKxRs3ivWNGuTIIpTmjTqZwmmyO+G/L1fYXa/Fti/9H+o2ua/QIAElCR45OBgu+nHooI+UX/QndK4IMJguDnhefMSfhv4Hm6fOoDitf8LBNw6uhtFqzewWV5VfiPC4eCsPb+SX+Pze+/EXqwd0gqhQQHIVqQM0uXIi/fHLECd/t/g8fmjeG4QCLOmSDPye+ufNdgxojUyFSytgjCeF44gV4U6sHfSue9tld/wcDg4ReVX8h0Zbnw/XF73C/75ug0KNuigfn94eBsKNeqMpGL3tsbfXuXX+PzK8HJnf+qFfI3/+x4I9noGz0MbUOC9qO83W5C/U46O/322TbV02bR6Obr3Gxhnmi2EhYYhjYur+tnJ2dnob3lYWBiOHjyApu+3iTXNZvkNC4WLa1R+Xd1cERoaYtTT6NXLFyhbvoLe80qOnLnwQbv30bNTO1y7ctl2+Q39L78urm4IMcjv7Zs3kDV7duTImfO//Pr7w9HJCa2bNcKAXt3h6WHbIWqsPSeRNRciSnz3799HtWrVVPlr3LhxZr3Gy8sL7du3R7p06VCkSJEYhwePyYsXL9CsWTO1zx49eiA0NNQm+5Xh9fr27YtKlSohU6ZM+Oqrryx6vSV5SIzy3tRn99VwcWcDzXttaGQkfnvlgQFZcseaZi09/zmmhou75+OPsEgN0jhGPSO5ODggNEK/F1OP4gXx3cnLaLJ1P+rkyqYq2cUPp6+i8R/70L14Qavm1S8wWA0fJ0tgcCjsdMrVrmmcERJmXuPETt/PR4VB32F09/etmFvgfmCgGl5u8OW4n3EkQOSk83xqb2eHSI0GmZ2cMPv+fQy6eBF7DVrBJ7ZdJ86r4eL6TJqL0PBwuLwNSkmvoDCDZ9HY1u84ehZhERFoVv2/5z5r2HDnkRombtalmwiLjEQah//uXfldV3ZXF+R0c0Gz7fvx+/W7aFvwvwZge588R/3c2WAr9jrX2dnJWZVTdE2b/D2W/joXbVu1xPCvR6u01UsX4cS/e/HPts14ZcMhpeU6y2dLt9xnyXpbk6CP69uGXG6uLggx8Xfq9v2HKFX/fcxduhq9O7eDu5sbGtSshgLVGmPouB/wSa9uVs9nYEioGj5OFr+gYHXutOfPLY2TWd9lIzo2QbPRc1B8wAQMbKk/VJq1hOqUodykzPd2uD6t+g0b4cyV6/hrz35UqVYtOiC0dtVKVC5TQj0/FC5iu4aioaqMqr0fXNXfeF3+AQF48PARatZvhMHDR6j16zZuQo8PusLB3raNHmw1J5E1l9QmRQSJLl26hPHjx8PDw0MFTPbti7uL77p163Du3DlcvXoVp06dwsaNG3HgwAGz9ymvv3fvHu7cuYNly5bh5MmTZr8uIfsV58+fx3fffafGw5aH/lmzLJtDxJI8TJ06VX2paZd8+fIhPpxc3eH39IEaNi7w9VM4u6c32qbsB1+oIeeurJuh5il58O8mHJk2EHd2LsP1LfONehhYi72TM8KD/BEW4KvmI7KPIYBStv8klOw+CjfWTY81zdoyZM+Fp3euITw0FA8unkLmXMbXyDV9RnT/fj5qd+6PPYt+ik4/t3MjKjaPqni3lUy5C+DxlTOqtcLDSyfV74ayFSiGTuPmomCFWji1ZaneXEal6re0aX4z5MynegUJz6un1e+G8pavgRajZqlecneP/QMHRyfVuyh70bLIUqAYgvxsN8Zv2hz58OLGOfXzi+tnkC6HcX4L1GqJel/8Au+HN+B1/zru7FuPC2tm4uD0IXh66ZjJuYysxT1HXry6dU4Ft7zuXIZbVuPeK0Xf+xA1h87E/X83I8TPG/f2bcCp+aNwYfkUPDi4Fc8v227cZ+k9FPDsPsJDAhH82hNObhmMtpHh5Ur1noT7b4eb8/e4hUd7lqFQm8FwcHa1XV4dHREYGKDm6fF8/Ch6yDOtu7duYtuGtejQqDa2rl+NUYMH4sqFc0ZpAf5RQxJaW45cuXHz2hWEBAfj7IljyJNPvyWm9Bj6+49NaFW3qpq7aEifbibTbKVgocI4eeyo+i47dvggChUuord+8/q12LhmFRrVqIxrV69g8Ed9UahIETy8fx/fjp+I9p26YNbPthvzWfJ3/Mjh6B5ChQ3yu3TxQiycP1fNPSTzJk2eMBaFixbD82dPMWX6TJQtXxHLFtt2TG0iIl0XL17EH3/8gevXr2Pp0qXw9o77+UrKTLly5VIN+tasWYPBgwer4Im5Jk2ahNKlS8PT01M1EJTykzkSul8h5bXt27erfUpZzZwybnzykBjlvf5Z8qjh4kq5xN2r1y8iHDNePESHDNlU76GY0qxpau0Kari4vGldkT+dG84891IV7BdfeSNvWje9baedvaaGopvToAqOPXuFR2+HkvqoTBEsaFhNDUfnY1BpmJjcXZyxfszHanFN44T0bi648fipanX/6MVrZE5nXk/qGR93xboxgzDiV+vO7ZrbxQVTSpbEt0XjrhR1treHf3g4fMPC1HxEjnZ2KlAkjS2aZsuGIYUKYfEj686ZU6d8STVc3KRB3ZAve1ZcuP0AIaFhOHjhOgrlzq63bUzrF/6xBzcfemJkj7awtub5cqqh4noUK4D86dxx5oWXehY99eK1+l3X+ZfeeBEUjFl1q6BL0fyYf+W23lxG7+W3fkBWy97OHo+feKihrwICA/UasoncuXKiVMkSaNG0Cfz8/PTSt/65A4VtOORx1gzpcM/jOfwDg3HpzgNkTp/OovW2Vjh/Phw/e0E1sDtw7CSKFjKuXymQJzc2L56D3p3b4/uZ83Djzj0cPnUWm3+fjXHDP8PXP/xs9Xy6Ojth3bcD1OKexhkZ3V1x9eFTFTC69/QVsqaP+7tsxMLNWDi0BzaOGYhZf+yHf5B+AMQaJMBz7G0Z6sihgyhcRL8MJaTnkKurK7Zs3IjyFSuptPdatcLqDZvh4+ODA/v2wlaKFi6MQ4ejyqiHjhxFUYP8FitaRK2bP/sX1Sjknz378Oui3zHwsyHo+mEfbNryB9ZusO2wpJRy2K57QQLUqFEDFStGdZGsVasWnsUwKaKuK1euoEuXLqonjWjTpo1Ka9SokVn7lIf1zp07I1u2bGqRFmbmSOh+RZUqVVC1atScLB9++KEKUkkPKnNZkodvv/0WI0aMiP5dvkTiU3Ao0qIXjv38iZpDpHibqBbqN/9crIafy5C/BE7O+RKBrzwQFuiH7GVqqd5Gdb6OqozyOLUbji7ucLVBqzKt/I274tjEqMrG0h+OUf9eWz0VRdt9Cue0GXFiSm9oIsIRFuiLvPU7qfWm0mxB5hcqUbMxZvSoDzs7e3w4ZZFKXzP+U/SYuAC+r55j+dd9odFEItjfF80Hfq3WhwT649Xj+8hRqARsKWu+InBNlwFzezVQPYl6TFuByIgI/DH1C3QcOxse1y/g71ljVY+ukKAAtP06KuDm5fEAadzTwSWtcaW8NeWrUBsXty3D6k+bwzVDFlTu9DF8X3jgyt9r1DxFEhQ6u+k31SNGAh0V2vRWPYvObJiPsKAApMuWG9W6f26z/EoA6MDUQbi5axUyFSiBLEXL4cWNs3jz8CaKt+iBGztX4sHhP1UPLsc0LnDPlhtV+45BeGgwgrxf4tbuNchbrYnN8lu4URccnfm5GuIxT9UmcMmQBY+O/616QeWu0hiX1vyMl9dPqyEq3bPlUfMX1fj8Z3V/eN27gjcPriFzMeu22NOVp2E3XFn4herqnq9JL9U75/G+VchYvCrS5SuJa8vHIvjVExVozlisinrN7Q3TEBEWgotzPkGajDlQZoDtAgN9Px6C9+tG5WPyzKig1fgvh2LoN2ORr0BBrNy2S6UtmT8LNes1RNESpYzSXN1sM4ybq5sbmrVqi0aVS6nWhgtWRVViDO7TDfOWr0P12nWxdGPUxKWTx4zE0K/HoljJUkZptiIBFOkVVKVUEbi4uGLjjqjz1r9nVyxZvQGdu/VAwyZRzwZDBvbDtxMmIneevKhQqTIG9emJ4OAgjJtkuznL6jdqjDm//IxKpYoiS9ZsGLpzDzyePMb82f/DDz/NwPCRo9B/4Mdq208/6oeBnw5B9hw58OvcWRjQqwdCQoKx4PfleBc42EUtVnt/6701UarWtm1b5MkT1ZikZMmSqpeP9LKJjZRxfvzxR6RNm1aVGSXgc/fuXWTOnNnsMt/PP/+sKoAaN26M4sWLm/W6hO5XyMgR2nKXHLvkxc1NP4CRGHlIjPJeDqf/hgTzCA3GhKf38CI8VPUukuDRq/AwPAwNQvuM2bHa6xkuBvmrbcSPeYpik/cLo7SsjtYbZky3Mr1n8YIYuP+U6tXSqmBuZHJxxta7j5He2QlN8uVEhawZVa8hNydHuDo6qPmLJp26oirlA8MiUCCdG9JasTe+PBPJUHFaX3ZpgaZfz1SVfGN7tlZpk1fvQKsa5VGpaH6s2HMcP677G97+gbj95Dn2/vQl2k+YhwfPX+GNfyCaVykDa5KhAgvo3KdLHj3CjufPEabR4F5AAMaXKIGpt2/j04IFkdHJCV1z50a3c1EN3Ma8HfqoZ968mHTrlhourXYcn/GESufmqoaL02pVpzIKdx6sghrbp49SaZ2+/Rmbp45EWjcXo/UPnr7El3NWqIDR8p3/4sMW9fH1h+bXzVgqo1Suv+3Fks016vfG2/bDxdEByxrXQESkBl8eO68CSTLU3HVvX4w4eg7+YeEY9Xa4wQd+AUjr5KjucVsZOfxzVK/fRN23P0+dpNJGjZ2Agf36oGiRwmjWpgOePX+B16+98L+fop6PR0+YiC3bd6g5Ures1Z9nx5rkMzegbRMU6zJY/b5uctT3Y7exM7FiwlA4OzmaXP/13BVYv+coNNDg0p2HmDvyI5vkN2f2rKhWsRzyV2sMRwcH7F73u0rv0H8Iti6ZqwJCnQcOUyMKefv4Yv7U8ciZLSs8nj7HBx9/AV//AIz/4jOr51N6DZXIG1UHKb7s1BTvjZ0jVYT4pmtzdd6nbdiNJhVLoGrxAlh94BSmrPsHXn4BaDTqFxyY9gUqFc2HfjNXwNnREQWyZ1ZBdGt7v01b/DZvDjasXY38BQpi6s+/4Orly2qu11Fjx+PG9Wv4qHdP1SDjw959UbFSZaxavhS/zp2N0NAwddzjJ/4AW+nfpzeatGyNCZN+QIP6dVGyRHH8tXMXnr14gQF9eqPPhz3Qsl1HrNuwSTUi/XHyJCxb9JvqMXXz1m1s3LIFLZs3x7sgat4gK5bM7FNEv5pEZacxNVhoMiLDr0krqR07dqjfx44di6JFi6ru+aYelIcPH469e/eq1lS//fab6kUjf3TkIVx65MjDvzlWr16tWrBJrxyJDDdt2hTTp09XgSNr7leOt0mTJqo1WbFixfD555+rfyVIpN2HKYmVByk0SAuzFr/sUb2DLBERGqwqprW9iEJ8vVTwR4aYCnzlqYZtkgCMs0EQQAJHMuSbpfsTQWGmJ0w1h+zXzsEBjmmiHnJlbqI0GbKqL5mA54/Ul71zusxwfJsvU2mWqlci/l2+A3y8VcDI8W0XeO+nT5ApV141vJuX52M110+6LNnh/HY4p7CQYBUoSptJf8xqSwSFxv/8+nu/hFv6TNE9tXyeeyBDjjwICw6C78uncHByQrosOeDwttAZEuCnAl0JCRJ5+ce/pUmg90u4ZsyqrrGc02DfN3DPnE3lK+jNK3Ufu2fJAXt7B4QG+iHQ+5W6Z90z67dAs8SbwPiNXauJjESwr5fKr5AAS0RIMNKkz4QQX2+EBPioXk+umbLrdZGXQJf01HHNFL/7MDQ8fpMMy34luJomfVSlRWiArwp4yvxJQd4vVP6dXNOqAJKu8JBg9b0iQ+bFxyuvwHi9TvYpw+I5uUV9l4X6ecEhjZsKbgd7PVXBQlnn5B51r6q0t93O5X53iUfAe2GfqEBPfPj5+qiHbve0Ua3cnnl6IGv2HHot+LxevYKbu3t0V/qY0sylOxSKpaTnk+w3zdthE548eoC8+fWHcnn+1FMdg27vKFNp5krrHP8HvBfPnyNzlizR5/PRwweqAKHryeNHyKvTM+rVy5dIkyYN0qU37lVrjoQEOJ49faqGlJPPvgzXJ3nJlTt3rPmV18jffgnkWcrP1xclC+ZRz0oyTG5S0j7DTHAtDJcYhpJNDMGaSHwfdC9ZHDPRu0KGX3vz5o0q6wkpe0maBIsMLV68GA8ePMDkyZMxdOhQNazRtGnTVJBFGsfduHHD7GCNlLXk+13e6/Tp02jRooUKtOTNm9eq+5Vjk/eTMq78XZMgkQR8ZMg97T5MSYw8aL8rNxcuD/d4VK6EaSLxTGc4aQkgSQAmVKNBBgdHeIeHwT/yvzJELqc0qieRYZr0KjFXnuq5LM6nXp4jIuEXFqbmGhLSM0h6tKR7W4nuExqm5m+Robsk/VlgEILCI5DOyQlZXaNeY4kiPaKCO/Elw86FhIcj09teT8+9fVUPIxkCSypUX76J6onh6GCPIrmz48GzVwgNj0CW9O7IEo85aM5OXR3vvHqFhsLv7bBsMiRajjRp1NxDMk+RdnhWWS8/u+k8wwVGRCAoIiJecxJVmxJVeR9fr974quBRmrfX//7TFyiUK7vJ9TLknPQq0cqUPi2yZ7KsvHpzfsICIK+CQpAxjRMc31ZYegQEIo971L0RHhmJ50HByODsFB3M9AsNUwG4+AaJSiz4b6QPS/j5+SNSE4kMb5+NPJ8+Q9YsmeHs7Iy79x+oMqz0HNIGw58+e47AoCAUzJ8vXs/3wu7qQcSXl6+/GlbQ7e33gsxJlT9HVH2AqfWer7zg+3ZenbSuLsib3WA+HTPYF4h/EPfFq9fIlCE9nN5e5/uPnqBQ/rwICQnF/cdP4OToiDw5c8DlbX7l78OTp8+RIV1aZMwQv+fV0HN7EzwEXXBoWHSPyOdv/JDONQ3c0jirQPcL3e+yXFF1FC99/FVvvrzZLA8g+zWMX+BOqsWlzKcdljs4OBi+Pj6qQZ38/PjRQ2TPngMZMkbVS3h7e6kylqurG3LnyaM33KIlMiI4Xq+TXmUyRGPOtx0DfH39EB4Rjsxvg+4SMPTwfIpcOfXrBGTouTc+PsiR3fI6LNlH9vyFkkX5R/sck7HJt7BztF7vZE14MN7sm5osjtlWUkRPovjo1q0bjh8/jlKlolosf/bZZ2YHarSvl0CNjO9ctmxZVK9e3Sb7FTI2tQyvd+jQIRUwkvmFpDBgSd4Tmof4kApUWbS0FcLCLWvMlabScyApGO7XJdN/rR7ccxhPSGkqzZbcM+j/kZQAkZChz7LlL2y0vfTikSWppDUIREiASDi5uCJLPuP8Si+ipOSmk185pxIg0ubLMG/ObunUkpQtJrQBIiEBIVmEBIpkMcXe0THeAaKEkP3qfh/oDkcpgayYSE8oWWzN8LtMAsNaLpmNKylMpdlSuvT6BdWcuY2H9MucNatZabaQKYt+gcowQKQdms6cNFuQwoEuwwCR0A24iKzZbP8508qZ67/7UQqRhgEiU/nVfc2705PIel2J2JOIKPmQMlP//v2RM2dOZM+eXQVfLOnNI3Mf9ezZUw3bJg3rypQpY5P9ChkpQwI7EhgaOHCgKvdJEMhciZGH+HCyszceMs4O0DY5yeTohEzQr5w2lWZLTg72yOzwX7Anw9ueGtG/OzupRSunm+2GDzbFzcUZbvgvjzky/ffsLJWthkPQFcyZNM90Kj/OzmrRJYEiXekMhh4TEjDSDRrZUtaM+pV9ugEiw/VS8a7bEykpGAYqtQEiIYEj3d+FNvhpa+nS6QcoJSCkVaSQ8fOzVGAnpcwGAdUCObPFuj531szInXQfNWTPql+GkgCRSJPGGSWLGtevSOCtQN6kKT9pSTBIFq0cGf+rQ5EguDYQritbBssD3QklgUHdeVtdXFzUov25WHH90XkyZcqslqQigR9tgEikNxgSUeqB8+U1/t6SRozxCRAlV3Z2Vp43yC71lfqSfU+i1Maw55StJaQnUVJISE+ipJCQnkRJISE9iZJCQnoSJYX49iRKKvHtSZRU4tuTKCkkpCdRUkhIT6KkkJCeREnBmkOlJbbk2JNoopv0JLLeQ32wJgLjA9mTiIgSr+eULSW0J1FSSGhPIltLaE8iW0tIT6KkkNCeRLaW0J5EthbfnkRJISE9iZJCQnoSJYWE9iSytfj2JEoq8e1JlBSSY0+iTE3Hws7Jij2JwoLhvXdysjhmW0lZtSY6gRSJ9OouhhPjmfs6ia6b+15Jtd/EyAMRERFRaqCdk8iaCxHZJpBiWO6pW7duvF4no0OY+15Jtd/EyAMRERFRqiD16lZckES9XZNSiowuVK1aVY0ZaavX2Xq/sfUiSugxEBERERERJXdDhgxRi61el9T7tdZ7EREREZH1BQUFqc4dXl5eKFGihMm5NZOzFBkkIiIiIiIyReYjsu6cROxKRERERERElFRkPiJrzklkZ+F7r1u3Tg0jXKBAAbi7u+PgwYNo2LAhNmzYoOaDSgkYJCIiIiIiIiIiIiIiIrKQzFt09uxZNV+SePz4MYoWLYpVq1ZhwIABSAkYJCIiIiKid4a9lecNSpETehIREREREb0jkltPovfff1/v91y5csHNzQ2BgYFIKRgkIiIiIiIiIiIiIiIiesvX1xe6ZOi4mIaPk7mI/v77b/j7+2Pz5s2oVq0a+vbti5SCQSIiIiIiemdwTiIiIiIiIqJ3l729g1qstwMH9U++fPn0kidMmIDvvvvO5Et8fHywa9cueHt749KlS+jVq1eKmY9IMEhERERERERERERERET0lswtJPMNacUW9ClUqJCag0g8efIElStXRqZMmTBmzBikBAwSEREREdE7w8HKcxJZsb0aERERERERxcHO3t7KcxLZq38lQKQbJDJX3rx5Ua9ePRw9ehQpBefeJSIiIiIiIiIiIiIispCHh4fe7yEhIWrIufz58yOlYE8iIiIiInpnsCcRERERERHRu0t6EVm3J5GDRdu3adMGNWvWRPny5REQEIB169apQNG3336LlIJBIiIiIiIiK3r27Bl+/PFHXLlyBTlz5sSQIUNUISImMhnq3r17jdKzZcuGrVu38loRERERERElEydOnMDatWtx6tQpNW/Rp59+ih49esDFxQUpBYNERERERPTOcLCzU4vV3h+Wvbevry9q1aqFkiVLYujQoTh8+DDq16+PAwcOoE6dOiZf0717dzRt2jT6d41Gg/bt26Nw4cIJzj8REREREVFKltx6Ejk7O6NPnz5qSakYJCIiIiIispIFCxaoQJH0AJKWZG3btsWtW7cwbtw47N+/3+RrSpQooRYtCSh5eXlh4MCBvE5ERERERESUqBgkIiIiIqJ3hsPbeYms9v4ay7bfs2cPmjVrpjfUQLt27TBo0CA1TrUMRxCXxYsXq55I9erVi0+WiYiIiIiI3hnJrSfRu8A+qTNARERERJTSSO8g3UUCPqY8fPgQefLk0UuT3yMiIuDp6Rnnft68eYMtW7awFxERERERERFZBYNERERERPTOsH87J5G1Fnl/kS9fPmTIkCF6mTp1qsn8hIWFGU1Y6urqqv4NDQ2N83hWr16NyMhI9O7dO1HODxERERERUUpmZxfVk8hqi13q60nE4ebIpIhIDewjLRxPheLkHxyW4u4Dsh4HeyuOh2QFKe1+sHdIOe0gQsJT1rl1TDmnVgmNSFnnNwXdugjVpKxzm5geP36M9OnTR/8e07BxmTNnxuvXr/XStL/Lurj8/vvv6NChA7JmzZrgPBMRaeWukgNpnZxSxAmZu+YqUpJZH3+IlCRN+riHPU1O7BxTxn2r5eCcgh7shF0Kym8KuxdS1LlNpcNt2ZR9CrofUljdFcUPg0RERERE9M6Q+YisOifR2/eWAJFukCgmlSpVwtmzZ/XSTp06pYacy5YtW6yvPX/+vFqmT5+esEwTERERERG9I+wcHNRizfdPbVJQ2JKIiIiIKGXp37+/ChL9/fff6vcnT55g2bJlKl3r8OHDqFu3Lp49e2bUi6hw4cJo3LixzfNNREREREREqQN7EhERERHRO0M7d5A1398SderUwYwZM9CpUycUKlQIDx48QKtWrTBmzBi94eeOHj2K4ODg6DT5ec2aNfjqq69gZ8XjISIiIiIiSkns7O2tOiSiXUoaDjCRMEhERERERGRFX3zxBfr164fbt28jZ86cyJcvn976+vXrq95EuXLlik4LDw/H9u3bUaFCBV4bIiIiIiIishoGiYiIiIjonWGrOYkslTFjRlSrVs3kusyZM6vh5nSlTZvWKI2IiIiIiCi1k15E1u1J5IDUJvX1nSIiIiIiIiIiIiIiIiL2JCIiIiKid0dym5OIiIiIiIiIEg97EiU+9iQiIiIiIiIiIiIiIiJKhRgkIiIiIiIiIiIiIiIiSoUckzoDRERERESJxd7OTi3WYs33JiIiIiIiotjZ29upxWrsU1+Zjz2JiIiIiIiIiIiIiIiIUiH2JCIiIiKid4adgx3srNjyy449iYiIiIiIiJKMlPesWuazZ08iIiIiIiIiIiIiIiIiSgXYk4iIiIiI3hn2DtYdn5pzEhERERERESUdGd3BmiM82KXC0SM4JxEREREREREREREREVEqxJ5ERERERPTucLCHnb0V20HZaaz33kRERERERBTnnEHWHD1CwzmJkp+QkBA8ffrUrG1DQ0Ph6elplXy8evUKAQEBNt+vufuI6TzZIm9ERERERETx5ePjA29vb7O29fX1hZeXl1VO9pMnTxAREWHz/Zq7j5jOky3yRkRERETvrmQ/3Nzly5cxcOBAs7a9desWevfubZV8TJ48GTt37rT5fs3dR0zn6eHDh/j2229hbX6e93B4Uk/sH90eLy4fNVp/fvE47BpSH/+O64LXN89Gp0eEBuPgd93geXo3bMnr5hkcHd8ZxyZ2h++jm0brT/30EfYPb4STU/vA3/NejGm2cmXfNszv3RArR3SDzwvjoN+c7rXxS8fKWDuqNwK8X6m0Z3euYvmwzpjZsRJObV5i0/yeWP8bFvSqj80TPkZIgJ/euoA3r/G/jpUwp2t1/DF5CMJDglX6/TOHsfTT1pjVqTJuH99rs7xqNBocXjgJKz5qhP1zxiAyIlxv/bObF7C4RzUs61sXhxZOUtu/8Xyg0mRZ3r8+bEnO1+FZI7FlSDNcWD/baP39o39hff9a2DK4Ka5s+11v3ellU3H813E2zC0Q4uuFo9M/we6v3sf9A5uM1t/8czH++qwedn/VCo+P/aW37sT/hqr1thTw/AHO/twPJyd1wesrR4zW31gzGUdGNcWpKd3hfeuMSjs/6+PotNfXjts0v6ePHUarWuXRrl4VXDn/33errpfPn6FJxWK4cPqE+r1322aoXigbujSphVvXr9g0vzs2r0fjSiXwwXsN8OTRA5PbXLt8ATWK5cHrly/U7/XLFkaVQtnRq21zvHhmXqORxLJg1gxUK1UIvTu1hc8b05WXO//8A9VLF47+vUSuTNHLs6e2baQxduQXqFSsAIYO7IewsDCT2yycOxvtmjVUPz+8fw9Fc2RSS5kCufAutSqzc7DikgpblRFZ28qVKzFv3jyztt2wYQNmzpxplXx07tw5xgaK1tyvufuI6Tz9888/WL58OWwhMDwcQ/49g3qb9+B/F24YrfcJDUPfvSdQe+NufH/qst668ScuYczxi7Cly/DFQjzEn3hucr0PwrAJnliGxzgHH5V2Ct7qNSvwGLfgb9P8Xr33GJX6fIUinYfg72PnzFpffcC3SNu4F0p2G4Z/TlywaX5PvfJC2wNH0engcVz38TVaP+/mXTTdewjdD5/EXb+oc/n56QuoteuAWv7ysM2znbevP94bNhF5WvXHTyu2WrT+g9E/Y8T/bFue1vr10m3U3bAbA/eehF+o8bPdT2euoda6f9Bz11G8CgpRaRVW/a2WG17G18NaTp89hzJVaqB4+So4eMS4Puj5ixdo0/kDZMpdAN1691Npsm267HlQrlotHD5q2zLU9kOnUKjNR6jUcxiuP3hi1vp6H42CW93Oavnz8Cmb5nfestXIXbEuGnT8EM9fRtX3aEnDBpeC5ZC2aEU079YP3m98EBwcgg8+Ho4spaujRff+8PJ+A1u7/ugpqg6ZgiJ9x+LPE5eM1n/z+xakaz8Mn8xeHZ02ctEmlSbLhJV/2iSf4eHh+Gxgf5QqlB8jhw1R9T26bt+6iWb166BkwXwY/+0ovXVTJn6Hbh3bwZZ8fHzRql1H5ClYFFOmTTdaL/kfNXoschcoggzZosp36zZsgnumbGp5r017vFNzEtlbcbFLfWW+ZB8kooQpVqyYTQoMt7b/hhLtB6Pa57/g5h/zjdZX7P89ms/aj0oDJ+HWn4ui0+/vW49sZWpCE2nboVvubFuA8oN+RMkPvsTd7b8ara828jc0nLEHhVr2x4M9q2JMs4Xw0BCc2rIEPaevRoUWnXFqk3Gl+eBVh/H5+hMoWrMxLuxcr/4w/DXjG1Rp2wvD1p9CtY5RD2K24PvyKe6eOoA+c7YiR9HSuLxns95694xZMHTjaXy68hDSZ8uFW8f2ICTQH7vnjEPjT8bi8w2nULRmE5vl9+m1Mwjy8ULXX7bKX1Q8OndYb33OEhXRf+Vx9Ji/C4HeL/Hi9iW1XammHVV6r8UHYEsPj++CW5acaDV1I17duYw3T+7qrS9Y+310WXQY70/diCdnDyDYN6pi+/XdK3BOmwGayEib5vfe3rXIVakh6o1ejocHtyAsUD9oWLxVf7SccwB1Ry3GnX9WRqd7nNqNDAVKQhNpujWvtTzctQQF3xuAcoN+xv2dxp+1Et1Go86Uf1C6z0Q8/CeqsFjx8/kqrWSPMXhyYI1N8zvrhwn4eeFKjP9pNmZP/c7kNit/m4v6Td+LfuBdunUXjt7yxOCvx+H32dat7NIVFhqKxXNmYsW2f9C930Asmj3DaBvJ46aVy1CtVt3o/B64eBsnbj5B8zbtsW6Z7YKGEuDZuf0P7Dx4EpWrVcfaFUuNtgkOCsKJI4eQr0DB6LTqtevi6uOXasmZK7fN8nvq+DF4PnmM/SfPwcHREXt26gddxfOnT+Hn5wP7t0OxyTnu2rM3bni8xKV7xgVkIiJKGbp06YIvvvjCJvv6874H8qdzw/ZWDXD6uRfu+egHUf5+4IEq2TLjn3aN8NgvEHfeRD37nX/phcwuzogwqICztrJIh37IBw1M7/cq/FAW6fEBcuM2/BGOSFRDRnyE/GiPnDj/NnBkKxMWb8CUT3pgx8/fYMxva81af3zhD3izexkWf/sppq/ZbtP8zr15Fz9XKY9RZUqogJCu4IgI7H/2AivrVMOXpYtj/q2ohpaR0kiveQO1vJ87p03yueiPPahTviTOrZyJjfuOwuPla7PW7zx2DkXz5UREhG3LUOJpQBD2P36ObW0aoEyWDNh0+5Heek//QFx69Qa7OjRGgzw5sO1u1LPc2R7voWPRfPKkZ7O8jvluEn6bMwvrVizB6PHfG63/dvz3KFGsGB7fuoo1y6IaMl4/fwreng8xbdL3mDVvAWxp/K+rsXPOd5gwsDu+X7jGrPXurq7w/Xe9WlrXrWazvL7x8cWiletxeudmtGveGLMWr9Bb7+DgAP875/HqyknUqlIJyzf+gdMXLiM4JATXD+1Ey8YNMH+5bcuoYuLqvzDhw9b4a9IQkwGfKf3a4+TsbxCpUz8hnzNJ8948ExN6trJJPnfv/Bv2dvY4evY8nj17hmNH9OuDNq5bi7YdOuLImXN48vgRzp2Naih65/Ytde5j6n1sLb8vW4ZqVavg/Onj2L7jLzx6rF+G27z1Dxw6chRnThyB17OodXKO583+BT4vn+KvP/Tr54hSbJDoxYsXel8g5pDu+NL9Pr5evnxp8T4Tsl/dYePkeBN6nmw13FzIm1fIXq420uUqhDTpsyIsSH9oPpkbwN7BEQ5pXOHkmlalBb5+CjsHB7hnywtbkp4i9o7OSJe3KDKXqIqwIOPWYXb2DtH5dXybX1NptuD7whM5i5RGhhx5UKZJO7x8cNtoG3sHBzg4OsEpjSvSuKeDv9cLdc7LNmkHBycnm0bAXz28jcLVGsAtYxaUbtQWrx7cMpFfRzg4OcMxjYvK78t7N5CzeDkUqFBTrbNlfr0kv7WawSVdRhSp8x68Hpk6v45wcE4DR5379/ahv7ByUBMc/X2qUWsTa3rz+DYK1noPadJlRL6qjdXvuuTcafPr5Oqurr/k78GxnShS37atXISvx13krdkSrpmyIUuJKvB//jiG7wYXOLunV2nhIUHweXQTWUva7uFbK+TNC2QpWxduOQrAJWN2hAX6GuVXvrccnOV7IN3bNAeVJjIWq2qzvMr3vHz2S5WrgCq16iDQxJCoF8+eQvHSZZE+Q8boNHmYdXJyglvatEiXIeqc28IzTw+Vl7z5C6JN5+64e9O49fHfWzeiVaeucHB00Muvo5MTXN3ckS5DBpvl99b1a2jUrDmyZs+O9l274+Y1415Xyxf9io8+G6qXdv7MKVQtURCf9etp8ppYy41rV9CqfQdkypwF7Tt3xY2rxvld8tt8DBoyTC9t2+YNqFKiMEaPGBav553kyN7BzuoLEVmHVLbEpwz0/PlzVY6KD3lOkjJffMR3v9ph4+R4X7/Wr6yOz3my5XBzt7z90KpQHmRycUaz/Dlx620QSCujszOc7O3g7GAPJ3t7pHV2VEGBvx88RYfCUnFtW3Zv/4uJC+zhADu1pFH/j/rPHnZwhL1Ks6Wnr7zRslYllCqYFzmzZIRvQGCc6x0c7OHo6KACYU2rlbdZXsMjI+Fsb4/i6dOhetbM8AvXH5HBxcEBVbJkQpdDJ7Hy3kO8DP7vs/Le/iPoe+wM7vrb5lnp8t2H6NGiPnJkzoj361TBNYPGMabWh4aFYffJC+jYsBaSwi1vXzTMmx1ZXNOgbeG8Rp81NydHuDjYI436rNkhQxonle5ozXkZYyBTM9StXROVK1ZQz+6Gz5RHj5/EmFEjkTZt2ujGSuoZ39FRbdu0UQOb5fW1jx/y5ciKkgXzon3Dmnj07KVZ6/0Dg5Dn/b5o+PFo3H3yzGb5vXnvPhrUro48uXKgd9cOuHzdeCQcOY9p0jjD3c0VGdOnQ+XypREYFIyC1Rvh/JVruHjVuNxlbY9feqNNzfIomS8n8mbNBC8//c+63AcOJnrnNxn1Cyp+OhkHLhrXI1nDtatX0K5TZ2SWMlTHzrhuUIZq3bY9Vi1fhuoVyqq6gKuXo3pFLVu8CIM+Gwxbu3zlGnp064rs2bKhVcv3cO3adb31Bw8fwbAhg5E7Vy71GdP6+tsxKFyiDH5daNuRWqzJqr2I7FPn6BEpIkgkD+3Vq1dH+fLlUbFiRQQG6j8omSIP661atULhwoWRL18+9OrVy6KK3KCgIDRs2BAlS5ZEiRIlcP26/gfPWvuVYePatWunjrNcuXLqmKXwEd/zFNcwdZJfKVToLvGiW6mvfjQ+5iNT+uDID31QoFEX9fujg1tQ8O3PSSmmQsO/XzbHlWXfIW/ddrGm2UQcQRO/188xs0NFnPljOUrUbYHIiAi4ZciMJYPbYUHfxrh2wDZddbWigzx2qtONkSdXzmBW58rwvHERectWVYE7Ozt7LBrQFEs+fh+PLp20aX7f3rQq36Y+rzcP/IElvWqqrTLkyo+MeQqh9+//4oNZ29Vweh6XTtg4u9rzKyfYePWlTfOx+ZOGyFKojApqPTj6FwrVax3nfWQ1b/cbdV8YZ/jsonHY81UrFUwSDw5sROGm3ZAk9L7LTN/AMrzc2Rn9kLtux+g0v0fX8fLSQeRv+qGtcqruVd2AqmFwVQpce/78A606fWD02vplCuCrj3ujS68BsCVtHk0FggMD/HH31k1UqVFbL12GnauUPwsWz5mBlu06ISn+Ppj6bvB4/EgFr3R7EYkLdz1w7NJNFC1WAhvX6Lf0s+V3g2GL6bOnTqJchYpIl+6/wGDBwkVw8e5jHDp7EUGBATi4b49t80tEpOPMmTMoXry4Knv17NnTrHPz4MEDlC1bFhUqVEC2bNksHg7u7t27KFq0qCo/1a1b1+yyUEL3K8PG9enTR5UZixQpgg4dOpgdqDd1nmIbpi4xynu99xxHxbU7sfdxVMWo9q941L/6f29q586Gf2SYofX/qJ5D2V1dsO3eE3Qums9mj6IyTJwMF/caoXFuWxTuOA4vLMVj5IWLChOJ3XiJlXiCEnC3al59/APh3vhDtQQEBRsUq+X5Q3/7mNafuHILO46cxdc921o1vzKknAwT13q/8ZBipi7v6LIlcaRFQ3xfoTT83g6FO6daRexsXBefFC+MBW97F9n0OVTOm4kyieH6hX/sweAuUeUTW1l4+Y4aKm7csYsGeTIulmRwdkK+dO6ounYXNt5+hCb5bNMryxT9MklUOUVX5kwZ8dGnnyNHgaIYM2FidPr+g4dw+tx5fDIwacok6mcTd66p9QcXTsGjHUswpGsrTFmywUY51c9PTPUVV2/eVkPOrdm6A62bNYK7mxv2rF+KwHuX8GHHtiqQbG0BwSHRQ8X5BASZU6w2MmNQZzxaORUrR/XHuBXbk8X5LV+xIk5euIx7Hs+RN28+ONg7qN5HzVu+D2cnZ5vl0dwyqgyfd+DgQWTLk18NLffmjQ+6f9AFng/u4ujBfVi+ajVexaNxCqUOKSJI5OHhgb///lt1/ZMH4r17456vRIZYc3Z2VoETed39+/fx55/mV5SvWLEC2bNnVy20zp07Z3ZvnITuVzuP0K5du1RwqGXLlvjf//5ntfM0depUZMiQIXqRwFZ8OKfLjFc3ziDwlQeC37yK7m2hq843S1BvzApc2/CL6ml0f99a7PmiGa5tnIVLKybD94lxDw5rkF4LkWEhCHj2EG/uXYKDi6vJ7er/9DcqfjId19dOjzXN2tJnz63mF5K5hm4c3oWs+YsYbZMuSw4M33QGdXsNxcElM5AuS3aE+Pui47i5aPfNLzj753/jvFpblnxFcP/MITWE3M1DO5G1QFGjbfKUqYIh60+pHkdnty5D5nxFEOjjhe7T16BOr6G4tGujzfKbSfJ7ah/CgoNw78QeZM5vnN/iDduh79IjcE6bHneP7oq+jxydnNWwThqN7VrfZ8hbBI9O7UVYcCCenP0XGfMZ3w/lOn6CjvP2wcfzHl7evoh7h//Enon98deoTnh0ei+u/mk8bJa1pMtdGJ6n9yDU3wevb52Fe3bj75jKA75H02nb8eDgVjWHkcxdtG90R5yc/QXu7FoBzzP7bJZf6QnpffMUgr2eItj7GZzcjXuuyPByVb5cirt/RM0J9fraMTw/tweFW3+qehXZirQMCg8Lw71bN3D5/BmkcXHRW3/n5jWsW/qbmn9o1aJ5GNS1Dfx8o4ZM2X/xLuat3IzJo4bbLL85c+fBzauX1RxJu7ZvQeFixfXWHzu4H0vm/aICQv/u3okuzeup9CzZsuP0vef4ctwkTP9utM3yW7RESRzcv0edsx1bN6F4qTJ66//6Ywt+mjhOzT108ewZfNC6eXRLPidnZzinSWPTnjnFS5bGPzv+VL2X/t62FSUM8rtp7WoM/+QjNf/QhbNn8OXgj6PzmyaNi8rzu9KTSPX4s/JCRNZpiS4N86Qsc+LECbN6FE2cOBHdunVTr7lx4wamT58e45xCMc07+8knn6jXLFu2DI8e6Q/nZK39ansTSYM+OU5/f3+zy4yWnqfEKO8tbVITZz5ogSZ5c6BoxrQqCBQQFo79T56jaIaontVav16+jc5F8+NQx6YIj9TgoMcLbLn7BB/sOoq2Ow7h7weeWHJNf1iyxPYh8qrh4rIg7gq8o/BGfWTBh8gDDwTD+21gqRmyqrQr8EMQrDecUIa0bvDZvVwt7q4uyJ4pA/afvYJ7ns/h+cpbrddlav32w6ex6cAJTP20R3QvDWsplSG9GiZuW6PaqsdKSGQEHvgH4KL3G7g5Opp8TWhEJH6/8wAVMkf1bLe3s4OjnZ3qaSS9zGyhTOF82LTvGHz8A7Dz+DnVEyuu9Yv/2IPyPYaj3sej8dvW3Vi8zfqNaQaWLaKGi/u+VnkUzZgOh568gH9omPrcFM+k/1k7+ew1vIJDcLBzU/QvUwSzztu+t4iWi4sLzpw7j6vXriM0NEyvB4PImjULvvh8ME4e3o8t2/9UwevV6zbgwMHD+H6s9efQ1pUlQzrVO+jB0xfYeews8mTPYtZ6OSZnJ0e4pnG22X0rihcqiIPHT+O19xus/WMHypbUL0OJMiWKwe/2OQzu1xMTpv83b7GPrx+mL1iMujWsP9qFu0saNUycLBncXZE7S0bsOXcdD569wqMXXsiSPu6Au3x/OTk6wNU5ajQUWyhZqjR2bPtD/W3dsW0rSpXWL0Np3bp5A1s2bUC1mjWxbvVKdO/UHsUL5MGRQwcxYYzt7uEypUth45atqsHHzl27UapUSb315cqUUb2Irl08h5w5cuDfQ4dUMEnKfK6uruo+tuVIONYkf0usvaQ2pv+KJzPSKyZr1qzqZ+nZY05Xegm0SE8g+SDIUq9ePbMf+oVs26BBA/UBSpcuHWrWrGnW6xK6X1G6dGnkzBnVCqRx48aqpVl8z5O0TovNt99+ixEjRkT/Ll808Sk4FG/zEc4tHIPwkECU7f6VSru6bgZyV2uGTEXK4/CkDxHw/BEcXdxQoHFXNQxWi9n/qu0eHdqqgkrp8xaDrRRuPQhnZ38Oe0cnlOsX1ZLl/LwRKN1rLNKkz4wDXzRRc6E4pc2IIq0HqvWm0mzB0TkNqrT5EEs+a4u0WXKgw9ioP/rLh3VGn1mb4PPCE4sHva+++NNlzYGmn4xVAYzqnfpj9Zc91LYN+o+0WX5lWLz8FWthUf8myF64FNqOnq16Nq39uid6zlineglt+e4Tld9MeQqg5Rc/Im3mbChcvSGWD24HZ7e0aP75f62LrC132eq4dWgHVn3cBHkr1EaByvXxxvMBzm5YgCbDp+H63s04uuRHVSmYtVApVO44UPUsOrzoBzWUW8HqTZCnvHnfD4mhYO2WODLna2wb1hKF6rVBxnzF4HnxKF7dvojynT/DhQ1zcHPXGnUP5ChdVa1vNGq+aroT8PqZ6mVUunVfm+W3cJMPcHr+KNz441cUa9VPDSl3f/8GOKRxQ/46rXHmt9F4fvEw7J2ckbtKYziny4SmU7eqB5fXN8/i9a3zyFWlsc3yW6B5X1xbMR7hQf4o2jFqbP+72+YgS9l6yFikIs7N/AgBz+/DMY07ctWJmvTx/l+/IejVEzw9vg2uWfKg6te26z0yZNQ4fNytnfp78+P8qDmShvbpirE//g/FS5XFibtRPVFnT/kODZq1RLr0GVCnRF6EhgQjU5as+ORL2z3QShCizydD0L5RDWTLnhOzl0aNo9+lWV1s3HMETVq2wZn7URVdIwb2wpgpM/D8qSfeq1FOpeXInQejJ9suQJ87T140aNIcdSuUQskyZfHriqj8tmlUB38eOIqBQ4ah/6dDVNoHbZpjzbadOHvyBHq0b6mCx9Vq1sbc3213L9SoXQeb161G1VJFULdBIzR/vzUe3r+HH78fjwXLVuGHGf/DpOlRrcy7tGqO6XMWYPP6NRg1dLC6Ni1atUHDplGBLiKipCBlL2lsJwoVKqTKMtJoL66y18cfRwW9c+fOrcpA0nAuV66oyZrjIuW0IUOivsulR5Hs1xwJ3a+oXbs20qRJo36WXkyPHz9WFa3xOU/WLu9FDQkUVVnStlBeDDt0Fo227kOnIvlURfaBJ89xzcsHg8sXR+O8OTDm+EVMP3cdJTOlR7ksGbGsaU3VM+OxfyB+vXwH/UrFXkZNKBkqTusBAlWvIKkSk95FMj/RdfirYeaKI63qSXRQ9TmKRG64IB2csAsv8AhBcIQdisAdrlYeck6GitOaMKALPhj3C/wDgzF3ZFTvimG/LMUHTWujdrkSJtd/u2ANHjx7iQVbd6NE/tw4t+wn6+ZXJxD1afEi+OTkOTW04JSKZVXa56cv4LvypZEljTO+OHMRp155oXymDJhUIarytcHugwiNjER+N1d8V6E0bGFg++bo/M00TFuxBcO6tVbDia3aeRCvfHwxvFsbk+vPrvhZ9X44d/MeVu06iAFtm1o9n6oy923FZJ60bqiVKysabd6H0pkzYHbDqoiI1KD7zqPY0KouSmVOj9fBoWiyeZ8aku77mlHPzC3/OIAnfoHYcucxBlcojkHljBtBJrZJ48fgg179EB4ehsUL5qq0PgM/wagRw1G6VElMHDcGPfoOwLPnLzCwX2/13Tdq7AR4v3mDmbPnonrVKjiwawdsZdyAD1C731fIkNYdG378WqXV+2gU9s6fjDTOTkbrQ0LDkKlxd9WDo3ThfFgx8b/vVGvLlDEDenduh+J1mqNowQLYumSeSq/SogPO/rMVF6/dQI33u6h7p1C+PFg4fZJan7dyPfj5B6LD+80wsEdXm+TVUSc4OLb7++g+dbHqVTTr06iRLUYu2oR2tSqgXtlimP/nvxj1+1YVcNstwaQVU1BnxE+4dM8DuTJnwC+f2GbUoRbvt8LWTRtQtlghtGrTDnXqN8C5M6exdtUKTP/fHFy8cB7vN2mIDBkz4stR36JY8RJYvGK1alwnjfP69+qBCZN+gK0M6NsHnbv3xIz/zcaQTz9BwQL5sW7DJnh4euLL4UPRq2d3dO/VF0VKllVzFzWoVw/DRnyFxUuXIV26tBj5xXBke1tvTGTITpPMQ4jSnf67777Djh1RfzDGjh2rHuD79jWu5Lxy5QqGDx+uetD89ddfGDVqFBYuXKiGjuvfvz82b96MqlXNi6BLS67x48djwYIFqpWXDNk2b948dO7c2ar7leOtX7++aslWrFgxfPPNN2jTpo0KPGn3Ycl5kv3G9jpDUmiQFmZNZ+xWgZyE0ERGRre21U4+b6qVvfYWjM88NNIqKbEY51fGoPzvAdhUmqUqFfhvXpCEksCLzEcS9XPUcG2J3bpZHkITO79yveVcqnlcEjky7x+sPw52YtwP2vxK0CV6nYm0+PBLpPyqz5AMPSb5jYxUPZtiypvufZ5U51f3My/5kWFKYvxueHtc8eHjF7+5CYzyIXm0s3ubX+PvMpWm/ajYmf6ei8usbhWRqHMVGZwz3TSZx0DNXZWA7wvHRByfV/Jj2NpQN7/SZV5+Tkh+3ZzsrZpf3TTJr/yckO+3xBwRwtT9oM2vfMbkZwkyxpefny/KF8qrnpXSp7fdHFexPcNsL10Z7gbXKDEFRESg7bVzyeKYid4Vc+fOxZs3b1QZRjRt2lSlSfDF0OLFi9Vwb9ILaNy4caosNmHCBFy9ehVffvklbt++rRr5mWPMmDFqe/n39OnT+PTTT9VIEHnz5rXqfuXYfvnlF/z+++/q+1jKjKtXr8alS5ei92HJeTpy5EisrzP1XXm8SzOkdYqawySh1LPx21a9cZEKQUtb585dczX+eTMYWEwCSNqUmIYd175GN9hkiVkHpyIxmfpbrksme9d9vrZ0aKlLPybuHBW611jb40L3mkfI872Z94spFcZ+mij5tKQeIq5rEJvbvy5DYpfRTc3lortee487vC3DWKLo7KU2um8jDO5by5/d7G4cSVAeDfMTWx7CwyNgb5+wMpR9XuO/afEVV/lD1kteE1ImCTufuCN6yD0h+dEOkxbxdhQD+S6W7y35LhPxHR7Pt36/RM2rbnk0pvJSQr4bMtrHPSRqYnyXSR5lm/h8xnSfHbLnLZgsyj/a55iSQ9arxsfWEhESiBtzP0gWx2wryb4nkbSmktZZWlmyZInx4khrhDx58qifZV6gO3fu4IsvvlAf2EmTJpkdqBESmJEH/sGDB6vxpuXBXSbZs/Z+ta3Djh07hh9//BFNmjRRrdTkoV+7D0vOk27ebE23Yje2StPEDhRYK7+2HEbKHNoAUdTPyf6jHJ1f9VCQAvKrvR9M5Te5HYP6DGnH0ZUHwVhGEk0OwyTpjfEcS350jysppbjvBhPnVDctIQ+H1mAqP7r5TUgAw1b51U1Lbvk1dT9o86sdeoCIKKllzJhR7/tKyi/a3jKGpGJAyjraXjISMJEGhNLrSBrnmRuoEaNHj8bIkSPRr18/tG7dWs0NG9P3YmLuV1velEaIEpSSPMjcstJDSbsPS86Tbt6Sgno2NnNbWw/fIjkz3GNcuTX1mqQUVwWkLeYbsYTuNTZ1vSVwkRxYUg9h7WH8LBFbgOi/9XYp4L5N/mWSmHr8JQdxlT+S2/k1vCdUOcQgj8npu8zc8mhy+G6I67ssOeSRUoZk35PIFBm/VLrjG34oZNJPS18XGhpqVACJ6b1ssV9vb2+9HkGJlQdzJWZPIltIzJ5EtpCYPYlsITF7EtlCYvYksoXE6klkKynt/CZWTyJbSMyeRLaQmD2JbCExexLZQjIqH6XInkR/lqti9Z5EbS6fTRbHTPQuk8+YzPOqS8bTj6sBnKnXhYWFwcmg10xM72WL/W7dulWvR1Bi5cEc1uhJZG0J6UmUFBK7J5G1JXZPImtLrJ5EtpLYPYmsLbF6EtlCYvYksoXE7ElkC4ndk8jaErMnkS0kVk8iW0iOPYlKDY2axsCaPYmuz+6aLI7ZVlJk81Hpyt++fdRcELqRXZk01NLXmepSGtN72WK/mzZt0usRlFh5ICIiIiIiSilk+G9pPKdLRmhYt26dxa+TeVtfvXpl1nvZYr/Saym2lr3xzQMRERERUaoJEsnwbzKkm61eZ+v9ynxGiZ0HIiIiotTAzkGGBLVedyw7pKxezEQp1YcffqgWW70uqfdrrfciIiIietdEzRFmvRFONCls9JTEkIIGNCEiIiIiIiIiIiIiIqJU3ZOIiIiIiMgUewc7tViLfTKYDJqIiIiIiCi1srOPWqz5/qlNKjxkIiIiIiIiIiIiIiIiYk8iIiIiInpn2NnZwc6KY0jbRbInERERERERUZKW+eysWOazS31lPvYkIiIiIiIiIiIiIiIiSoXYk4iIiIiI3hn2DvZqsdr7a9jGioiIiIiIKKnY28tivd4+mlRY5EuFh0xERERERERERERERETsSURERERE7ww7Bzu1WO39NalvfGoiIiIiIqLkQuagteo8tPapr8zHnkRERERERERERERERESpEHsSEREREdE7gz2JiIiIiIiI3l12dlbuSWTHnkRERERERERERERERESUCrAnERERERG9M+wd7NVitffXcLRmIiIiIiKipGJvZ6cWa9GwJxERERERERERERERERGlBuxJRERERETvDgc7NS+R1WhS3/jUREREREREyYa9deckgjXfO5nieBlERERERERERERERESpEHsSkUlBwWGQ/5K74MBQpCR5K+VGSvLaP2Wd37QuTkhJnP1DkJIUzp4WKckNT1+kFGVcA5GS2IWlrPziwTWkJBGvnyGlcAsIQrIcn9qKLb+sOfY1EZGtFJv0I9KnSxnPdh9Py46U5OnsoUhJCjYtj5QkMtAPKUmxn+ciJbELeI2UIvTxLaQkrllzISWxz5AFKYlTCuv5YRcSgJTCLjT5lf/trNyTyC6F3U+JgT2JiIiIiIiIiIiIiIiIUiEGiYiIiIiIiIiIiIiIiFIhDjdHRERERO8MOwd7tVjt/SPZxoqIiIiIiCipyPDiVh1i3J7DzREREREREREREREREVEqwJ5ERERERPTOsHewU4vV3j8y9bUqIyIiIiIiSi7s7OzUYs33T204XgYREREREREREREREVEqxJ5ERERERPTOsHOwU4vV3p89iYiIiIiIiJKMnX3UYs33T21S4SETERERERERERERERERexIRERER0TvDzsFeLVZ7/0i2sSIiIiIiIkoq9vZ2arHm+6c2LOUSERERERERERERERGlQuxJRERERETvDHsHWazYqizSam9NREREREREcbCzt1OLtdixJxERERERERERERERERGlBuxJRERERETvDLYqIyIiIiIienfZ2dmpxZrvn9pwTiIiIiIiIiIiIiIiIqJUiD2JiIiIiOidYW9vD3sH67WDso9gGysiIiIiIqKkYm9vpxZrvn9qw1IuERERERERERERERFRKpTsg0S+vr44d+6cWdv6+/vjzJkzVsnHjRs38Pz5c4vymJj5ieu9rHns5ooMD0NESJDJdaH+bxDs9QwRYSGxptlSRGgwIsNDTa4L9X2NkDcv1DFphfl7I8T7GcKDA5AUggP8EBkRYXKd9zNPeD/3NFof6OuD8NCkOb/B/r7QaDRG6ZLHN8884Pf6RfT68NBQlaZdgvx8bJ7fmPapzZu/96votNDgQL38hgYFJMn5NSU0KBC+LzwQ7PfGaF1IgJ9aZ+q6WJMmMlLdvzEdx5vnHggJ9I9OC/TxVmnaJab73loiw8MRHhxocl2InzcCXz9V3x+m1sX0OmsKDAxCSIjpz/nzFy/x4NFjtfj7B5j1Gmvz8/dHeHi4yXUeT5/hweMnagkODtFLCwi0/bkVb/z8Y/zMPPB8Hr1EvL1PTaXZkref6e+j8IgIPHz+6u3yWqUFBIfopL2CX6DxfZ0S2TnYWX0hosT16NEj3Lt3z6xtnzx5gtu3b1vlEpw4cSLGv48x5TEx8xPXe1nz2M3lHxCAsLD/ykimPPF8ZvS33lSatT31eAzPx48QEmz89y0sNFSte/nsKSIjI/XWvfH2ivMYrSk0IgKBobHv/6V/kFp0vQ4MjvN11iD7DInj2j7z8dfbRp6tPN/4wSvAdP2BNfkGBCEiItKi9fL7oxev4RNgu+dR+bzIc7Mpb3x88eDRE7W89vI26zXWJs/uQUGxP0s+f/kqehv5Tnjk4YmQENN1Mrbi7R8YY/rDF17w0lnvL8/OL7zwOobnbWvy8fM3+q4SYWHhePDkKTyevdBbL2lPnuqn2VpYeDj8YylfGK738vXHg6cv8eqN6boOa3rj/d/nyJB8X7169TK6TOjn54dHDx9GL4E2LqdKPt74mK6/CggIjC7/P3v+Qm+dr5+f3vfFOzEnkb0VF7vUV+ZL9kGiW7duYfz48WZt++DBA3zzzTdWycevv/6Kw4cPW5TH169fY/PmzYmy/7iOzZrHbo43dy/g+PedcHxiZzw9scNo/cN/luLC/GE4/l1HPD+7J8Y0W3l1cR/O/9QN56Z1w5vbp43W3/tjJq4tHoHz03vA9/5Flfbon0W48tvneHHqT9ja7mVz8V2nepjQoTY879wwWv/byP6Y9WkXTOxSH88fRRVet8yahEldG2JChzq4e+GUTfO7fcZozOnbDPMHvKeCQYbBmBVf98bvQ7vg14/bINDXGy8f3VFpsiwa0glX//3LZnmNjAjHylF9MbdfcywZ/gHCQvQfYl48uKXytfDT9ljxdR9EhIfh3rlj0fn97bN2eHztgs3yG+zng7VfdMKqwa2w86cRRusfXTiKPyZ8hFWft8W+ueOigzS7//cNln/ynloXbiLAYS1eHg8wt29TzOvbFPt+n260/uLuzVg9qg/m9m6Mw2vmq7Sj639TabL8PqQjIsJsV3B4dfsC/v7yffz9VRvcO2D8/X39z99x6KdP8feXrfHw6H/3aWiALw78MAAeZ/fDljZs2YaCZSsjf+lK+Hv3XqP1A4YMR/N2ndWyc89es15jTT/N+Q2Fq9RHwUp1cf7yVaP1Lbr0QbOOH6rl4tVrKq3/0K9Ro3l77Nht23MrPvlhLoq3H4RSHT+Bx4uowIqu8h8MQbNPx6rlpbdvjGm2IJUXrUb/D6X7j0WdoVNUAEiXBIaqfjoRzb+eiTZjZqm0fy/cUL/LUu3TSTh6JWkrHoko9dq+fTvWrFlj1ra7du3C8uXLrZKP4cOH4+XLlxbl8erVqzh+/Hii7D+uY7PmsZtj3pKVKFilPvJXqosTZ86b3Ob46XOo0Oh9PHjsEWuaLQzr2w3dWtTF5fPGjSkfP7iHgV1bo1ebpmhXrzI8nzxS6WOGDkKH+lXxfo2yuHXtMmztxOPnqDF/E2ou2Iy1F43/Lnv4BqDF0u1oufxPTDzwXznWOygEbVf+jX9uP7ZpfjedvYFy3y9C2e8WYf+NBya3ufnsNepPX4nLHlGfrUtPXqDipN/x3qx1mHfgrE3z++PqP1G850gU6T4Cl+89Nmv9rpOXkLfzUDQZPhXr95+0ST4vXL6KQhVqokil2pg+e4HR+l/mL0SDVh3RvGM3zPp1sVmvsaa/9/2L/PLdUKU+1m8zXZaXoFDV5h2w/2jU92X3T4ajSadeyFe5Hjbt2AVbk+BPtS9/QbnPf0L3n1carV+69xRafr8Q5T//CX3/t1al7blwS6VVHPozPp63wWZ5/frHOSjUoINa7jzUv2/vP/FE4w8/Q63OH6Fo406498hDNVSTtNpdPkLxJl3w+KnpBu/WdPTSDRTo8BkKdx6MhX/sMWv9gi270XjId/hmgXnPA4khKCgIzRo1QOUKZdGudUujxgwSPKpdvSrq1qiGJg3qqoDQX39uR9tW76mlbs1quH4tqtxqC0+fPUeZGvVRokodDPzcuD5o29+7ULtZKzRv3xUjx36nt67nR59h/JSfbJZXSnmSfZAoJStQoACmTp2K1ODR3lWo8Nn/UGPMWjw5aPzHslinL1Bz7HpUH7UCHke2xJhmK56HN6DCF8tQ5uPZ8PjX+A9Qyd4/oNLIVSjRaxKen4wKChXp9DUKtR0OWwsJCsT5A39j4rYT6DFmOvavXWS0zTerduG7LUfR5pNROL1zi+p5cevMUUzecRr9f5iPk39vsll+Xz2+D58Xnhi54Tiqt++Nczs36q13z5gZQ1fsw/DVB1GmQUvcOr4fuYqWVmmyFK1WD6UbtLRZfu+eOYKMOfPiq00nkbdURdw4qv8Ak7t4WZWvEeuOIEP2XHhy/QJK1m4and/8ZaqgUMWaNsvvzUM7ULx+K3y0/DAiIyPw7JZ+4bVorWbovWAn+v9+AD7PniDA6wXuntgL3xeeGLDkgFrnlMbVZvk9vX0Vmg4aheFrDuPhxZMIeKNf0V6jYz8MXrYXQ1cfwrWDf6u0ZoO+UWm9f16NknWawcnFdvm9+dcy1P1yDt6b9gfu7Ntg1IOkYo+Ral2zHzboBZFu/7Ma+WvZ7r7Vmj57Li6fOIRje//GjzOjKv4NHdn9F66dPoouHdqZ/RprkFaDa7dsx8PzR7Fi/kzMmG/8XZY3d04c27kZN0/uR40qlVTaPxtXYMwXg2Frd588xe1Hnni2ZxW+7NUBC7cYF15rly+JY8t+xu1ti5Aza6YY02xh/4XryJ0lI55u/AX1yxfH1sPGPZy7NKiGY3NG49Liier3VjUr4ObyKWqpW64YmlYpjXeBnYO91RciIq0WLVqgd+/e7/wJkUrH31asxb3TB7FtxUJMm/ub0TbSYn3Vpj/wftNGsabZyoY9R9G5V3+T6woXL4m/jl/CrtNX0bZLT5w4dADeXq/h8egB9l28g29++Bl/bbFdJbDWvOOXsbZbcxwa1AG/nzGueJx55AKaF8uPM4O7Yk6b+tHpi05fRacyhW2cW2DOgTM4+W1f7BrWDTP3mG6Y+PvRi2hXsXj0799uPYBvW9bGpQkDMa51XZvlNSAoBJsPnsb99b9g0Vcf4ZcNu+JcL41wPvtlGTZNGorba37GoDa2uY9nzP0NqxbOxf0LJ7B601aTPXR+nzsDFw/vxcTRX5n9Gmv5cfZvOLlzMy4d2IHp86KCVob+t3ApenZsG/374e3rcPvEPvyzfinWbLF9Q9y1h86hW71KeLL0O3WdT9+OChRrjWjfENfmfYP7i8bhlucL+AYGo0PNcirtwe/jcO+5F556Wb9B2IvXXjh46jyen9yJn775HLOX638vFS+UH/f+3YpHR7ZjULf2+OfwSTg4OESnfflRD6z707aNssW0ldvw94zRuLV+NuZu3mVUrja1fkzfjtg5c7RN87lj+zbUql1bBdfy5cuP/Xv1G1Hu37cXrdu0xY27D1C5SjWcOHYU3Xr0xIUr13H+8jXUql0HVapWtVl+f1+5BsM/G4Snty6p3kI379w12mbCNyNx5uBurFoU1QhXrN20FU0a1sO7xMHezupLapNiSrnSne/YsWN488Z4CKWYSBdxGYLt/Pnz8epiKRFlaRUWU2uyuPIYnyHgtEPX+fj4qPeS3+MjocduqfAgf6TNVRhObumRJmMOhAUaDy0V4vMKfh634Zold6xp1iZDyDm6uMPJPSNcs+aFJsJ0l3wZWi7g6R24ZM2DpOT9zAN5i5eBcxoXlKhaBy+fPDTaRoJCrzwe4dn9W8iWryDsHRzgniETPO5cw+NbV5EtXyGb5ff1k/soWKEG7OztUbhKHdWTxJAMgefl+RDeT58gU6580ekSQJDAh1t621Wsvn7yAIUq1VI/F65cB69N5FeGk3v1+J4aBi19tlzR6Z43LyNbwWLqfNvKG88HyFeuhvo5X/ma8Hn60OQQbl6P76hr4OyeDs/vXEGZph0RGuhv86HbvOT8VqwFBydn5C1dGd5PjVvsBfh44fm9G0ibOZte+sXdW1CmURsb5laGwfRFxvwl4OyWDu7Z8qhhMQ0FvXmJN49uwj1b1PeWj8dduGTKBpcMWWyaV2nl5O7mhhzZs6FIoYIIDze+tjmyZUXtpi2Rv3RF7Dt4yKzXWMtjz6coX7ok3Nxc0ahuLdw3aAWnrYCq3rwdilVriCs3biEp3XrogQZVysHe3h5Na1TC7UfGrZ/vPXmGKj2GocIHQ/DC602MaTbJ75PnaFyppPq5aeXSuO2h32LQ0cEB+89fR4WB49H86/+zdxfQUVxdHMD/cXeHECAkuLu7W4sVh+JQaIuUtlAquFNarLi7uwYNkBDi7u7uHr7z3pIQ2RD5skMC93fOnGxmdmbfbja78+a+e992ZBQqR+PgEwRzYwN+H0II+ZQiIiJ4Hygrq/xZxKwP9vLly0qXYWPVH1ifLzU1tVJtrEwJuPzSdWFhYbC2tq7Q863K514RrEyUuWk9qKupomPbVoiIKtlHPn7+MmZOGldksmdx66oTXvYsJAjd+wyAqqoapKVl4OPhCi83F9Q1NZP44+e9e4eQxBS+sLYkZWahsZ4WNJUUUEtdBYnFMoMdw2MwtXVDxBUq0eQWGQdjdVXoKCtKvL1ZObkIiktEaHwyv62uKA9dVWU00NPiZfKKu+3sgyHNG0BBVnSOwZ5jYGwiRrYyR7wAZW5Z4CcwIgYRcYkIiopBa7O6UFKQR9+2TeEbVrTihbjtITFxMNTWQOsGdZGcJlxpPP/AIPTq1hkqKspo1awpQsLDi2zX1NTAd8tWoE7zDthz6Fi59pEk1qeob2IMIwN9KCkqlMjGsLR+i/atW0JdXbXEvvZOrpg0Wpg+X2RCMi8Xl5qRBe+wGPRpIfof79PSDD7hH8rLFxaVmILW9WtD/f3/FytD5xYUAWkpKWioSP5/zicwBF3btoCcnCwGdu8IL//SswX9gsMwvG+3gjJ0voEh8PANhFk9Y4m3k/1vs1JxbGHXIOOSktG6YT1oqqnAtJYBYhKLXiMsa7tQfLy90au3KPjbp28/+PoU/T41bWAGfz8/eHp6ICw0BCZ16xVse/rkMTp37Spoe718fNG3Z3feR+3Tsxu8fYuWwWX//9t3/wfTlh2x7Le/CkrQObq4ok930XuDkNLIogZgpdT69u0LeXl5+Pv7w9nZGTo6H78Yx4Is3bt3h6ysLK8rra2tjcePH/NjlEd8fDw6duwIdXV1HnBRVlbmx6tIG8PDw3kJuEfFItEfw0rXTZ8+HXJycvwx2XGfP3+OiqjIc2fbC9fdrmxQCvgwKiAvNwvSMiXfWp7nNiMlzBtmX//w0XWSxi6csxJjBUqZa8Lj+ErkZWegwbhPV8aPkZGT4/PiMDk52ZCRLfnapiTEYc8Pk6GgpILpa0SZAaYt2+O/Jd9CTkERMzfsE7a978uDsTJhMrJyJe4T4euOyxuWQkPPCFpGJgXrHR5cQfPewwVra357c9/XHM8ppb3eb57h4cEtPKtISU2jYL3dvYvo9LWwI0elZeWQm/Ph9WW/F+f28DIcb59G/Y59+P8i7+y+eojXp/7h9x+15jA0DD8E5ySJvZ5lvR9enNoN9xf30WnU9IJ1rESev8NrdJs4H58Km7NMWqZke+2ObUB8gDtaTVzCf/d9dB4Nh0xDyJuHkJaT54FocX+XqsZGiLGT/3zi5s05vEf0eeDh7YMff16Jvj17lLmPpMjLyyHz/WcZ+15l31HFPbx8iv+88+gJ9hw6gX3b1uFTkZeTReb7zwb2U16u5N/U67ooG2rXuZs4cesxfpo2Wuw6QdorK4PM939b9lNermjAp66BDjyObeC3l+47h3s2zvi6W1v++9F7lpg97MNo5JqOfc+zRZLHJ4RUvQcPHuDKlSv84qKmpma5+kAODg4YMmQIGjRowPtNo0aNwq5du8r9mGxQHdu/cePGyMjI4AGjiraRlYBjj71uXfm/s1jpurNnz/JBhey7mPXTWLCnIsr73Kuqv8e+x/ODWewCYPFa/WyOFDsndhGqC1LT0hEeGQVdba0S68zq10V1wc5H/l67CuOmzoS+US0+WMW8STPMn/A1VNTU8M/RcxJvQ2pWNiace8BvP5z5IcuCycrNg1yx7xx1BXksvm0Jt6h4tDDQwfFx/XDc3hMLOzfHNTd/5Oa948EaeQkN/PCJjsfUw9eho6qMuz+MR06huXuKn1dmZOfAwj0Ai/t3wDX7LEQmiQKxOipKGPLveQTHJWFUm4bYPq4/JOWpgzuW7jmNdg3rY+2sMQXndlk5OZB7H7jKJ8+unxTbLiMtzS/Gtpv7O2ISkvHr5BH4acJQCHOen82vC2VmZZY4D12yYA5f2P9dnxHjsHD2t2XuI0nvCl0PYvNgsrbkY58Xxy9cxW+Lv+PlpqNj4vhnKOsLHD9/hfcPxg4fLEg7lx25ATvfYGybMbLIuXNWdm6J9wPDAkd/X3+KrTNGFilDt+fOSwzv0BRyAgywYn/HzPeDu9hP1kcpjvXvft68C7PGjUQdI4OCMnQDp/8AXS0NLJo2TuLtTM/MwsDFa/ntt0c2FdnG+1Ji+n7l3S5JcvLyBX1U9n/Dfi+stnFthIeHYeSQQWjRshX09D4MbD194gRWrxf1r4TC3g/538VsPq/i/+dfDR3MF3afQaMmICwiAkdOnsXXw4YgLCKSn3ewecvUVEsGbGsaaQln++RV08EtklQjernspJadNLNl6NChePr0aZn7HD16FJ07d+aZNKxOtJaWFj+hL69jx45h4MCBsLW15Sfghb/kqqqNpWGj2NhoNjZCbfny5RXq6FT0ubNyeBoaGgVLnTqVu3CsqFMbYa9vIM7ThmcVySgoITs1kV9kZTITo9Fw3FI0n7UJ/ncPlbpOCFLSMrxTE+9pjWj7h5BTEwUcs1Pi8S5PNPopMyESjaavh+no5XwuIr49LQk5qQn8+WUlxwnWXm3D2gj1cYef01s8OXsQdRq34OsTY0SjxHNzcpCdlYmF/55G38lzcf/Ybp5V5Gljie/3nMX4X9bjxt6iX9KSZGjaGJ6vHiHc2xVvrp3kgRUmf24ilkWkqqWHqZuPoUH77nh7+1xBp8LL6gkvNyck1j6WsRLp5wm7OxdQq2ELPu9QaoLob5yZlorajVrx9ipraBeUo2Pr48OCoFdX8iMMC9Nv0AzO9y4g2t8Dni9uQ8+0CXIyM3j2EMN+mnUbiK9XH0RydBgivZxh0LAFNGuZ4Os1h3kd+TA34Wp/GzVsDptrJxDq4cjLzekY10NmWgrPzsrPHus8dhYmbzoGt+d3kRInGpXqZ2sJk+btBZ8sUNXQBL5PLiHK7Q0yk+Ihp6yKrJRE5GZnFmQRtZ6yHN0W/w3Xq/v5a8/ua7l9EXwenoXX3RNIjiiZ3SUJ7LVhF2xu3n2AsxevQE9Pl6+PjIrmFzmYoJBQ+AUEwurNW2ioq5e6jxCMjQzh5ukNSysb7Nh3GG1bNuPr2ckqw9ocEBwCH/9AvHVwhoaGmuj5RMcgLiERMbFxfBFKS/P6uPX8DezcfbD73E20byr6Xw+Lji3opAWERcIzMAQuPgF8FJy4dUJp17AeTj58DSe/YBy68xztzOvxTnrU+4lfU9IzEBgZA9eAUPiFRUNTRZmvZ3MXuQeFo0V9yY8wJISQjzEyMuIVFZycnPiAvdDQsuev2bRpEw/OWFpa8sF2LPjCAibltXXrVuzYsYPPPcsCQGVVrahMG0ujq6sLR0dHfiwW6KlIX7Uiz72q+nu62toICY/EE8vX2H34BJo1Mi/yPe7s7ol7j59jwNhpsHj+CnOWrhC7TkjREeFISUpCbFQkkt7/bdk6Ji01BT/NnYoOXXtCU1sHyUmJcHdyQLC/Hw5duo0FP63E3m2Sv/inpiAPy3mj+aIkJ4u6mmo44+CF5/5hSMrIgrK8HJ9vKPN99re5riamtm6E8xMGIjo1nWcgvQwMx6TzD3HU1h17rJwRGC+5EflNjXRhu2oWHiyeyAMokAIeuvnj/Fs3GL7PEolKTkVuXh78YhLwzCsQo/Zewk0nH/xy+THSs3NgoK6CLWP64uqCMXgh4TmUhnVpDc9TW3Hmj+9gYqADJ99gvHLxxs6L99C2oSgjICxGNJG7uO2slK+sjDTO/7kIl9f+gGuWwvSj2rRsjp37DsHS6g1cPbxgXMuIB4TS3mczsUy+gKAQ2Ng7FPSXxO0jFD1tLZy7dhs37ltAQV6etyk6No4HrcIjo/H0pRUGjJuGo+cuY9WmvxERFYN1f++Bp68/hvTrzQPIQji1dDIvFze0fVO0a2CMgw+s4BQQhnMv7NHGtDbSs7J5phDz1icYSw5dw/fDeyImKYWfV7NtY7q2xJ0/5yA4JgFvipWok4TGpnXx1NoW1g6u2HH4DDq0aMLXh0WK+s0sAD/u+5Xo0aENDHS1kZCUjCweTJLD41N7MGPscGw/dFri7VRWVOCl49iirqIM8zq1sP/qQ1i8debztGqoKiMuKQUZmaJrhOK2J6WmITQmHqnpGQgWMx+sJLRt2xYnjx2Fq4szD/q0aduOX9+Nez9o5OTx4+jSrRtu3nuABmbmuHxJNJ1CdHQUUlKSUbu2sJWH2rVpxbMH7R2dce32PbRq3oxnCiW+HwASGxfPy9DZO7kgLiEBykpKeP3mLabNW4QFS5bj1r2HuH5b+DnASM1QIzKJGjVqVJAFY2hoyNPqyxIXF8dHhDHsC4odg60rL3Zftg/DRo40bNiwyttYGhMTEygpiebgYI/Lgj0VUZHnvmLFCixdurTIyLLKdBxMh82F18VtyMlIhflo0ej6kGcXoduiO9TqNIbHmY3IiAvjJd7qDhBlXohbJ5R6wxch4PYeniVQ/yvRPEOBd/ej7tD5vD3uR3/hGUby6jow7ivKboh4eRkxTqKJ0zPjw2E+4XdB2soyL75etAJX/1kLDX1DTPhFNM/VmQ2/YMGOY0iOi8bu7yfxQKZOLRMMn78cOrXqwLRlBxz9/XsoKqti8EzhMrVYyTCWXXPz71UwNGuK1oPG8BJnt3auwsS1B3jw6Orm5TyDR9fEDIPmizqLbD2b20dcpokk1W7cCvVad8K1rb/woJVp2668JJrVlaMYuuhPeFk9xpNjOyGroMgDSmyOHMbX1hLN+wib9cSYdR2ICE9HWOz+A036fgVNIxOEOL9BlI8r2o6aAZf7F+D26DLkFJVh3LITDxCxTKJwDwfc2/YTdOs1RIMuoucghA5fT8PdXX/h3p416DX9R8grqcDxwWXIKSjzuaeenfgX/vavoKCiiiY9BxeUnPN5+xwdvpoKoTUbvQB2xzciKC2JB4N4WywuwLB5F2g3aA7bo+uQEhkEBVUtNB4+k78v2BxFTJDVPcjIykPDWLjA4c5N67D411WQk5XFrm2iYPDK1euw/o/fYGigj5HjpyAjMxNNGppj2/rVpe4jBDZacNvq3/DTnxtgZKiP/dtFF17mLP4Vt88dRVx8AgaMngI5eTm0a9UcO9f9wbevWLsFL16LatyzEnXb1vwmSHv1tDSwdMrXmLtuN1o1rI9ZXw0UtXfNv7i9azUPAk1YsQWKCvLo3a4Fpg3vC2fvkuuE0ta8Lga2a4ZZW4+iX9umGNi+GYKj47Hl3B3s+XEqHrx1xYpDl6GmrMAziHq3Fp0nWNi5Y3zvDvicSMtI80WSxyeEVL0WLUQDoRgDA4MK9/lYFQbWj2Hr6tX7UA6mvH0+NrCOPW5Vt7E0+Y+b3+djQafyVr2oyHOvqv4e88+637Fi/VboaGli3xZR5tT85atw9dh/6NG5A7ytRX0l9l3//ZzpqGtcu8Q6Ia35+Qf4eXvi5dNHCAkKwKzvl+HPZYuw9/RlBPn7wcfDDdtWi/oiU+ctwpjJM1DLxAQ/zZ0GTW1tLPpFdC4ipJ97tsGK+1ZIycrG2gEd+bpDb90wyNwELQ11sKx7K6x8YI3gxGSMamoKE001HmBiLrv4Qk1BjgeShLJ5dF8+xxArJ7dtbD++7s8bL7D2q54FASVm2wMrDGpqyoNea7/qhV+vPEFCegZ+Gijc3K7s3HfTvPFYvvcMaulpY9/Sb/n6eduO4OamZWK3s2tBa2aOwaJ/jkMKUlg9Q5gM8ZVLv8f8pb/iyq272L7uD97XP3rmPJo3aYQBvXvip1Vr8MbOAXVqG+GfTatL3UcoW/9agUW//oXsnBzsXLeKr1v/9158N3MyGprW53MPMXuPnkaThg34PKQsoBSfkIiLN+6gVbMmuHR4N4Q0uktLWHkGYu6eC5japz0aGOriqYsP7HxC+HxEDx084RsRg9Ebj/D73/trHs6/cMBRC2uoKCrwUnUdzT9URZEUVRVlrF48Fwv/3ALz+ibYv1ZU6Wbmr+tx7+hO+AWHwsnDGz9tFFWQ+HH6eAzo3gnD5yzl72mzenWw7VfhrgflWzN3PL7behAJyan4Z8kMvm73pXsY1KkVOjUzF7v9ytM32HBCNGBi3MrtsDok+UB9n3798fz5MyyYOxsjRn7N5xdycnTA1SuX8efqtRj3zXj8tORH3Lh2DeYNG2LZzz/z/R7cu4uJk6dAaDMmT8APP/+GuT/+hB/nz+Fl5K/fuYeIyCjMmzEN/+w7gHNXrvNM3t9++hGaGhq4fVEUJHT39OJzE00ZPxafA0nPG5T3BWYSSb0TstZMJbASAH/99Rdu3brFf1+1ahXMzMzw7beiL/TCXFxcsHjxYl7ejWX0TJgwAevXr+dzC/3++++8HED+iXRZ2GiyqVOn8tFZrHwba8O+ffswduzYcrexffv2Be2pyPPt3bs3Vq5cCXNzc2zYsAHLli1D69atP3qsqnrurNPARph1W3+bz9tT3WWkVa5+96cysU8D1CSxKTXr9a1p4lKK1hiv7jSUy3/hojrwCJP8RKJV5ezXwpT/qypS2aIRdjVGcMmJn6uz3NgI1BRJqekwGLuYnyuxEr2ftC3vz2HsZo6EqrzkBhywC3dtj9yoFs+ZkM/F7t27eRYP60cx/fv35+vE9V8OHTpUUN5t+/btuHjxIn766SdeQeHIkSPw9PSEoqJiuTOJWAYO60exftg///zDs3KMjY3L3UaWyVPRcnNsv82bN+PPP//kF3LZMfP7cR87VlU89/zPyhgPWz6/UE3gkqePmkTrX+EvyP4/VAyFnVfz/6XWsiVqEumWovlOagqpbMnPFVVVsl5dR02i1KYXapLsMH/UJKltvkJNopopXJWM/1dScjL0TJtWi/5P/nnMoJ0PIackuevW2empuL94QLV4zkKp9plE7A/frl27IqOsWJq/OGpqaujQQTQatlu3bjyow2o9s9Ef7AS6vAEipkePHrzTcOnSJTRv3pwHgViGUEXaWLg9FcGCROwNyNo8f/58TJkyBYGBgR89VlU+d0IIIYSQGktGGlKSzPahTCJCqlzdunWLzDnL+jWqpdTLZ1kw+Rk3LLjDKjCcO3cO+vr6sLCwKHeAiGEZNqy/dP78eQwfPhzz5s0rdf/S2li4PRXB5qH19fXl89myAE+TJk0QFBT00WNV5XMnhBBCCKmpKJPoC8wkKi1qyObrKYyd3LM5hCq6H8u0yS/tVtaxhHhcbW3tIllJVdWG8qJMIsmiTCJSGGUSSRZlEkkOZRJJFmUS/Z+ZRHO+hpoEM4mSWSbRwWtf1KgyQj4FFjBxcyuaCcpKwnXq1KnC+7H5BRQUFMp1LCEe18bGpkhWUlW1oTwok0jyKJNIsiiTSLIok0hyKJNIsiiT6MvKJBr67yOJZxLd+aF/tXjOQqn2mUTiREZGYufOnUXWsTT9sgIl4vbLyMgoMeKqtGMJ8bjsfoWzkqqqDYQQQgghXwIpaclmErHjE0Ikz9HREXv27CmyjlV4KCtQIm4/NjiPDdIrz7GEeNxevXoVyUqqqjYQQgghhHwJKJOo6tXIIBGbq+fevXuC7Sf0465evbrK20AIIYQQQgghNcWIESP4ItR+n/pxJXUsQgghhBBCPssgESGEEEIIIaVmEkkw24cyiQghhBBCCCHk05GVZouUxI7/7gssHvEFPmVCCCGEEEIIIYQQQgghhBBCmUSEEEIIIeSzweYjkpKRkeDxcyV2bEIIIYQQQgghH0dzElU9yiQihBBCCCGEEEIIIYQQQgj5AlEmESGEEEII+cwyiSQ4J5EEj00IIYQQQggh5OOkpaV4NpGk5Erw2NUV9XIJIYQQQgghhBBCCCGEEEK+QJRJRAghhBBCPhvS0tJ8keTxCSGEEEIIIYR8GjJS0pCRYL9MRurL6/N9ec+YEEIIIYQQQgghhBBCCCGEUCYRIYQQQgj5fNCcRIQQQgghhBDy+ZKR8JxEMjQnESGEEEIIIYQQQgghhBBCCPkSULk5QgghhBBCCCGEEEIIIYSQL5Dsp24AIYQQQgghVYXKzRFCCCGEEELI54vKzVU9yiQihBBCCCGEEEIIIYQQQgj5AlEmESGEEEII+WxISUlDSlpaoscnhBBCCCGEEPJpVMdMopiYGLx58wZ5eXlo06YNateujZqEermEEEIIIYQQQgghhBBCCCEVtHjxYh4Y2r17N/bs2QNzc3Ns3LgRNQllEhGxcnPzIJWbV+1fnby8d6hJ0rNyUZNIMiovCZrKcqhJcmvY+1dNsWZ9ZRhrK6GmkE5PRE0ilZmEmiQrKhQ1SZq/36duQrmlpWeiuqE5iQghpGwO836AimzNOLdruWA4ahLp/j1Rk9yf9jdqksEXOqIm8VqyEDWJ+e4jqClk9WrWKP08BTXUJLnRNasPlV3Drq+8k6s51yveyeagupGRkuKLJI9fESxAtHnzZigoKPDfr1+/jq+//hqDBg1C27ZtURNQJhEhhBBCCCGEEEIIIYQQQkgFTZ8+vSBAxAwbNgxSUlJwcXFBTVEzhg4RQgghhBBSDpRJRAghhBBCCCGfL2kJz0kk/f7YSUlFq6iwQFDhYFBpHjx4gHfv3qFly5aoKSiTiBBCCCGEEEIIIYQQQggh5L06depAQ0OjYCnPPENhYWGYO3cupkyZgtatW6OmoEwiQgghhBDy2ZCWkeaLJI9PCCGEEEIIIeTTkJFwJpHM+2MHBwdDXV29YH1ZWUTR0dEYMGAAGjdujIMHD6ImoSARIYQQQgghhBBCCCGEEELIeyxAVDhI9DExMTHo27cvDAwMcOPGDSgqKqImoSARIYQQQgj5bEhJS0FKWlqixyeEEEIIIYQQ8mnISkvxRVJyK3js2NhYHiDS09PDrVu3oKysjJqGgkSEEEIIIYQQQgghhBBCCCEV1L9/fwQGBmL+/Pm4cOFCwfp27dqhRYsWqAkoSEQIIYQQQj4bUjLSfJHk8SsqJycH586dg4uLCwwNDTF58mQ+yqwsfn5+uHLlCpKSkjBo0CB069atkq0mhBBCCCGEkM+DUHMSlVerVq348ubNmyLrWZ+PgkSEEEIIIYR84ViAiE1eGhoaigkTJvDyAxs3bsTr169hampa6n7Hjh3Dd999h2+++QZNmjTBH3/8gXHjxvHRaYQQQgghhBBCqodjx46hpqNMIkIIIYQQ8tmobplEJ06cgJWVFc8KMjIyQm5uLrp27YoVK1bg/PnzYvfx8PDA3LlzsXfvXsyePZuv++WXX/gxCCGEEEIIIeRLVt0yiT4HkutBE0IIIYQQ8oW7fv06+vXrxwNEjIyMDC83d/PmTR4wEufgwYPQ19fHrFmziqz/WOYRIYQQQgghhBBSGZRJRAghhBBCPhtSUtKQkpZgJpGU6NhsnqDCFBQU+FKcl5cXBg4cWGRdgwYNkJ6ezkvQmZiYlNjn7du3fP4hBwcHPieRiooK+vTpg06dOlX58yGEEEIIIYSQmkRGSsKZRFKUSUQIIYQQQggpQ506daChoVGwsHmGxElLS4OamlqRderq6gXbxElMTISjoyOmT58OOTk5hISE8CBRaY9BCCGEEEIIIYRUFmUSEUIIIYSQz4aUjAykZWQkenwmODi4INjDiMsiYliAKCEhoci6+Pj4gm2l7ePj44OAgADo6urydc2aNcP333+PRYsWlbofIYQQQgghhHzupCU8J5H0FzgnEQWJCCGEEEIIqSAWICocJCpN06ZN4eHhUWQd+51lH9WqVUvsPiwgFBsbWxAgYtq1a8fnMAoKCuLbCSGEEEIIIYSQqiC5gu2EEEIIIYQITEpGWuJLRXzzzTd49uwZ3N3dC0rMHT9+HOPGjYPU+1rXzs7OWLx4MeLi4vjvEydOhJ+fH/z9/QuO8+jRI6iqqvL5jAghhBBCCCHkS8WyiCS9fGmqfZAoOjoat2/fLtd9Wcf6+vXrEmmHpaUlfH19K9TGqmxPWccSog1lyYiLQGrEh4sZhaWG+yHB2w5pUUEF65IC3fi69OhgfArp0UHIiAsTuy0lxANJfg7ITIgqWJcW7otEXztkJkRCaO/evUOYjxuSYsQ/tp+DNXztrZASH8t/z83Jho/ty4Ilws9T0Payxw/1ckZakqicTmE52VnwtXuFACcbZKQm83UZqSl8Xf4SGxooaHuzMtIR5O6EzLTUEtvSU5LhafMSfk5vkZWZUbA+Iy0F/i52SEsuOnG5EDJSkhDu5cJfy+JS4qLhb/8KYZ5OyMvN4eviw4P5uvwlPTlR0Pay92WolyvycnNLbIsNDeLv0TBvtyLrE6MjEOTuiNz3z0FISVFhiAn0FrstNtAbwU7WiA8NKLI+Kz0VYe52iPRxhdA8fPzg4y/+f8bB1R0WL17xJSxC9Hn27PUbPHlphaBQ8Z9/kv4sc3T3Ruj7thT3ys4Jj16+4Ut8ouh/65m1HZ68tkVkjOjzTUjZObmw8wlCbFKK2O2PHTwKlszsbCSnZRRZ5xMq/nlKSnpWNuwCw5GcUfKzIZ9neAyeeQQW3Kc8+5D/z5gxYzB+/Hj06tUL06ZNQ4cOHZCXl4f169cX3IedY/7zzz9IShK979l9WWm5jh07YubMmRgxYgQ2bNiAQ4cOQVFRkf4k5LPDAqVv374t133d3NxgbW0tkXZcunQJqaklzwc/1saqbE9ZxxKiDWVJyMrGm7h4OCaUfj6ZnJ0D96RkZOflfXSdUCITkuHgF8o/e4tzDgjDEydv/jNfSEwCHjt5wd43BJ9CYEQMXANCxW7zCo7AYzs3eASFl3sfSfLLzoBjZgqicsWfR3hlp8EpKxXRudkfXSfouah3AMJi4su1PTMrGxZvnfHM3g1R8cL2oRj2/+Iam4CETPGvb1ZuLhxj4hGTnllkfUZOLmwihT139vXzh6eX+D4Uk5CYiBevXsPZ1a3c+0haQkoqbD39+N+5PNt9QyNhYesCG3df5OYK91nGMsntnN0QJaY/xN6zFi9e8z5ebLyovHF6RgZfl7+4e4u/lilpQdHxcAks+lmVz8YriH/2ehXqL3mHRfPPXvdgYa+5paakwN7OttS5QiPCw2Fn+xaZmaL/s9jYGDx/+gRv31gXrBNSbFwcbO3tkZ0t/n3L1r+1s8OTZ8/578EhobB4+ozvI+57kJAaU24uMDAQ+/btw7Bhw8q8b1hYGHbt2oWvvvpKIh2G7t27ix29WVob2YcFq1dfFcp6bkK04WNinJ7B78ZuSMspQq9NP9Qd+G2R7bEulkj0d0JGTAiMun4N494TEGlzhweWWLDGsPMImPSbCqGEW15E9Ntb/CK6cb9vodt6QNHnY38fGbFhyIgJgsmQBdBu1hOxLs+Q6P0GOi36wKjHeAjp6o7fEehih/SUJEz4bTtMW3cqsv3FhSPIykhDdJAfpqzZA/26Znh69gDflpoQh6bd+sPQtJFg7T21cg5va1pCHKZtOQpd4/oF27Iz0mF5/iDycrIRFx6MWTvPFqxjEiJD0Wvyd9CpXVeQtrLAz47ZoyCnqMjbvPTAVSiraxRsT4gKg8Wp/5CVmc4DQssOXkVOTja2z/oKGrr6iIsIw5IDl6GpZyhIe5NiInDypylQUteClLQ0vt1xlv/MFxXgBatLh3kgSV5JBZM3HUOImx2cLUTB4kg/D8zecw1CCfF0xolVC6CioQU9E1NM+uOfItsDnN/C/tF1HgA1adoGY3/eyINGFzb9AjVtPSiqqGLOjpOCtdfX+jGeHtwAOQUlmHUdiK6Tvy+y3c/mKUJcbJAQHozmA8agw9jZiA3ywbU186GmawTDhi1hYCZcGaidB49jz9FTyMnJxR9LF2LGhDFFtv+28W9k52TzbIUlc79FLUN9/H3gGNLS0+Hq6YNNv/2EqWOr/juzNN//uZUHfRKSU3Bs25/o17VDke3frdoMfV0tfnvrih+hpaGOf46eQ1JKKjx8A3B+9wZ0a9dKsPaOWb0XUYnJiElMwa1136NxHaMi26dtPoxWpsb8dsv6sxCfkortF+/z34Oi4vDTuEEwq60vSFszs3MwcNtpKMjJIDEtE/d/mgxtFaUi95l3/DasfUNgqqeFLeP7Q15Gvcx9aqLKZPtU9PgVur+UFE6dOoUXL17AxcUFo0aNwpAhQ4oEe1q0aIG///4b2traBeu2bt2KyZMn8wu/WlpaOHDgAIyMir4HCflcsGw7NndX+/bty7zvq1ev+HxdnToVPR+uCtu2bUPnzp2hoqJS7jYmJiby8pBVoaznJkQbyhKano6TAUEIz8jApa4l2xmenoHv7BygKScHWSkpHGjfBhEZmSXW5WdSShq7EDlh63FoqymjuYkRji6eVGT7dWtnvPYIQHB0Ajo3rosDiybA1icY/917iXd573BvzQII6coLW/y07xyUFeQxvk9H/D6t6HmahZ0bzj2xRtO6tbBvyfRy7SNJLzMSYZuVgh6KGhij8qFEakF70xMRlpuJkJwszFYzRDdFdbHrhLJw2yG8dvFCQnIqjv2+CL3aNP3o9mb1jbHtzA3k5r6Dq38QLqxbim4tGwvW3gVPbJCUlY24zCwc6tsR9dRVC7bl5OVhyoPXkJWWQnhqBnb3aodmOpp823EPf1z1Dca9r/oI0s5jp85g3ebtkJGRxsK5s/HDd/OKbLe2scXYydNg1qAB+vTsjhbNmpa5j6SFRMeh7/droauhxuceeb7nT0gX6leL2/7cwR0XHlshNjEZSgryeLr7D0E+y6Ys+hkevn6IjUvAlcO70L5V8yJBoi17DyMnNwdevgG4e+YgtDU0+DomPDIKMyaOQRNzYTPRb75xwZJDV6GiII9RXVrir0lDimw/+cQGvhExPDA0vW9H/DZ+IJ46e+PsMzuY1dLln8VCSEiIx5C+vaGuoYH09DTcf/ICSkof+kMP7t3Fz0t+QG3jOpCRkcHlm3cQ6O+PXX9vR0pKCnKys3H70RPIy8sL0l5vHx8M/mo09HR0oKmpiXs3rhbZnpycjH5DRvBgkKGhAfr06glbO3vsP3wEMTGxMKljjMvnTuNzIOlsH5kvMJOo2geJajLWkWeTC38JbQh7eQWtFu2FnKoG7HfOhcmA6UW+LE0GTOM/c7My4Lx/KQ8SmY/9ia/LTk2E+8m/BA0SxTo+RPOFB5CbmQ7vM3+UCBLVG/Ej/8mCWiGPjvAgUZ0BM6Fq3BgZMcJmPrHABcsi+vHwLYT7uOHpmf0lgkTTN+znPz2snsLV8iFMmrbG7G3H+brr//yFNgO/Fqy9rI2KqmqYvuUYXJ7egf39Kxgwa1nBdiU1DUzffJTftrp6Ar5vLdFu6DcF686t/h5Nuw8UrL1uLx+jadc+GPndL3hwfA8cn91DlxEfgoBGpo3w3T+iIMXlnasR5OGE7MwMtOk3HCPmL8ejk//B1+EN2g0YKUh7XSxuoOuEeWgzeBxu7ViJIJe3qNuyY8F207bd+MKc/2MekmMi0KLfV3xhmVK3d/4GNR1hLloz1jfOYvzKbWjQpjP2/zgJCVHh0NT/cJGz3eDRfGEnt//OEb2Gbq8eY8KqHTBt1REHlkzhmU/sfSMEh9unMW7DCaho6eHMsm/QecJ3kH4/ST3TYewcvuRkZuDy7zN5kMj6wn60+3oGWg+fDKEdO38FjhY3kJaegWFT5pYIEjGrl/+IOrUMYWwkCmReO7qX/7R1csWeo6cFCxIlp6TijaMrHO+egbOnL1b/c7BEkEhHSwOrF8+DWV1j6OmIgkVX/tvCfz54YY2r958KFiRyCQiFgrwcrP5diWsv7XH8wStsnFX09W1obIA/poxAk7pGUFdWgq6GKm6vF31/TNpwEKO7t4VQHrn5o3fjulg7ug/+fWiN6/aemNG9dcH2N36hcA6JhNXvs6AoJzr9u+3o/dF9SNXq0aMHX8Rhg5BYubniWrduzRdCSPXVpUuXL6oNzTTUsattK8y2sRO7/W5EJGbVr4fhtQyx2tUdTolJsI1PKLGulaYw53YH7r/CiSWT0bVJffRftQfh8Ukw0voQlFg1fhD/mZObix6//Mtvf9W5BQa1bYyv1x2C0PZes8Dzf1ZAT0MNnb5bg1VTRxbpVy/4qi96tmqE3VcflXsfSZqqZgDzDCWElpJJtEBddN4fmpOJkylRPCAkbp0QWODH0TsQtkc3w8E7AFtP3ygSJBK3/eyaxbi74ze+/exDS7xy9hIsSOQelwg1eVkc6tcJ9wLDcNUvBEtaf3jskJQ0GKsqY0ePtngQFI4XYdE8SMTWs7++jqIChHLgyDG8ef4I8nLy6DlwaImAzx/rNmDL+jWYOG5MufeRtDMPLPHLlJGYNbwP5mw+gBdOnujVuslHt88Y1psvTO9Fq5GSngE1ZckOsAoOC0d0XBzsH1zFY0srHDh9oUiQiAW27p8VfVadvXYbtx8+xS+L5hSs+3bxCkz6uuxB91Vt721LPF6/CIaaauiy/G/8Pn4QDwjm+3ee6L2QkJqOcRuP8iDRnEFd0bO5GbZffSxYO29dv4aJU6bih6U/YfXvv+HRg3sY8dWogu32tm/x1agx6NajJ06fOI6njy0wYNBgXLx+i29fNG8OPN3d0aKVMH3UYydPY8v6tRjz9Vf4ZvI0ODo5o1XLFgXbjxw/iUaNzHHysGgQNvP1yOF8YXoPHMIDSWpqaoK0l9Qs1b7cXD5vb2+cOHECXl5e5d6HjaY6d+4cLl68yCO8lcneOXnyJGxtbSvVxsqUessvG8cmNGbH8vHxqdD+VdGGysjLyYKCph6kZeWhqGWInLSSJbhYebnIN3egXrfoKPtoBwsY9/xG4m0saGt2FmSU1CAtpwA5VU1+cVocVm4u2v4B1OoJN2pdHFZ2y7B+I/7lX8u8Gb/IXlxOVha8bCzhavkA9Zq3K1JGLT4iFLoCZeUwCZFhMDIX/Y1rNWyOhIiSZQ/YRX9vmxcIdLZBrUYfvtDiw0N4gEleSVmw9sZFhKJOI9FJVp3GLRAXXrKsRGJMJFxfPkZ0kD8M6pqhVoPGCPZwhoulBXwcrGH8fn8hJEaFweh9poqheXMkRpZ8fePCAuFh+YCPHlHR1ClY7/rkJhp1Ey4Ax8RHhqJ2Q9Hrw36yTLHiWDlE+4fXoWtcj/9er0U7OD6+Beend/n7XkFFuBOYnIx0qOvXgoycHDQMjZEupmRihJcTXC2uwsBc9N5lmUT12/dCuKcjcrKESzdnaeTqaqpQVlKCrraW2IsBrZs3xh9b/kGX4d/g9JUbRbZduXMfi2ZOEay9oZHRaGpuyv+mrZqYIywqpsR9TGoZ4rdte9F62GQ8f2PP12VlZfMA0ZV7j9Gjg3AXy1kmUBszE367rbkJAiNFc8UUpiQvh9+OXkWb+WvgG/ahVEJwVBxUFOWhpixcWbDguCS0MhEFAlubGCI4tuj3sHt4DEa2bsR/hsQllWufmoplV0p6IYRIBsucY32ZmJiS3xGlYeUa2Txfd+/erVQZFVdXV97nKzwHWEXaWJlSb/ll416+fInTp09XOAuoKtpQEW5JSbzEXFxW2aVJIzIy0FhNlO3QWE2NZxaJWydJlm5+vGRRSnomgqPj0aaBMT9Pam1qzH8X56G9J6b1LTp4RQgsOMXKx7GFla/KyMqGsZ42H6hS10C31JK3hVVmn8pKz8vl5eVcs8SXZRSHlZZ7nJ6A5vIqH10nCZFxibxcnK2HH0Kj49DMtA4/F23TsD5Coor+331sO/vbPLVzw5TB4gd7VJXQlDS8joiBb2IywlLT0VRbFExtpq2BsJT0Ivc1UlFCfGYWLIIj8Cg4Aq30RAOsznoFYlrjD5U8hMBeM20tLaiqqkBdTa1EGSw//wD07NaVZxTlbytrH0lgpeNYuTj2vgiKjEXbhqLXqV3D+giMiC5y349td/YNQpfmDSUeIGICQ8LRupkoeNWuZTMEBpcsGZ6aloYHz17i7pPn6FxoQB0rP5eengEDvZJZflWN/Y+wz122sM+11MwsmOhpQV5OFvUNdBBdyufShRf2+GFkT3wqIcFBaNm6Db/dqk0bBAd9mCKDmbPgO2RmZeL8mVOo36ABgoNEZd5DgoNx7/YtREdFol594f7fgoKD0aa16G/ctk1rBAYVHcTu6u6Bb0aPwmvrN4iL//B9x8o6XrxyFRoaGnyO088BizlKdk4ifHFqRCYRO3Fn2TCGhoZYsmQJHBwcUKdOnY/uExoaymu+d+3alZdcW7FiBQ/2sH+I8mAl2lgaf+/evXHmzBneaWDl5irSRpb2X9Hyd6xs3MKFC2FqagpjY2MsW7aMd3jKU3++Mm1gr03hGpr5tfAr7B3wLi+PXzjJSU+BjHzJL0tWXo7NTaTX6kPKc8iz85BX04Z2U+FGv0nJyiEvu9CF3HfiO5MRry4hJyMFBp2ES9cXR15RqWDunszUFP57cZnpKXh+7gByc3KKbHd6coeXmhO6vaydjKjkWcmAT2JUOF5eOMQzNKSkPnzy2t27iFb9RgrbXiXlgteXZW0pKJXsqIR6u+Px2UNQUlPjJQpZ6Qk5eQU8OrUPCsqqkBZopB4jV+T9kAQVrQ9BoHxBzm/h9OgqNA2M+fxQMnKi1Gd3y/sYv/o/CEleUfT6srJxpb0fWPabp/UzmDRrwy/uKKtpINzXg89TZNSgMfJyciAtUPo2w+ZOYu/NrNRkyCmWbK/b4+uI9veAWRdRBiLLcnpyYD0yUxKRnZmOidsu8CCTpMnJySGj0Od3bl7JOZ/W/7qU/4yOjcP4eYsxefRI/hr/tulv9O/RBW1bFC2xIUkqSkpITBZ9NqSmpfMLGcUd3foH/2ll74JD56+hZ8c2SE1Px9YDJ5Gdk4MxQ/oK115FeSSlijrjCSnpUFUqORozP2vo7BNrXLa0w8/fDOa/H3v4ChP6fMjwE4KyvByS3tejT0zPhIqCXImA1nPPQDxlJXXikrB5XL8y9yGEECFduHCB99fYxXw2BxcLfBQu/SPOo0ePMGnSJAwdOpQPstuzZw9u3RKN7i2Pe/fu8bnCWMnuw4cPl1mmW1wbK1P+jpWN27t3L8zNzXn5mpUrV8LR0bFc+1a0DVXR37sXHgX/1FTMql8X2mWckylJyyAlRzSnJPtpqKQodp0k7b39AolpGdg1bwyUFeX5bVaKjY1YVxGTXXHB0p6Xm1s2SpjSXMXLxW47f4/f7trcHO/wjp+rsfd+QmoaVMvxWlVmn8qKy8vB5bRYqErJoFk5AzzXUmOR+i4Xw2S1P7pOEjwCQ3m5uOamJvhuzCAkpYrmHGHneEoKRd8LykoKYrenZWRi2b8nsHj8UBi9z3SXFPvoeFzzC0E3Iz000lJDcpbo/yYpOwfKch+qGzBZuXnQV1LEETc/XnJORVYW1hExYBWS7GPikZKdw+czyi9BJ0ms9HU+Nh8O66cUpq+nixnzF/LB2zLSMnjx6G6Z+0jCuUev4OIXgl+nfsU/CxLf/73Z/43++4BcvtK2v3T2xNVnNti8oGjpSklRUWZ9KFH/PyEpGaoqJfun0bHx2LrvMB8AXbjs2cnLNzBuhKh/ImnZubnYfvUJv92pYb2CwBHLHkpMzYCqmM/eXTefQ0tNGSM6CjfotjhlZRUkJYnmG0tKTCxRblZbWwebtv3Nby+cOwut3mf429na4MiB/dDT1+dBJDUh2/t+7l52vVdFtWh7lZWVsO/gYT6/YnBICB7cus7LPD60eIxL166jaeNGyMrKgkKxzz9CmBoRF2OBHXYCz0aIsdrs7CS4LGw02JQpU/hcQjdv3uQp+NeulX8eDlY7fs6cOTh//jwP0rCgTVW3sTTKysqwsLDgo8O2bNnCJykuj8q0YePGjXy//KWs4FtpVGubI/DeIYS/ug6W2ywtJ4+0qCAeMGKSAl15cKh2j3EIt7rJ1/nd3IuMmFAeJGLbhcI6VLKKKoh4fRWhj49DUVf0nFPDffg8Oby9fg4w6DIK+u2GItJalInF5k5Ki/RDRlwYUkPLn9H2/9I0qI2YYH/Y3r+CO/9tQr2WohrkQW4O/Gd2VibCfdzRc8JcNO7SBzZ3Lhbs6/DoOlr2ETa1uHbjlnC3fAiXZ3fx9ORu1GvZgZ+shLiLOr5pSQlITYxD9/GzoV/XHG4vRPN35ObmwN/RGvVaVX19+Y8xa9MJzy+fgOOz+3hy9hAatOnIM7DCfD359oToCMjIyqHf5Lk8MORl+xpuVk/5vEX9pyyAjpExHJ+KnoMQWGm5V+f384CP06NrqNO0LdIS43n2EBMXGgANfSN0GTubz18U5S96r0b4uEGrlglk5YU9GWAl4+4f2g77RzcQ4GLL58ti2XBsDiKGBYPqNmuDXhPnwtvGkj+X19dOoWXvoegzeT4i/b0RFVSxjMr/h75Zc7w69Q+c71/gATYW1IoPDeABLt5eTyceHGozchpcHlzi6wwbtULdNt3QfsxsHvTMTBOdxAtBT1ubzzG09u89MK8nyhh0cHVH5vvJbdkEpg+fvcS/h06gbp3afB0LFim87zyw+wrF2EgfASFhOHbpFpZv/LcgK4gFhPIzhh69fIN7z17zAFE941r8edi6eODneVMxsl8PHD5fNBtKktqa1cVtaydcev4Wa0/fQvfm5nz9Gw/RSPPE1HQ8dvDAXRtnXHj2FvUMdD6MorN3R59Wws0Dx3Q1N8aRF/a46eCF3RY26GpWB+lZ2XANFY167NrAGKqK8vhxYCeMad8ETiFRYvf5HOTPSSTJhRBS9Vh/7erVq7hy5QrU1dXLNa/q7t278d9//+HYsWO878OqMHh6is7hyoPN6XrkyBEcPXoU9+/fLzMTqTJtLE3Hjh15xQd2rMGDB5e7r1rRNlRFf29pIzNeYq61libSc3N5VlFqjuhnWk4u4rOyEJImGljRRksTxwKC8Dgympeea6GuLnadJJ1ZPh23/5wHU0Nd9GjWAL+fvI2zz23x1jsIDWvp8WwiVnaO2X/3JS5aOqBNg9p46iw654xOTMEzF1/Es+96Jy+JTlDPLkTf2byUL4rycmhjVhd/HL2K/TefQApSfJ1nUHjBxeqgqFi8cfdDeGwCLJ1F5/ni9pGU2rIKWKNVFz9riuZkDMnJRGBOJsJzsuCTLXoP+GWnI/t9tQ6WMTRSWRsDlbRwOz2u1HWSwsrJsXJxWxdN5VlWXsHhOHnvOX7dexrdW4lKt1m7evOf4ranpmdg6LKNaG1eD2Ex8fAOLlnVoyoNr1+bl5eb0dQULXQ0eZbQvcBw7HPyRnt9Hd6vdopJ4Pd1iUtEQmYW5jRrgM6GurjsG4To9Ey4xyXhsKsvotIzcCugZNaJJJiZ1seGrTuwc/c+6Ovp8XWubu78QjXTrEkTTPpmLH5e8iPiExP5RWpx+0ja1oVTcHf7r7ysXM/WjbH51A1cefYGJ++94JlBMYnJ8A0V9VPFbb9n5YCV/53DoE6t8NTBDenv+1yS1KyhGV7Z2OPM1VtYtXknenQWXQ+yshNdX0lJTYNvQBB+/m4WenRsh9NXRNfbmKt3HmLEAGGC3+xzh33usoUNtmvXwBi/n76DQ/dfIysnhw+48wqNQnyK6LPst5O34B4SiVra6rDyCODrQmISYO0ZiIj4ZJ4RKoSu3Xtg365/cfvGdRw+8B+6dOuO5KQkeHl68O0pycl4/vQJ9u/ZjSePHqJ3v/4820hTUwuLFi9FZkYmbN+8gVB6du+GtRs349LVa7h28zbat2mDiMjIgoyi3j16wKyBKX5euhhtW7eGl7cPPDy90KRJYyxf/AOcXVx5Zt/nQLJZRFJf5JxENaKXyzJq8svosBPh9PSiabbisC+dwllD7HbhEVTl2Z89VuH9q7qNpWGpf/nHYo/L2lIelWkDy7Bi0ef8pbIdnXpD5yInM40HexqOX8HXxXvZIDNRdHEqwvo2Qp6eQ5zbS5iNEY1sZyXp0mPDEPzkLN8upHpfLUNauA+yU+JhMkQ0GWmc81PkZYtKH4RbXkDEy8tIDnRGveE/8HWJPrZICXRGVkIEYp2Eq5HKRoR9s2IrvN48h6KKGnpPFNXqffs+GJSRkoynZw/A8tJRXlpu8BzRXE8p8bEwadaWZ3AIibVx6KLf4frsLoybtkaLviP4yaz9gysF5dIszx+E1dWT/EJblzHf8vUsmNGq31eC1c/OZ1jPDP0mz4Ptg+voOnIC6jdvi9TEeDg9f8C3h3q5weLUf3h59TQM6pmjbf/haD/oa6hoaOPV9TM8Y6TrVxMFa695pz4w69ALbk/voN+s5VDV1kN8eBDPHmICnd7g9cVDcLx/CY27DYBxU1HqdLi3C9oMGgehdRk1BVoGteH+0gLjftkMWTl5hHg48eAQ4/LiPp6dOwiHRzfQb9pCqGrpYNDsZXweLpvbF9G850DUMhMu26XrlB+Qk5WBMDd7DPxxPV8X5PgaydGiDqGbxVXYXj0KvzdP0HfB73xdp/HzkRIbCfcnN/gcRcoakh0RWdh/W1bD2d0LIeGR+GfdKr7u4s27SH7fIWMBJLawOYu2/fFLQRbPa1t7bNl7iN9XKOx/+9i2v/DghRWUFBWw8rsZfP3Ri6LAD8sY2rL/BHafuIDahvpYNnsSklPT+Lq/D5+Fb1AINv2yULD2slJx/y6ciCuW9mhnXhdT+okC2MceiAZghETHY/vF+zh05wX6tGqMcT1FHTa3oDCM792xzNHvVc1MXxtLBnbGJRt3TOjUDF3MjJGQnskDQIyxtjq+7d4a56xckZObh4V924vdhxBCPpXCwYvK9PnY5y7rR1W2z8dGXxeerLqq2lia4n3N8vb5KtqGqurv5UvIysbJgCDoKsjzn6wEXXBaOhwTRCOxe+npoIO2Fh5GRmFxQzNoK8iLXSeUhUO7o7aOJu7YuOHQ9xMhJysDG+8guAaKzu1C4xKRlpnFR8DvuCYaBe8bHoN/bz6DrroKX89GyAtl3awxSE7LgJWbLw4vn8nXPbR15ecdjI2HPy4+tUFWdi72XX9c6j5Csc9KgWt2GqLysvEsQ/QeeJ6RhMz31TqupMbgWlos3LLSsEDNqNR1QmCfEUdWfocH1g7QUFXG8smiChbHbj8tdTsbFKSkIIdrL97wjKQntqLBTUJQY2WNOzTj8w210tPEsHq1WAEXXPMT/Q931NdBGz0tXPQJQkRaOuY2MysIMrFlZP3a+KWdMP2oHZvXIzgkFC7u7tj3z3a+7ubd+4h+X0pzy/rVsHd0wtmLl7Fl3V/881bcPkIa1rUthnRuhQsWr7Fp/kQY6WjCLzQSzx3cS90eFBkDZUUF7Dh3G9vO3EJ8cvnLLlaWvLwcDm5bi5sPnqC+iTEWThddezh6TnR9JTI6Blv2HsY/h04iPjEJq34UXd/yCwzG8P69oSDg521hqycP5aUwX7r749APojY/dvJG8PsgZ2xSGgKj4vhn7OGHVnydrU8wzr+wQ25eHs8IFUL7jp0wedp0XLl4Hj8s+QkNGzVGeHgYHj98yLdHRkZg19/b4ezogPNXb0JLSxu2NtZ83Ykjh9Ghc2f0HyRMthYz8Zux6NihPS5fvY5df2+DhoY63Dw8YW1jw7ezuYfMTE1x9MQpdGjfDoMHDsD9h4+wfee/OHH6LGZOn4YmjYUdzEhqDql3pU3IUk2wWs1//fVXQdmAVatWwczMDN9+K7qwXJiLiwuf9JeVHWC3+/TpgwULFvATZza6zN7engdSylsnul+/fvjuu+/4yTTLKPr3338xduzYcreRlavLb09Fni8bScbKw7FjsJFxbGGdgY8dq6rawMoPsE5K5zU3ebZNdZeZLvm6tVVpdA9hawP/vzJzJDdqThI0lWtWqaSoJOHmr6kKaoo1okJpgfCEyl+4Edr2DjXrvSuVWbPmrsnyED+5dnWV5i/MyLmqkJyeifo/7+LnSoUveH4K+ecwoQdWQF2C80ElpWWg9tyN1eI5E/K5YBlBCQkJvA/D9O/fn69r3LjkBPGsygIrrbZu3TpeeWHNmjWYPXs2L73NSrbZ2dlBVrZ85ywsg2jTpk2YOXMm35dVZXBychLbZyytjZaWlgXtqcjz3bp1KyZOnMjLzbHnxPpzLJvpY8eqijbkf1Y+6tWNl6eqCVouEE24XVNIq0m+vFdVuj9NVEqpphh8QfT+rym8959CTWK++whqDEfRhfyaQqaBaABnTZH1RrgBhVUhub8oQFZTaEKy8/NVpaSkZOib1K8W/Z/885hN9x34IHVJyUhNxq+DWleL5yyUan9WqK+vj+HDP5wUdu7cGXqlpKLq6Ojg66+/5rebN2+Ox48f83JzbESZlZVVuQNETIsWLXD79m2e9s9KAbD5iBo0aFChNhZuT0Ww/Vkw6sWLF7w2Njv5Dw8P/+ixqroNhBBCCCGEECKEli1bFsmIYf0XLS3xc380a9aMz8HKsPmE2G3W72vTpg2fi7W8ASJmxowZ0NbWhrW1NebNm4d27dqVOqFzaW0s3J6KYIGtWrVq8blvWfvZoMCyjlXVbSCEEEIIIaRGZBKJEx0dzTN7CmOpwSzrp6L7sRqpxScmK+1YQjwuC0gVzgiqqjaUF2USSRZlEkkWZRJJFmUSSQ5lEkkWZRJ9WZlEYYd+k3gmUa3Z66vFcybkc8YqOzx79qzIOhZUGT16dIX3y87OLjEpemnHEuJxw8LCimQEVVUbyoMyiSSPMokkizKJJIsyiSSHMokkizKJvqxMoi0PHaEkwUyi9NRk/DygVbV4zkKp9plE4rDRUx4eovks8rE0/crsl5aWBmVl5XIdS4jHZdlAhTOCqqoNhBBCCCGEEFJTxMfHl+j3sKBLZfYTp7RjCfG4LGPpY3MJVbYNhBBCCCGEfDFBIhMTE157Waj9hH7c+fPnV3kbCCGEEEK+BFLSMnyR5PEJIZLXs2dPvgi136d+XEkdixBCCCHkcyMjJcUXSR7/SyP9qRtACCGEEEIIIYQQQgghhBBChFcjM4kIIYQQQggRi2X6SDLbhzKJCCGEEEIIIeSTkZaS4oskj/+loUwiQgghhBBCCCGEEEIIIYSQLxAFiQghhBBCCCGEEEIIIYQQQr5AVG6OEEIIIYR8PqSlRYskj08IIYQQQggh5JNgxcVlJFgRTgZfHurlEkIIIYQQQgghhBBCCCGEfIEok4gQQgghhHw2pGRk+CLJ4xNCCCGEEEII+TSkpaX4Isnjf2kok4gQQgghhBBCCCGEEEIIIeQLRJlEhBBCCCHk8yEtI1okeXxCCCGEEEIIIZ+EjJQUXyR5/C8NZRIRQgghhBBCCCGEEEIIIYR8gSiTiBBCCCGEfD6kpSWcSURjrAghhBBCCCHkU5GWkuKLJI//paFeLiGEEEIIIYQQQgghhBBCyBeIMokIIYQQQshnQ0pami+SPD4hhBBCCCGEkE9DWorNGyTZ439pqJdLCCGEEEIIIYQQQgghhBDyBaJMIkIIIYQQ8vmQkpHsnETs+IQQQgghhBBCPglpaSm+SPL4XxoKEhGxUuIzIKNQ/RPNMtOzUZM8cYtCTSJTwz4UZ3avh5pETb5mfQQbqimgJjHTVUFNkRf0CjVJbnw0apKIxy9Rk0TYBaOmSM3J+dRNIIQQUgmtfl8AdRXlGvHaRZgPQE2idmk9apLB51agJslLjEVN0vjnH1GT5OVkoqbIqWF9EtmcDNQkuSnJqEkkWXpMEqRyas71TKncmtNWUnk16wolIYQQQgghHyMt4UwiSR6bEEIIIYQQQshHSUtJ8UVSpCV47Oqq+qeKEEIIIYQQQgghhBBCCCGEkCpHmUSEEEIIIeSzISUtzRdJHp8QQgghhBBCyKcrLyjJEoMyX14iEWUSEUIIIYQQQgghhBBCCCGEfIkok4gQQgghhHw+aE4iQgghhBBCCPls0ZxEVY/qZRBCCCGEEEIIIYQQQgghhHyBKJOIEEIIIYR8PticQSybSJLHJ4QQQgghhBDySchIS/FFksf/0lAvlxBCCCGEEEIIIYQQQggh5AtEmUSEEEIIIeSzISUjwxdJHp8QQgghhBBCyKdBcxJVPQoSEUIIIYQQQgghhBBCCCGEVHNpaWl48eIFnJycEB8fDy0tLbRo0QI9evSAiopKpY5JQSJCCCGEEPKZzUkkwYrKNCcRIYQQQgghhHwyMlKiRZLHr46Cg4OxceNGnDx5EhkZGTA2Noa6ujqSkpIQGhoKeXl5TJ48GStWrEC9evUqdGyak4gQQgghhBBCCCGEEEIIIaQaun79Olq3bo309HRcvXoViYmJ8Pf3h6OjI//Jfmf3ycnJQbt27XDx4sUKHZ8yiQghhBBCyOdDWka0SPL4hBBCCCGEEEI+CSkpKT4vkSSPX93o6+vzgBDLHhJHSUkJ/fr148u6devg6+tboeNTkIgQQgghhBBCCCGEEEIIIaQa6tKlS7nva2RkxJeKoHJzhBBCCCHksyElLSPxhRBCCCGEEELIpyEjLSXxpbpjpeVcXV357bi4OIwaNQotWrTAgQMHKnU8ChIRQgghhBBCCCGEEEIIIYRUc35+fli1ahXMzc357xs2bEBwcDAmTpyIZcuWwd3dveYEiVjDDx48WKXHjIiIwL59+8RuYxM6sbp9lfWx/T/2uFWlrMcQog1lSQl0RKLXa7zLyy2xLdnfHrH2d5Hka4N3796Vax9JevcuD8l+b5AaaC92e4LLQyQ430daiEvButyMZCR6PCuyTii5WZmIsH+CBH/xjx3y6hZCXt5AUrDnR9cJJSs1CWFvHyE5zE/scwl6fhXBr24hNSq4xH4hr29DaLERYbC8cw1RoUXbU1hkSBCsH92p0D6SEhHoB+sHN5AUFyNmmy+eXj4Fq3vXkJacWLDew/Y13yclMUHg1gI+rk54cfc60lJTSr2Pi81reDmL/h8DvT1w59xxvLa4h0/B1eYVrC3uIic7u9T7vH36AKH+PhXaRxLy8vJw+6UtHlg7iN1+95UdDlx9wBePwNCC9c4+gTh60wJPbYX9PMvIysa1l/awchdfG/fCMxscvPOcL2GxCcjKzin4nS0Pbd0EbW9iRiZueQbAIzpe7Hb2ffYiIBxnnLwRmJCM6NR0nHb0wgUXX/jGffj/E0p0ZiYeREYhLD1d7PasvDy8iInF1dAwxGdllbquxpOSBqQluLDjE0LKxdLSEg8fPqzSV8vKygp37nw4Jyts586dfNLeyvrY/h973KpS1mMI0YaPycrOxvWnVnhh71qh7Tau3jh8/SHeuHpBSBfOnMSpo4cRGR4udntychJuXbsCF6ei/fzUlBS+36fywjMQtx28kJObV2LbEzd/HHtuj4fOvgX96pv2njj63J4vp15W/ppHpc9FX9nhoY1zubcfuf0Uh24+xnOHil/A+n+lZ2bh6kt7vPHwF7v9wrO3OHjXEv4RRftZr9x8cfjeSzj6hQjUUuCVgyuuPHqBjMyS52dvXb2w/+ItPHj1tsS2h69s4ewt/vlJ0oPHz3D7gQX/mxdn8cwSB46d4ouTi+h8/szFKzh04gwePX1R5BqRkLzDonDR0h5RCckltoXExOPQ/Vc4/cSmyPrnLj58n7jkVMHamZySiou3H8LOxaPEtvjEJOw/fQlHzl+Ds8eH/qlfUAhOXb1dZJ3QXrgH4KatB7JzSl7ve+LqhyNPbHHPwQt5eaK/v7V3MA4/fovn7sK+f4MCA3Hl0gWEh4WJ3f7G2gqXL15AVFRkufeRJFd3D1y4cg1x8SX7qN6+fjhw9Dhfbt69X7DezsERZy5cQmRUND4XrEfGkn0ktqB6s7CwQM+ePSEvL1+QVfTvv/9i5cqVmDJlCh49elThY36y5xwZGcmfQFWSk5ODgYGB2G3Pnj2Dt7d3pY/9sf1jYmJw+fJlSFJZjyFEGz4mxvYWQi0OIcb2JkLu7ymxPSsxAumRvoh8fQFhFgfLtY8kRT0/iji764h+dQqxb6+U2J4R44+MKF9EPP4PcXY3eBAr8OJKpPjbIMbmIuIdhe2cOR75HWFWd+FyagMi7Z+W2J4S5ovEQHc4Hv4d0a6vS10nhHd5eXiz83tEOjyD7d7lSAgoevKfl5uDpBBvJPg6w3rn90gO/3AC4PfgFPzun4SQUpMS8fu3o/lF/z9njkFcVITY+107shvXjuyt0D6SEObvjR0/TIX9swfYNGcssrMyi2xnQaAgLzc4PH+IjXPGIi83F/dPH8DlPZvhZGmBPT/P5euEYvfyKTYtnYtnt6/hr3lTxN4nMT4Wh7etgZXF/YLffd1dcOX96y2khxdP4ujmP3D/3DHsX/Oz2PsE+3ji1M4N8HKyLfc+krJy7ylsP3UNqw+ew6bjJb8D9ly6Awdvfzh6ByAuUdT5OXv/BUYsXY/Xzp4IjiwZaJSk6VuO4PjDV1jwzylcel6yU7vj8kM4+YXwJTUjE7l5eQW/s+CSvW+QYG1lndXJFy1w1ysYc68/g31Yyddq+f3X2PjcDq5RcUjMyEJadg7couPxNjQKEy484uuFkpqTg/n2DrCMjcVCBydEZmSUeD4/OjrhTHAwvFNSkZmXJ3YdIYRUJQcHB1hbW1fpMdXU1KClpSV227lz55CcXPJiX3l9bH8XFxe8evUKklTWYwjRho+Zt2EvDlx7gB+3HcTRG4/Ktf2fszcxadVWvHXzRniM+EEXkuLh5orTJ47Az1f8RdJp34zCjauXsGDmVLx8/qxg/cG9u7D77634FA49tcWqSxY4+MQWy858uMCXLzA2EU7Bkdhx7zVWXrTg63wi4+AcHIm3fqG4+EZ8AE9Sftl3Bn+fv4M/D13EtrO3yrXd2TcI9l4B+P7vY/j34l1B2ztly1GctLDG3H9O4dqrkoOsPEIicfqxNRx8PwwEXHvmNubuPAU7nyCxwQRJOH/vKeav2YmjV+9jyq8bS2yPjI2DlaM79l8o+poHhEbg578P4rG1+MGwkrLp711Ys2UHtv27FyvXlGzv0dPn8MbWHo4uboiOFZ0fu3l6w97JBSvXbMCfG4X/f2N/65Gr9+OWtTMG/LYbmdk5RbYnp2fCyT8Umy99GOjwz/Wn+OPkLdx964qJm48iV0wgVxJGzFqMczfvY+L3v+LhC6si29IzMuHo5gUre2cMmb4Qz6xtERAShsHTFuLBcyvM/209Xr4VP6BQklgA6Jcz93HQwgY/HCv52RAckwinwAjsvPMKy0+LBoeGxSfjhUcgzlg6CdZOFuT5athg3LtzB8MG9UdiQtFBtefOnMYvy5bC4uEDTJs4AampqWXuI0lv7RwwatI0XLt1BwNGjikRlLWxtcOlazfg6OIKX3/RtTaLp88xefZ83Ld4jN5DRyKjWD+R1EwZGRnIyRF9bgUGBvI4S6dOnQrOlSvzd5bFZ0RHRwejR4/+1M34IsW7PoHZpE2QVlCG1+GFPKhSuGa/Tush/Cdb733yp3LtI0mpwY6oP3kn3uVkIfDSSui0L/q+Mew9l/9kgSKWVcQCH+9ys6GoZ4q87Awkeb6AVquhgrQ1KzkeuRlp6LB4N9KiQ+Bx6V8YtOld5D6Nx/7If8Z6vkWs+xvoNesidp0QEgLdoG7SCC2nrkS0qxXCbO5Ds16Tgu1ySipoPkl0IT3oxTXE+zpBzag+koK9oKRtAFllNQjJ9vkj9B8zCaPn/IC7Z4/ybJAhE2cUuQ9b16HPIAR6upV7H0mxeXgL4xatQIcBw3F2x2qeIdSiy4f3g1nLdnxhdi+fi/joCCRER6JWfXPUMm2ImPAQeDvaoFHbzoK01+L6Rfy64wDMmrXEX/OnICzQH7Xq1i9yn6tH/8PX0+bxDCKmZcduaNa2E36dLvzn+bNbl/Hn4YtQ09DCsjH9eWaQrJxckfvcP38cA8dNrdA+ksIyiGyObUVWTg56zVuFX6ePKXGf7q2aoK6hHrq2bMx/33DsEi5uWo4OTUVpyUKJT0lDdGIyHm/9CYGRsfh+z1mM7dm+yH1UFBXQpWkDNKtXC+a1RYM+di2axH/O23kSk/qKTnqE4BwZh0a6mtg+pCssA8Nxzd0fbWrpFmwPTkzBq6AIPJ4xEopyH06n1vcXtXG/jSu/TzN9bUHa+yo2DsMMDTGzXl2eFfQ0Ogbj6xgXbH8bn4CcvHfY37Y1pKVE9ZVt4uJLrPscSHreIJqTiJBPq1mzZvQn+ARYNoObXxCsj29HVFwCJv62DTNG9i9z+6ZjF/Hq6FbUryV+MKck/bFuEzat+UPsNh9vT+jpG+DA8TOwe/sGp48dRreevRDg5wsFRQUYVnCy56pywdoVN5dOgpqiArqtPcwHzMiwLNb3vu3Rmv9k6wdtFg2uWzJY1M+7YO2CPIEzMizeuuDt4Q08Q6f/j+vx08ThZW7/+4dpfBsLFB29XXIApKSw89DktAw82PgjfMOisfzQZXzdVfR65vtj8jCoKysWyYLfe/MZHPf9Dn1N4fqpJ248wK0962FipI/u0xYjPikZWuofHn9Yz85o17QhFq77t8h+/56+iu8nfY3k1DQI6eqtu3h2+wpkZWXRoe8QbPrrtxL36dS+LcxM66FPj27893WrfuE/g0JC8etf6yG0y5b2WDdtBMZ0a43lh6/imbM3Brb9cN2iSR1D/Dt/HHr/srNgXXhcIpqYGPJtQVHxeOXhhx7NzCTaTp+AYGiqq+Lyf9vw1skN/x49iwE9PvTlaxnoYe/6lfz2wbNX4BcYAnVVFWiqq6FNs0Z83xOXb6Fb+6LvdUk798oJt36ZBm1VJXRZ9R/PJpKT/XB+Pq1XG76wLKK+a0WZm6M6NkUzY33suP1SsHbevnUD3/+4BDPnzMX2rZvx4P49jBs/oWB7RHg4zMzN0bJVa0RHReHxo4eIjIz46D6SdPbSZezZsQUD+vTG7IU/wsbWHp06iK7/5GvcsCF6de+KIQP68d9fv7HBtvVrMGzQAEyf9x3/n2to1gA1nYyUFF8kefzqrGvXrvj999/Rvn173L59G0OHDoWMjOh/7M2bN1izZk2Fj/nJs6eePHmCP/74g6dJlVWa7sGDB/wFYE+WTci0ceNGbN++Henvy6oUL7lmZ2eHP//8ExcvXqxU2yq7P6v7t27dOt4+Fs2riMLPlb0uT59W7qSpvG3IzMxEUlJSkaWy5dtkFFUgJSUFOXU95GaULCuV4P4CYRaHoNm4e7n3kYS8nEzIKKpBSkoa0nKKkCrl34AFrRKcH/AAkrSsHGoNWozc9EQo6psiJ024kQKZSXFQ1qvNbyvrGSMrpeQIvJzMdAQ9v4IQy2vQb9mj1HWCtDcxFip6oguTKgZ1+O/FZSREI/DpJZ7hpNNI9IUW8voOTHp8DaHFR0fB6H3QolZdU8RHf0ghZrIyM+Bpb4O2PfqVex9JSoyNgr5JPX7bwKQ+EmOiStzHz8Ued47vg4ysLLT0jTByzhIYmJgiNTEe9Zu0QoKYfSSFvTa16pny27XrNUBcsdfK180ZRib1oa4lzIX0srAADwv2MHq1jJGcUDQTxOrhbXTqP7RIEKisfSQlOycH6irK/ERASUEBsu9PCAob0qUdXji44Yfth7D99PWCizjsYsO5By+QIGCZhMj4RDQw0uO36xroIC655Gd+r5YN8cLFG5M2HMS1Vx9GQCampiM+JRXGuuJHjksCKx1XT0vUGa+vpY6o1KIl3Fh5ue51jfAiMBxP/EL5xZrC+/rEJWFAgw9BGkmLzcqCsZISv11HWQkxxUrHhaSno7OONiyiomH/frSbuHWEECIJp0+f5n2bj9VHZ6Xp7t69i5MnT/L7sn6Rl5cX7/+xdaWVXGMdY9Zne/mycheVKrs/68eytu3atavC2Uvsud6/fx/Hjx/H6tWr4eFRsoRQVbWhqvp70QlJqPc+0KOvrYn0zMwyt8ckJMHEUA+hUbG49Ogl0jKK7vMpRUVEor6p6CKZaQMzREWKzlFPHDmImfMWCtaOrJzcglJxrLwcu2CqrqTI+8jGWuqITy05EvjqW3f8dtECX7UTDQDKd9nGHSPaNJJoe9nf+eANC5y6b8kDP9rqqpCWloaKkiKki03y/bHtObm5OHzrCZYVCypVNb/waF4+7qaVEyLjk2FqJBrw06CWHg8alSU0JgGtTY3hGhjGM48yBSotzSoAsAARY2ZSCxHlyMJj2UM927fkr7XQWHBIUVGR/9RQU0N2sdepX6/usHNy5llGf2zYUrCeZbXv2n8Yq35aLEg7b71x4SXk/CJiEBGfhAbv3w/mtfQREV/2++G3CYNgZqSHuOQ0tDWrU659/l/hUTEwq2fCbzesb8J/FycuIRGvbB0x8avBaNOsMX5dMAPB4ZHo3KZlqftUJfb5xUrFsYV9rrGFBYgYE11NxKaUDFxes3HDynMP8FX7D8E5oUVGRPDvAKZBAzP+e2HzFy5Cu/Yd+HdEuw4deICorH0kKSIyCmamousr5g1MEfH+uyufeYMGyM3LxZkLl9Fv+CieaTR0YH8cOn4S23ftRXxCIhrUF11PIjVbu3bt+Dkgi4uwc8AtW0SfrSxmwj6LWSm6GhUksrGx4Sf8rEzcpEmT4Okpfu4UljL166+/4vz58/xkacSIEfj222+Rm5vLgylsoqbiJddYKv6gQYN46hULtFQ00FPZ/VnHh/0hUlJSEBUVhY4dOyKsAjUq2XP97bffcOrUKf5HZXUEWQeiIirSBhZE0tDQKFjq1KmDysrLFZ0I5KTGQ0ZBpcT2jNhg5KQnIjslrtz7SIK0rALyMlMLAkHv3pUsvZWXlY4Ii73Qaj0ccuqikzMlo8bQ7zEDsirakJYX7sRLTlmNB4oY9lNWUbnEfd7l5iAl1IdnPKXHhpe6TvD2Jsby34vLzUxHcqhoTpKMuEg+31JudgaCLW8gKzkBkY4vBGuvqroGEmKiCwIaquqaRbbfOX0YsvLyeHDhBBLjYuHy5mWZ+0iSspoGkmJFj80CROz34lgQiM1NlJeXywNDSiqqGDp9AUZ/9zP8XO2hqSt6TwuBvVYsqMawABH7vbDTe7YhJzsLNk8fwtfNCeFBAfiUpCDFA4NMQmw0VAq1l33nsCyi8EA/uNu9geubV/w98bF9JElOVhYp6aLHZSd/LGhU3MJxQ7Dvl/l4vn89bluKyrupqSjxgNH1528wdPFaCEVTVRmR8aILU7FJKVBVVChxnz+mjMC+H6bg3G9zccf6Q+36M4+tMbpbWwhJXVGeB3sY9lNDUVTrNx/73S0qHrc8A7Hfxg1bLUVlHNhcRFteOOCvPu2LjPyVNDVZWcS9DwyxgBH7vch2OVm8iYvHq7g4/OPjh2th4WLXfRb43EEyElw++RgrQmqUI0eO8LJzrOzFgAEDSlw8zMfuM3fuXBTlvfwAAQAASURBVN4Pi46O5n2xxYsXQ1lZGZs2bSrojxUuuXbixAn88MMPvB47m0+oouXFK7v/mTNnMGPGDCgoKPC29OrVi58nlFf+c3VycuKla9j+oaEf5g6syjZUVX9PS02FZwgxqekZJQaniNuuoaqMpNR07Lt0B2cfPMc3v25GdaGhqYno9/NLsIt/GhqasHz+FLExMbh45hS/tsDmK5I0NsiElYpjS/6AE3ZxlYlOToWGcsnzJe/IWMSmpCMi8cOAG6+IWOipK0NFoej5SlVjgR9HnyC4B4RASUGe/33588jNKzGHUmnb2fnrd9uOYOHogahr+CFLWxLiU9Lh5B8C3/BoaKoqITJBdC7KzkkLZwx97PyVZY8cffAKxx+9xqwdwpRGV1ZU5NlDTER0HLTUVcvcZ8uR84iMjYeFtT2sHN0QKmBZafa5nl/2KiklhV/jK2zG5AnYt30TXty9xucnyg9gf//zKnwzaiSaNm4oSDt9wqJ5CbmElHRR3+R9+UAWMNJ6H9D4GDUlRSwb3Q9/TR6Kt95BMHw/oEyStDTUEBktGngbHh3Dfy+OlZdbunY7dvy+DIoKos+M0UP68d8TkpJhpC/Z/7OCz7KgSL6w2ywJIyNL1D+NTEyBpnLJ19c7PBZRSakIF6iMoziamloFcw2xABD7biiMBT9ZoOiPNWvh7OQEAwPDMveRbHs1EBklur4SHhkFzWKP3aFdG+zZvgWXTx9DHePaPGsoJCyc90tdXN2goqyM6JiSA7lrIlYJQ9JLdbds2TI+0Igl3piYiILJLAZQmfmIPnm5OTMzM95pYNiIJltbWzRqJH7kS+PGjXH4sCgF0dnZGdOmTcOoUaN4YGnOnDkl7s+CRezFYsElhp3kVURl92edF9bRYNE8Jisri49OE9fG0rA/LOuwME2bNuXBsaVLl5Z7/4q0YcWKFUWOzf4Olek4qNZtieBbOyCrqg1ZFU1IycgiJdgFClq1IKeqjXj356L10tKId30K9Jsjdh+hyGsaIeLpAR4sUqolGrWQ4mcDZZOWPIgUdHkVlGo3Q1qoK7KTo6Barx1S/N8iKz4M8c73oN1mpGBtVdTS5yXnfG4dRGKQB/Sai9KzI+wsYNi2H3Iy0hBhawHV2mZIDfdHnJct9Fp0L7HOqMNAQdqrWa8ZXE5vgu89LR7sMRs6k48QirB/CqO2fXgWUaynLdSMzZDt7YAEf1eo1zHnmV1srqLcrHSkRn2oAS1pLbv2wupZ45CVkQGLq2ewfOcRPueQj6sjWnXpCU0dPUSFBiElMR6Z6WmICQ8Vu49g7e3WF6e3/oFgb3e8vnsVgybPQWxEKA8YmTZvA19nOx5EqmPeBK/uXEZcZBift8jJ8jF8XeyRnpoC81YdBGtvh94DsPuv5WjRoQv8PV1hYtYIAV7uPMupjqk5jOubwd/TDbGR4YiPjUFKUgKiwkLw5ulDxMdE4c654xgweiLk3k/GJ2ktu/TArpU/QM/IGMqqapBXUOQBIQNjE6hpavEMsgBPV8RGhEFaRpoHh8TtI5TGdWvj+20HeS3qTs1Fnau7r+zQs01TPorw6E0LXl/b0sENrRuJst+6tGgMQx1NKMjJwd1fuMl3DbU0kJiWjj9PXOdzDA3p0IKvv/zCFmN6tOMXHU5ZWPELCLetnTCsk2h7/n1urPkeQmppoIOf71thj7ULLHxDsKCjqLzRbc9ADGtUF030tHigqLmBNrSV0hCfnsWzi7698gTT2zTi5ena1dJDYz1hsp86amthgb0jst+9w+3wCKxr1oTPU+SclITO2tpor6mFK6HhaKiqykvRJGVno7eubol1hBBS1VifbetW0XwTr1+/5hlCpu9HwRb3zTffYPNmUSDByMiId3rZz1q1avEBhuPGjSty/7Nnz/I+Yu/evfn5ZmnHLU1l92cD+lj1Bxb0yh/N6ebmhhYtPnx3lYUNfGSjPxl2cZX111jGRXmVtw1V1t9TVoK8nCxW7D4B35BwDOzchq+/8vgVRvXpInY7G9DSvqkZmpua8O/3B1bCzpNy+8Y1uLk48/4wKyHXtn1HHvgZ/vVoNG7aDA52tvhn22ZYPn2MSdNnQkZaBioqKnBzcUJ6Wip8fSo/p3F5KcnLYcfkwQW/92hcFwuO3YKhhip01VQgJyOD197BMNXXgoGGKq68dYeemgpkpKT5/EPrx4mqHZywdMCETs0l3l4TA13sXvqhzHaD2gZY8u8JJKSkoWsL0bnondf26NO2GQ8Sids+cPEG9GjVGJZOHgiKjMGgTq0k1t525iZ8yRedkILVp27x+YWG5p+LWtphTHfRYKQ7Ni544xnAgwnaaio8y52VQG5laozUjCy88fgwn64kDevZEd/+tgUtG5oiOS0dhrrasHHxhJ6WBurVNoRnQDBuPnmNgLBIHLx0B3PGDoV5XWM4evoiKCySZ+0lpwlXcq5D2zaYt/hnHrRu1lh0TY8Fg9q2agEtTQ2cuXgFySmpeGvviMbmogyM0VNno55JHT4vUVh4BJ/jRdIWf92n4HZ6VjaWHrwM54AwnH32Fj981RvB0fE8YNShYV0kpKbjkqU9opNSePbRV51b8kyye7bueOMViKS0DHRtXLHvnMpoam4KR3cvrPnnAJ5avcW3Y0XXoi7efohxwwbwLKEBkxdgzsRRuHDrAdq1aIr2LZvyEnOhEVHYeeQ0rh/6UDJPUhTkZLHz22EFv/dpZoq5B66ijq4GD3YrysvitVcQ6ulpwUhLDVfeuPJMo6a19XDmpWgOItfgSNyw9YB3RAxOWzpicnfJfTbk6zdgAGZMncyzgY4fOYzrd+4hOjoKXp6e6Na9B2JionHr+nW4ubnC1cUZ/QcOQmCAf4l9hDJ04AD8tPIPjBgyCPceWWDT6t/h7euHtLQ0tGrRHJavreDm4YnIqGi4eniilqEBduzaCz1dHTRu1BAXLl+Dj58fDA2EGzhMJI8lnLAKM7q6/19A+JMGiQpPOMq+TNjJW3nvm/97afux0WDsnyQfG6FVEZXdn42YKJzCnyJmFEVZWPk81klhWVOsDRXdvyJtYK8fW/5fRj2mIsb2JnIz02AyYjlfl50UAzlVHX47MyaYl2iTVVJHvdGrSt1HKIb9FyHe4RakVHV5thCTlRgBpdxm/L9CUd8M77IzkRntz9KNCrazgJF+t6lQMxNmfp98redsQDArG9eiB4y7f8XXpUaKAil5OdlIDvGGFCs5pVsbtbsOF7tOKDLyCmg7bxPC3txH/f4Tod+iK89mSo0UTTjPglosGMRK+Ok0bAvjrsP57fw5k5R1a8N0gGgOEiHoGdXGwnU7YfvsIWb/tgF1GjTk2SAsMMT0/uobvjAGtesW3C6+j1AatumIUfOXwdvxLb7b9B/PJIqLDC8oIcfmIArycoOCsjJGL1iOuo1bIMzPiweVDOrUw/gfV0FaTFkySRkwSlSbNyzAD3/uPckvfrDsm/ygz8yfRMHsYD9vhAb4wbx5ax5EYkvrLj3h5+HKM6KE8s2CZbh39ihSkhKxePNevi4uKhwa2jrQ1jfE3N9FF67cba0hLSvD3z/i9hEKyxLaffEOr+/MsoYYv9BIdGkh6py5+AYhOzcX3Vo1wYzhffm6bT9Ox55L9/joxLPrlwna3jMr5uLg3ecY3KE55gwRlcH0ChGNwmLtZMEjdoFpYp+OfGFiElMwvncHKBfL5JE0BVkZHPiqF664+WFG28YYYCa6oOb3PhtKVloaGwZ0wgUXX+irKmFyS3NEp6WjZz0j+L+/DytTJxQ9BQX82aQRnsfEYol5A5ipqiI+Kwvh77PNNOXl8IOZKS8t11xdHV/XMoKctHSJdZ9PJpEEs30ok4iQatnnYxUg2Mj0iqjs/qxvlb8fC/BUps9WuK+Zv39FspHK24aq6u8xJ9csxX+X76Fn2+aYP1p0QdcrKKyg7ypu+67l87D34m0+f+LxNUsgJD8fbxjXMUFGejqfV4LJD/ywiymHT53HxbOnMGHqdIwaN56v79JddH5iUrce5v8gbHuZlSN74MATWySnZ+LAzBF8XWh8Egw1RZkkXuExiE5Og46qEk7O/zB/p5ayEro1/BAMEcp/y2djz5UHMNbTxsIxokGJfqFR6N6yMZQUxG9v26g+UjMyeUYSK68npHMrZ+PQvZcY3qklZg8WDb70CvlQits/PAY66qIKJyxgwOxZNBH77zzno8oPLP4wL6kkLZ46BkqKigiPjsXF7aJ5tSJi4vh5cn45Or/QcHRq2RhOXn583a6Vi/jPt65efE6ixvWFez9sW/sH9hw8iuycbGz8cwVfFxQSgiaNzKEFDbh5eiMpORktmjXBrCkT+fbmTRohJTUNji5uqF9X+Pdut6am+GPiELz28MeZn7+FpooSQmJEQSKGDbRjWUf9WjXiPwe1bYKUjEweVDIz0sXGb0dCRkby2eWs/3z1wA4cuXAd08eMwLQxoms7nn6iaSWys3MwoHsnBASLqgfVNhSV/XT29OGfy2zfzm3KP4Chqqz8ujf2P3qDhLQMHJkvmjc3LD4Zuu//v1gWUURCMnTUlHHuB9Hnb0xyGiITUtCijiHc3vcRJa1xk6bYtHUHnj15jN37D6C2sTECAwIQ9j67NyM9Ay7OTjAwNMTdh4+hpKQkdh+hsHmF2DVje0dnXD51jGc6JSQkIj7xfTnx0DA4ubhBV0cbN86f5uc5q3/7BYeOnURQcAh+WDAX3bsIMz+1pLF/P0n+C8rUgOIRDx8+5JnpAQEBvCoZm3aGlZtjZYgLl2ouL6l37IzuE3j79i3++usv3Lp1i//OSsaxzCJWRq6s+06YMAHz58/nI77YC8F+Z7WpWekBVpKApVWxeXhYihUbWcUCJiz9asOGDRg7dmy52leR/Qs/blBQEN+vb9++on9ce3u+FO7wlPW6jBkzhmdUsRFeN2/e5M+blVnIfwxxqqoN7LmyMgTNl1yEjELJsmbVTWZ6zRr1bGIu+TTfqiRTrK50dTeze82qrZqUUbI8WHVmqFY1FxiEkpYtXHDp/zUyWVQ2p6bIjReVPKwpIh4LN/lpVYiwEy6b8v/FspUGWL5CYmIi1NWFC4p97Bwm1vIKn7BXYo+Tkgqd7qOrxXMmpLrbvXs3EhISCsqD9+/fn69jVSLKum+9evV4H4xdADl37hzv77GScIcOHeJ9QNYRZnMTzZw5k2fl+Pv780wa1jk2LucFm4rsX/hx7927h+nTp/P9WBtZX42VCGcX5Mr7uuzZswfNmjXjwQqWYeXo6MirWeQ/hjhV0Yb8z8roR6f43IY1QYS5KFuqplC7tB41iXIjUbZ0TZGXXLPmUZQzEW5AYVXIq1Xy87m6yrEUTTdRUyi0rvgcIZ9S2vMbqEkyR/2MmkQ959OV2auopKRk6NVvWC36P/nnMQ+c/KGiJrmSj6nJyRjYsn61eM7isGllmjdvjrVr1/JyyaqqqgXnj126dOGBovbt29eMTCJWUo1Fu/INGzaM/5HLc192MswCSoy2tjaWLBGN8GGlCL777jt+u27durx8HYuqsZNvFiypXbt2udtXkf0LPy5rK6stzU7QWcT/2LFjpT6v0rASAazMgrW1NX7++WceMIqNjS14DCHaQAghhBBCCCH/jx49ehTJzmH9FQMDg3Ld948//ijIjGnbti3v7+R3fFn/jBk6dCjv87B+2y+//MJ/Fq/P/zEV2b/w4w4ePBjPnz/HixcvMHLkSH6c8gaI8k2ePJk/ZxacYkEjNqCv8GMI0QZCCCGEEFLzPHjwgM/fuXDhQqxfv57HLfJ17tyZl2yuMUEifX19fjJb+ISX1adm2TCFsZFVrFZz4fsOGSIqqcOwaN748aLURB0dHYwe/SHlmo0AY5N55hN3fJaWz0ZdiXvM8uyff9/Cj8ue29SpU4vsl19vuqzHnThRlHrLTv4LdxDyn5u4Y5WnDYQQQgghXwJWapUtkjw+IaR8WrUqOp8A669YWlri0qVLRdazuYDYnKqFsQyffA0bNuQLUzyIwh4j/3HYQEJxx88vhybuMcuzv7j2sYF8hefTrcjj5s891KtXL77ky39ulW0DIYQQQsiXgBU+YmVAJXn86iwmJqZg4FXxc82MjAx+Dlqj5iQqjtVKZmUFCqvI5J2VOT6ba4hNTlmex6xs+yryuMWzpqqqDYQQQgghhBDyqampqZXoz+RnCUnq+NnZ2SXm6yntMSvbvoo8Lgt4fWz+I0m/RoQQQgghpOZq1aoVrxzG5rMsHCSKiIjg5YvPnj1bs4NELPuleKZOdTp+Zfev6H6Fs6aqqg2EEEIIIV8EaRnRIsnjE0IqrXDmTnU8fmX3r8rnJenXiBBCCCGkJmNZRDISzSSSQnXG5vlkmUSsdDGrssaqjC1fvhxHjx7l55B9+/at8DEpBYUQQgghhBBCCCGEEEIIIaSaY9lDt27dwoABA/jUNK9evcL169cxe/Zs3Lx5s1LzVFarTCJCCCGEEEL+L5RJRAghhBBCCCGfLZbpI/0FZxIxSkpKWL16NV+qAmUSEUIIIYQQQgghhBBCCCGEfIEok4gQQgghhHw2pKSl+SLJ4xNCCCGEEEII+TRkpEWLJI9f3ezduxc7duwo130XLlyIJUuWCBMkSkhIwO3bt+Hn54fff/+dr7O3t+eTI0lT55kQQgghhBBCajR/f39YWFjwuuazZs1CRkYGfH190axZs0/dNEIIIYQQQr4YnTt3xk8//VSu+7Zr167Cx69UkMjFxYVPjMRq37GOQ36QaPfu3ejVqxemTZtWmcMSQgghhBDy/5GSEc1LJMnjE/IFuHTpEqZPnw59fX0YGBjwIJG8vDymTJmCM2fOoEmTJp+6iYQQQggh5Av0Jc5J1LZtW75ISqWSp5YuXYr58+fzLKLCFi1aVO60J0IIIYQQQggh1U92djbv77FA0blz5wrWs4oRc+fOxT///PNJ20cIIYQQQgj5xEEia2trsXXtGjZsCHd396poFyGEEEIIIRXHRn1JSUtwqX6jygipaj4+PjyDaMiQISW2UZ+PEEIIIYR88i6fhJfq7N9//4WxsbHYxcTEBC1btsSCBQsQHBws2SARG0GWlpbGb7P61Pm8vb2hqalZmUMSQgghhBBCCKkGSuvvMdTnI4QQQggh5NPp27cvtLS0oKGhwYNBq1evxnfffcfXKSsr85LRdnZ26N+/P9LT0yUXJBo8eDDWrVuHd+/eFXQawsLCsHDhQgwfPrwyhySEEEIIIeT/J9EsovcLIZ85li0kIyODkydPFgkS2dvbY82aNdTnI4QQQgghn4w0pCS+VGcpKSl8UBcLBP3222987tCVK1fC1tYWqqqqaNeuHV6+fAlFRUXcvHmzXMeUrUxDtm/fziNWdevWRV5eHjp06ABnZ2fUr18fGzdurMwhCSGEEEIIIYRUAywwdOLECYwYMYJ3NJOSktCsWTN4eHjwANHMmTM/dRMJIYQQQgj5ItnY2KB3795QUFAosl5eXp7HbN6+fcu3Dxo0CH5+fpILEtWqVQsODg64ePEif1AWKGIpTRMmTICSklJlDkkIIYQQQsj/7Z2UNF8kRZLHJqQ66datGzw9PXH27Fn+k41E3LRpEw8SFS9BRwghhBBCiFAkPW+QVDU/1VVRUcHTp0+RmZlZJFCUlZWFx48f88wiJjw8nGcVSSxIxLBOwtSpU/lCCCGEEEIIIeTzoqenhx9++OFTN4MQQgghhBDy3tixY/lUQG3btsWkSZNgZGSEyMhInDlzBomJiTyRh2UQscpv+/fvh0SDREx2djZ/4OJ0dXX/n8MSQgghhBBSOZKeN4gyicgXJi0tjS+FycnJ8YlyCSGEEEIIEZq0lGiR5PGrM3V1dbx58wZbtmzB+fPnecaQoaEhBgwYgF9//RVaWlrQ1NTkcxaxuYskFiRycXHBnDlzeGNYqbni3r17V5nDElJhMrJU8oV8kJNbsz57cmvYZ2V8ejZqErnq/q1eiJSSCmoSqbRk1CSKOjXrQqa8agRqiqzsmvN/RgipGDbJ7ZIlS+Dr61tiW6dOnWBlZUUvaQ3mtGE/VGX/rzGjgtFtehE1ifKIwahJHs/chpqk34mVqEmcV29HTdL08AnUFLJ6tVGTvJNVRE0io66JmqSGXQ7COxk51BQ1qa1fEl1dXR4kYos4rDx0RUpEV+qscMaMGTA1NcXGjRt5VIoQQgghhJBq4UsvUE1IFUhOTuZlKpYuXYqhQ4eWmHeW1UEnhBBCCCHkU6AuX9WrVJCI1bN79OgRlRgghBBCCCGEkM8Myx4yMTHB2rVrP3VTCCGEEEIIIcUcPXoUO3bsQEBAAJ8SqLAVK1bgzz//hMSDRA0aNEBYWBgFiQghhBBCSPXCai6Xs+5ypY9PyGeubt26fO5Z1uFk8w8RQgghhBBSXUhDii+SPH51Zm1tjQULFvD5h1q3bg3ZYiWEzc3NK3zMSgWJNmzYwEvOrVu3jgeMite3q1evXmUOSwghhBBCCCHkE2OT3c6aNQuTJk3C8uXLoa+vX2S7goICjIyMPln7CCGEEEII+VJZWVlh8uTJ+Ouvv6rsmJUKErFOgZubGwYMGCB2+7saNhk7IYQQQgj5PLyTkuaLJI9PyJdAQ0MDly9fxqVLl0ps69SpE++cEkIIIYQQIjgJT0OL6p1IBAMDgxLZQ/+vSh3t+++/x+jRo7Fo0SJoampWaYMIIYQQQgghhHw6UVFRvHzFxo0bMXToUCgpKRXZrqio+MnaRgghhBBCyJds0KBBWLNmDRwcHHi5uU8WJAoNDcWuXbugpqZWJY0ghBBCCCGkSrBMH0lm+1AmEfkCREREoFGjRvjll18+dVMIIYQQQggpQlpKtEiKdDXPJGLZ/jExMWjXrh3q168PVVXVItvnzJmDhQsXSj5IxDoMAQEBaNGiRWV2J4QQQgghhBBSTZmZmfGOZ0ZGBmUNEUIIIYQQUo00b94cP//8c6nbK5NdVKkg0YQJE/gkphs2bOAdCKliRQAbN25cmcMSQgghhBDy/6FMIkL+b9LS0ujXrx/Gjh2L5cuX87rnhbHyc3Xr1qVXmhBCCCGECI5FIr7gKYnQuXNnvlSlSgWJWH1qZuTIkWK3v3v37v9rFSGEEEIIIYSQT4LVNz979iy/ffv27RLbO3XqBCsrq0/QMkIIIYQQQkhVq1SQKDg4uMobQgghhBBCyP+NMomoRBj5v7Vt2/ajfT4FBQV6lQkhhBBCyCchLSXFF0kev7rLzc3FxYsX4eTkhLi4uCLbhg4dWmpyT5UGiYyNjSuzGyGEEEIIIUTCJk6ciLy8PMycORPDhg2DrGylTvnJF0xeXp76fIQQQgghhFRDrK/HSkP7+/tDWVmZr9PW1oa1tTWMjIzQu3fvCh+z3D3Gt2/f8p/t27cvuF0adh9CCCGEEEKE9k5KCu9YNpEEj1/dsTlkdu/ezecR1dDQwNSpU3nAqEmTJp+6aaQaS01Nhbu7O1RVVVGnTh1+uzTsPjQPLSGEEEII+WRzEkmwWyZVyf1sbGzg6OiI/v37o169epCUe/fuITIyEh4eHti+fTuvJLFu3TpeMnrAgAGVis2UO0jUoUOHgvmG8m+XhuYkIoQQQggh5NPo2rUrXxISEnDmzBkcOXIE27ZtQ5cuXXiwaPz48VBTU6M/DynC2dmZv0fYfEM7d+7kt0tDcxIRQgghhBAi8uzZMyxbtoyXgGOBGlYGTpJBIhYcGjx4MJSUlCAnJ4f4+Hi+vnXr1ryvx+YU/fHHHyUTJMp/sOK3CSGEEEIIqTZoTqICmpqa+O677/hiaWmJSZMmYc6cOVi8eDHPLlq1ahVq1679Kf9apBphAwFZP4+VJ2Qdzo/1+aiEISGEEEII+VRY3QjJ1Y5AhY/Nyr/t27cPLVq04OfRkpaWlgYVFRV+29DQEK9evSqxvaJkK9LJFHebEEIIIYQQUv2w7P6nT5/yTKLLly+jVq1a2LBhA0xNTbF37160a9eOj0Kjc3vCyMjIUJ+PEEIIIYSQCurTpw//ycq+CW3gwIF8UOCKFSt44Ojo0aO4e/duhY9TqVls2UjE0igoKKB+/frQ1dWtzKEJIYQQQgipPFacWqIFqqv/nETBwcE4duwY7yCEhYVh9OjRuHXrFu+8SL1vPytDwGpVs7rZrG41IYUlJyfzeurisPcQmyC3YcOGBSMYCSGEEEIIEQo7H83v10jq+ExSUhKKxz3Y8ql98803vLQdY2RkhAsXLmDTpk18jtEtW7agZ8+ewgSJevToUeYLOXz4cJw+fbrUeuc+Pj68Rvoff/yBqhIUFIQDBw7wiZqKO3jwIJo2bYpu3bpV6tgf2/9jj1tVynqM0l5PIdrGvMvLRZzTQ+RmpkKn9RDIKCgX2R7nbIGM6ADIqetBu+VAyMgrlrmPJOVlZyLO6R6kZeSg1WIgpGSK/itEvjiOd3l5UNCuDc2mffn2zNgQJHpZQk5VG5rN+0NKgpNiF5eZFIcwq9tQ1DKAUYeBJbZ739jPX0/1uk1g2KZPqeuEkhoVgnDbR1Cv0xD6zbsW2ZaTkQqf20chJSsLnUbtoNu4g9h1QvJzc4KD5WM079gdDVuLn9zN390ZHvbWGDJpdrn3kRR3m5fwc7ZDm96DUMu0YZFtAe5OsHlwE0oqqug46Cvo16mHnOwsvL5zBUmx0egwYARfJ6Q3j24jPNAP3YePhY6Bkdj7vLp7FSrqmmjVrU+595EEliL85No5pCYlov+YyVBWUxd7v7tnj6Bhy3Zo0KxVufeRhPSMTBy8fBdysjKYPXoI5OSKfpYduXoPngEh/PbYAT3QoXkjseuEEp2QhOP3LFFLRxOTBhT9bGC2nruN2MQUfnvuiD6opauFv45eKdjewtQYkwdU7nu8MgJiE3HdyRtNjXQxoHHJ/5v07Byct/VAUHwSRrdqiOa1dLHu3mvkvsvDnK6tUEtDFULySEyGZWQM2utqobV2yazvhKws3AqJQExGJibWrwMFGRkc8wng22SlpLCoiZmg7SWSs2DBAgQGBvI61KysnLa2ttj7TZ8+HQYGBvSnICW4urqW2eeTl5fHokWL+GS5pWH10FNSUnhQsqo8ePAAUVFRmDJlSoltP/30Ex9FqaOjU6ljf2z/jz1uVSnrMUp7PYVoW2HWsXFwTUxCH3091FctGijMysvD7bAIxGdlYaChAYyVRSVXnkZGwzs5Bd31ddFEXdg50W57BSEoIRmjmtaHoWrJPqdLVBwsfEMhIy2FRZ2a47pHAFyjPpRbXNylBZSLnWNJSm5uHo7de46k1DTMGtYH6ipK5dr+1N4dT+xdUVtXC3NH9oNQ7iXHIiQrEx2U1dFKqeR5z9G4cOThHYzlFNBPVZufb1xNjEZcbjbqvF8nI9Dgi1svbeHiG4hx/bqiQW3DMrfbe/rh3KOXMNTRxJIJI/ApBKWk4UFIBBprqqG7oV6J7S8jY+AUl4C2OlropK9Trn0kITs7G4ePn0R2dg5mfztVbKml+w8tYGllDfMGppg2aUK59pGUxJQ0HLrxCFpqKvh2WB9IS0uXuX3PpbsIiozB6N6d0KlZ0T64pIWER+DU5Rswr18PY4YVvR7E+qK/btgOWRlZdG3fBsMH9Obrvf0Dcf76HZjUroUpY0aUeI6Sxtp1wuIN4lPTMHNAF2gU+yy7/cYFlm5+aGdWB2O7t+HrwuOScPqpDRTlZDFzYBcoK8gL0lYbayu8ePYU/QYOQqvWorbkYxfhL184j+CgQAwdMRJNmjaDl6cHzpw8wQfMjPjqazRp1hxCunv/ARycnPHNmFFoYGpaYruHpxeuXL/BgwXrV/9ZsP7m7TvIycnFqK8+zedZTVWnTp0iv//555/466+/8KmxwVqFDRs2jC//j0p9SrCSFS1btuQnozExMYiNjcX9+/fRvHlzXn/vxYsXfATjr7/+Wuox2ES6b968QVXS0tLCiBEjSu3ohIeHV/rYH9ufRRWtrKwgSWU9Rmmv58dek6oU+fIskrytkRkXiqCb20psl1FUhayKJtLDvRB6f3e59pGksAe7kRkdgJRABx4QKk5WWROyyhpI9HiOKKtzyMvNRtD19XiXm4P0SF9EvzonaHsdDq7kgaKQlzcQ/OJqie3y6tqQU1FHoMVZhNs8KHWdEHKzs/B2zzLkZmXA6/p+xLgXfV9KScvwtskqKMPt/A7E+7uIXSeUuKgI7Fw+DznZ2djz+488MCHuBOfhxRN4de96ufeRFD8Xe5zb9ieyszKxa+kMZKSKLqjnk1dQhIaOHrKzs/DP4uk8QHTr8L9ws36B3NwcHFu7HFkCpr++uncNd04e4AGUbd9P46WPiosKDcLjy6fh5WBT7n0k5eqhf2Hz+B4iggOwa+X3Yu/jbmcNy9tXEOzjWe59JGXhht1w8PTFM1tn/LLzUIntVx69hJ6WBgx0tKD4/iRb3DqhTFi9B+GxCTjxwBJ7rz0qsf26pR0MtDX4IicrC3bJIP/30Jh4JKUK997Nzs3FlOO3kJqVjc0PrfHYM7DEfaafvIPHXoHQU1WCopwMX8du2wVHIjq54jWA/x+xmZn4+a0Tst/l4S8HN/gnpxbZnp2Xh5kvbeGVmAwdBXnISkshOTsbjnEJ/Hdtgd8LRLL+++8/ODs78yBRaQEi5vvvv+fn9IQU17lzZ8ybN4/3I1i2GeuLsKy0kydP8rKFrIzhuXPn+O9sKY2/vz+8vb2r9AVu0qQJnz+ptIoX6enplT72x/Zng+9YeUZJKusxSns9P/aaVDXH+AT87eGNrNw8LLFzQnqOaBRrvoM+/ngbF4/cvHdY6+LOg0avomNxxC+AJ2KucHBBYlY2hHLV3R/H7D0Rn5GF2deelXw+EbGYcfUpct69g7aSIl+nJi8HPRVF6Corwio4UrAAEbP+5DXcfmUP75AIzNy0v1zbLz97g9lbDkBeVhZaasIOUFGVlkFSXg48M4ued+TTkpGFhrQsnqYk4ExCBF+nLi3D11mmJuJcQqQg7bxnZY+1hy8gJS0DI37aiOycnDK3s/NkA21NnH/4Ep8C+x/7/pUd0nNzscvVG1ZRsUW228XEY5+bD6QhhRPeAXCIjS9zH0n59ffVePHyNewcnbBo2c8lth85cRqLf1kJZSUlaL2fvqKsfSRp2up/4RcWiWvP32DTiavl2q6joYbQqFg4epfsE0gS6wuPmL4A8YlJ2HHgGM5dv1NikL6hni401dXw5/ZduGPxjF+/GDl9AbKys3Hi0jUcPH0RQtty6RFuWDvDLzwGM3eeLrFdVUmBB+Yf2H/4zpuy9Rhik1IREZ+ExQcuC9JOXx9vfD9/LrKzsjBr6mREFLvue3j/f7hx7Qr/O/yydDHi4mJ5Fon++0FW0yaOR0JC6XM4VrWHj5/gz7UbkJaejhFjxiMrK6vo8/H3R9/Bw/iAEn29D0Hi6JgY7D98FBZPnwrW1s9FcHAwEhMTCxY2mOhzVamznW3btuHSpUv8ZLRw/bvz589j3LhxPKBy+PBhXtpiz549EArLWurUqZNgj1cTCPWaJPnZwmzyZkjLysPzyPc8mFI4O0fDnLWhE/9g9TmxtFz7SFJGTADMpu/ij+l3elmJ7TrtvuI/Veq0QKLHC1HW0Ls85KYl8OyclEB76HebJExbE2Igq6SCxmN/5LddT61HnR6jitynbu9x/KdG3aaIcbPi2Ubi1gkhMcANek07odFX86HXtDPC3z6CbpOOBdtZFpnpANFrJ6+miZRQP2jVby52nRAcXz3FkMmzeIaQXu06sHv+CMOmzi1yn6fXzqHv6Ek4vuWPcu8jsfY+f4gx369Ey+59ebabh+1rtO75oUwQyyzKzy6KDPRDUlwMZGRlkZqUgOT4WOTmZMPH0QZNO318dHBVefvkPuau3oFa9Rrg34hQhPl7o3ax7Kf7Z45gyNS58HNxKPc+kmL3wgJ/HLgABSUlrJg0lAfj5OQ/pBKzE+6Xd6+j18hvyr2PJDl4+MLuwl4+IqjzlB/F3ictIxP1ahmgSX2Tj66TtKj4JCjKy2H7wkn89rQN/+G7r/sXuY+0lBTvnLcwrYM6+qIL20vGDeY/p67/D9/0Fe473iEkCt0bGGPloC48i+iCnSf6NqpbsN07Kg5hCcl4vngSpKU/jICd1701AuOKpqQL4XV0HL6pZ4wpDeqijooynkfGoL7ah9Hdr6NjUVtZEWvaNCtYl5aTxjuW7EJCT4PPqEww+86WZLavgJnElWVsbMwHcrHzwMLlEOLi4vhIXSFH65KaiQWFrly5gtDQUMjJyfF17P3EMlXYBYkTJ07w/h4bRMeyW1jG2qca0UmEfU2eR8dgcSMzdNbV4QMT7OIT0E3vQ+YVu+iXlJ2NuKwsPkDBKSERgWlpmFrfBAMMDRCTmYWIjAxoyIveV5L2wCcE2wZ1Rl1NNSy4+QL+8Umor/Uh6/ussw+WdGmBSS3NC9b1Na2NvqiN5wGVH2Ra6fa+ccKjnSuhKC+PTvN+58EKNnDmY9v3XXuE/T/NRr92H77jhdJdRZMHKUKyxQ/k+VpDdJGypZIqnqck8Nv91ETneL1zNPFfbJgg7bz+3Aa7ls1Gx2bmiE1KwVt3X3Rp0ajM7U3qGePqU8kOCC6Nc3wCuhro4Ptm5vwnyw7q/D5biGEDftJychGXmQUlWRk8DotC7rt3H91HUl68eo3XTx7wue3adu1VYvu+g4dx6vB+tGvTutz7SEpGZhZiEpNwfeuvSE3PwODFa7Hy2zFlbp8woDuysnOQIWCQm3H39kOzhmbY/NtPcPPyxdq/92LCV0MLtrNz+aXzZvDbjczqw9ndC13atUaDeiZYs/wHWNk58owiod2zdcOdNd/xbKDuy3cgMzsHCoUC7r1amMNQSx3brz4uWMf6VDFJKZCVkcZrdz/e95Z0BtTDe/ewZPkvGD9pMjQ0tfD40UNMmjqtYLusnCySEhMRExPN2/LssQVGjf0G330v6nvHREchODAImppaEALLBtq5bTO6du7EAxbWNm/Ro9uHCh3HT53Bgrmz8fuKX4rs98+efVj64/e4ck008PlzwLrghbrhEjk+o66uzpfqgM0pu2PHDixcuJD38djt0rD7LFmypELHr9QVeT8/P+gVikjm09fX59vy057i48uOprLOhYWFBa+RPmfOnFJLqbHSdewxWZbS2LFj0apVK2zcuJGXOVizZg3fVry02tWrV3kwi2U45dfpq4jK7s8yrFjpN/ZlN3v2bHTp0qXc++aXjWOv5fPnz9GvXz/MmjWr3PsXfz3LKjeXmZnJl3zFay2W27t3PNjDsHJsrIQcy8QpLMr6MlJD3KHb4ety7yOpUnMss4nhQSlp0ejv4nIz0xDncAcGvWbyTJe649Yixd8WivoNEHJ7K4SSnZoIRQ3RxTtFTV1kp6eUvE96CvzuHEVqVBAajVpU6johZKUmQiG/vVp6/PfiUqNDEPj0EtJjwtB8yopS1wkhJTEeerVEHWtW1iwqpOjIoJSkBMSEh/IgUXn3kXR7NfX0+W0tfSOkJpb8nHV++QT2z+5DTVsH2ga1MPTbRXB4/hAZqck804gdQ7D2JsRDS080ykZb3xDJxUbZuFi/QNMOXaCgqFTufSQpLzePB3sY1oaUxISCtjBPrp7lJeW8ne3KvY+ksNIMau/T9mVlZXjJueJmjh4Ev+BwXHlkCTe/IGxaPEvsOiHEJaXASEc0clBfSx3JaSUvJswc2hPRicnYeekeYpOSMWuYqFxCRJzoc0RHXbjRsfFpmTB4H2Qx0lBFfHrR9oYnpaJTvVo48MoRstLSmNiuCVQUhLngJQ4bla2rKAoGGCgqwK9YJlFURiaaa2rgkLc/dBUUMNzYEJrycuhjqMcvMPxi64xdHdvAREwZHlLzuLu784v5L18WHf3Mss537drFL+oT8jEsY4V1iPMDRB/r812//vELDiyotHr1anh6evJ+DevfiMPelxEREbz/wrJlWOk31jfZv38/f5xVq1bx9hQurcYuILH3NHtvV6bERmX3ZwPf8vtd7PVYunQp6tb9MJCgLOy5smoc7HkGBATg22+/LfV1Kc/r+bFyc1XR37sQFILw9AwMq2WIxOwc/j3C6CsoIjG76AXT2ab18Cwqhg9AkJOW5t9PfQ30sN7VA3ZxCfz+5hLOdtn52hlp2TmY1bYREjIyYaAqOl8yVFNGfHoW6he6lheRkoY+9Wvh79dO6GSsj651PpQhu+Tmh1W92kq0rRlZWfjriGjE/JpZ48Dy51kAiLdXWxMJKWnQ0/xwcUrc9sj350lrj1/FiK5t0dq8/O/FiorJycK1xBgoS0tjklbJkm3ipOXl4lZSDOZo1ypYdzQuDAFZGRirIerXSFpcUjKMdEV/eGM9bR4Iqsh2oVhHxeJVZAyaamlATloKevnndkqKSCgWnGiprYk/2zaDZ2IS2shp4nVULP9/+9g+ksJKXrPrXwz77mCl5Ap/f6SmpSE2Ng5rN23F2K9HoknjRmXuIynxyakw1BH9rVWUFJGbl1eh7UKLi0+AkYHo/8TYyACx8aJga2ERUTHYuu8wfPwD8c/a36ClqYH6JsaY+/MfCA4Nx8aVJQdGVzUWQPvztCgY9eekITybNL9cnJGWBuKS02Ck/fEL7ed+nYlb1s4w1tWCnU8wktIzoVmsTF1Vi4+PRZNmogC7Ue1aCAkOLrL921lzYGxcB1FRkdDXN+ADrphXli9w7colpKWlo2lz4crNscc3MhR99tauVQtxcUWvlYSFhaNXj+5YvX4j2rdtg2FDBsPewRHmDRpAR0uYQBaRbKY/Oz9u164d//xkt0vD7lNRlQrJNmvWjJ+kZxQqWcRur1y5km9j2Ik2a/zHsIAPG33GyhiwgI+d3YeLbsVLqW3dupX/HDJkCK99vWzZMj7JLvsi+e2330qUZGOd4h9++AH9+/eHoqIizp49W6HnWNn9X79+jRkzZqB79+5o27YtvvrqK3h5eZX7cdlzZBNMsQ4Sm9dp586dPGhUHuJez7LK1LH7aWhoFCyVHYUmJS2N3CxRaYbs5OiCIExhskrqUNQxRqLXq3LvIwnScgrITU/mnbu8nCyeIVRcdnIswh/tg0GPaZBVEtXMllfXh3arIcjLTOHl6IQir6aFtFjRKLb02HDIq5QMpLHXUl5DBwrqOoh2fVXqOiEoqGnxdjJpMWFQUC9Z5obNBaWoocf/FnHe9qWuE4K6li4vd8ZEhgRBQ7voaPrbJ/YjPjoCp3asRUx4GF7fv1HmPpJub0yYaD6Z6NAgqGuXDNgrKqtAx8gY0SGBiA0P4ZlE7foOQZehY+Bq9Ry67wNcgrRXWwfRocEF7WWl8Aq7fngXPO3f4PGVM3B98xKBnq5l7iNJsrKySEsWXTxhwUG1QiOCWHnBh5dO4cWdK7B+dAevH9xAbGTYR/eRJNaxYvXo2WdZZlY2souVe2FG9+uOn74dhzObV8DayaPUdULQ01JHYEQMvx0SHQetYnMYMN8O6YnlE4Zhz+LpsHL1LVh/4t4LjO8jbKawrqoSn2uICYpLgl6xDgrb7hQWjaSMLLiEReOvO5+mFEk+bXl5hL0PvIWmpZcoH8e228XFI+/dOzyLjMZpv2CoyclhaoO6mNfIFGPrGsM9UfgMKEl4JyUt8aW6Y+Wf2fkoO4ctbPDgwfyCNLu4TMjH1K9fn/crWICmcNnX6OhobNiwoUJ9PpZ1ZGhoiN69e2Py5MlF+pDFA1Ms+MEy4VjHduTIkTh06BCvTsHK2+WXtStcku3vv//m/TR2X1tbW15msSIqu/+///7Lyzqyvh67uMkG6BUOxJSFPVfWd2al+9jcTyy4U97/S3Gv58fK1FVFf09DTo6XJpWXloaWvBzC3g+cCE1P598vhbGBE/0M9THYyIDPXVRLWYkHhzTl5PkxWDmsgFTJlmRlZeJYuTjWFh1lRQQnigZOBCemQFel6Oci2/4sIJxnw6x5agef9wGXmNR0ZObkQl/CFyilIFVQWpeV45ORlkbK+9eXnS9pFwuoidteW08bd17b84vZk9fuQXJa5UsuloXNH6QlKwuNclYAicnJxu6YEMzQNoJaoX00ZeRQW04BL9NKDiiUBD1NDQSER/Hb/mFRMNDSqNB2obCMIB1FBajKyfJzOXZOx4SmpkNHzETprXQ08Y2pCVziE/n/Wnn2kQRW1YBdm2OBd3YNqniwR1dbG4+ePENObi7GTZ2BnJycMveRFFY2LiRS1CdJSE4tkqlXnu1C09PVRkCwqP/vFxQCg0KZm4X7hUb6ejw49PD5K0RGx8A/KAS1DfT5nESPnkv+ehDLaDLQVOMLu80GMCamit6LwTHx0FEreyCajpoKpvfvjAZGugiLTZR4gIjR1dVDUKBojtaggADovR+Qm49lDw0cMhRTps/A8yePYVJXNE+tsooKDx6x8nNeHu4Qip6uHgICRYOU/QMCoK9f9FqJvp4uHj5+zAMI6zdvxYuXr7Bu0xa4ubvj373/4Y2NLV5ZWeNzwL4zJb1UtCwdO289duwY/50NJGK/szhBVWFxhvnz5/MSwyx5hiWmsN/FLZUpQyxb2Y4nC2BcvHiRZ9mwjgMrMcculOWPTGRBFlaW7mPYyT8LNjGOjo5wc3PjT1gcVjKNnUgzN2/exMyZM3nAiD1pNulucWyOJBZRYx3k/ONXRGX3v3v3Lk/nYn+o/A4Ae2MUn1DqY9q0aVMwCRbrdFy4cKFc9eLFvZ6tW39I5xWH1VJkI9/ysS/nSnUcGnWH3/lVPBCkpN+AZ94keLyEsqEZ5DUNEGV1CTlpCchJT0ZmXFip+whF2bgpgq6uQV5WOtTMRB3bOKd70GjcEzLyyvA/uxwqJi0RY3MVCjp1oNViAOKdHiA9yg/J3q9g1H+BYG1lQRZpaRk4Hf0TKREBqN9/Ml8fYHEW9fpNRHZaEvzuneCZWZmJMZBVVBG7Tiia9ZvB+fRmOJ3cgHgfB7Sc9jsvixbw5ALq95uA1KhgBL+4jnfvcpGREI2s5ASx64TStmd/rJw0BJHBAXB69Qx/HrmCpLhYOLx8gp4jxqJRm44I8RV12tlnnJKqGpp36lFiH6G06z8Me5fPgcurJ7zU3FfzliHc3wcxYcFo0a0PnCwt4GlrxUuesQBLVmYGIgJ98eLaOQS4O0JdWxemzYtOxihJ3YaMwu4VC1G/SQskxETBqK4pPOysIScvjwbN26DTgOHITE9DemoqUhMTIKegKHYfoXQZNALrv5sMTR19GJuaQ1ZOns83VMe8MXQNa6PbYFEpymg1NUjLyEJGVk7sPkLp1a4lRiz6g49CHd5LFEQ5cvUexvTvAQ01Ffy++zgyMjNh5eSBgV1FI0nErRMCywJSVpTHlHX74B0cge/HiEpg7rx4D4vHDeYXNNafvME7jxa2rvh50nC+nZ1j3LZyxNLxQyCk1rX18VPoEyy+/Bg2geHYMboPX/+fpQPmd2+NJgY60FVRQmxKGuLSMmCqKxo8sOuZHWyCIpCcmYUxrRuit7kwJf266etg8os3CEtLh1VMHPZ3actHj76IjMaIOrXQRU8Hp/yCeEmSlOwcKMvKwC0hCQ/CInkm0cuoGBwS8P3wpWb3sPNldnG3a9euHy2fwUp8PXnypMT6UaNGQUWl7O90dnG/tHNqVVVVhISEoFGjD2V2CCmO9UFYSQvWyVy/fj3Mzc35BMisj9G4cWM+aS8LirD32tq1az/6Ak6aNInPb8SwwW/s/WdmZib2vqxsXX51CTZYjmX5sDlWWfaMvb292D7bP//8w/uK48ePr3CWXGX3Z/dj7WMDAxn2/8r6XqwfV14TJ07EggWiPgUL8rA+I6uSURZxr6ek+3uDjD5kSA8w0Mcvji54ER3Dgz+LGjaAX0oqIjMy0EVXB/4pqbgRGg7XxCSe2dpEXY3PSRSdmckvfAenpZfIPqpqU1p9KB03slE9fH/bEk31tZCYkQUTDVVYh0RBSU4GLQ10MKpxPfz31p2X7srJy4PC+8yGy27+GNlYdDFQkhTk5bDkmw+lo0b37IDhP2+FjoYqWpmZQEZGGlef26BNw3qoZ6gndvuYXh15GToWaGKle2XfPwdJ0JKRw5hC2T+WqQl4nBKP5NxcKEnLYIS6Lu4mxaKXqiaUpWWwLMwbrZRUcTkhGibyChiopoMjcWHIfQfE5eYgU8yATUkY378r5m8+gK4tGsLW0xftGpvC1sMXqemZ6NmmqdjtLGi078p9BEfG4pc9J/HrtFESn/OJZQexhWGl49bau+EvWxfYxcZjbbsWfLDPGd8gTDETZYv97eyJ2MwsnoF0pk9n/I+9u4CLKvviAP6ju7sRscBO7O7u7o4111zX7nWN9W+3q2u7unZ3t2IrFooo3c3/c+4IMjAoIPMAPd/P532YeW/emzsDM7z7zj3nmmprpdlHCg3r1UXj1u2hqaGJGtVk34tbd+5GzWpVYW1liWZNGuLO3fvifwv9fdI5kKJ9pKCpoY6CjrZoO2E+fPwC0aGe7LmX7DyEgS3rp7t967Hz2Hnqsuir0GN6NJb1DZStUP58eO/7Cd2HjcOt+w8w53NW0MJVGzCiXw+8//ARi9ZsREJCorhNAaLwiEg8efESzg524n6CBJ8zCgoNb/HlPWlduSSaT1sJKxNDFHawEu/Z/qv34e5kAxdrc1x86IW/T10TGUMzth3BxA4NcPCaJ87cf4bDNx5iRMtakELDJs3QolF93L19C6dPnsTxcxdENtGtm9fRrEUrvH/3DquWL8WTRw/FeVDN2nVw4dxZnDh2VMxj9ObVS7FeKu3atEKfgYNRuWJFXLtxA0sXLRCZQsEhIahRrSratW6FUeMmiLm/aN4iXV0dNG/aBP4B/qJfraOjDT1drhyhDIGBgclJGpTpTUFwuq+rq5upCmMZNXfuXERERKRbOSwrshQkosAMlRigyUqpw0tRYgrUdOjQIbnj+vvvv3/zOPRGJTdEXV2MJkhPytrpFBFNuk+3Fe1nbGyMFy9ko5Dpg5BUEiGjsro/dWJSZg5RCYHMpnjRiT51hKiTQM9Lx8yIzLyfSaiGYcpa9VllWaEVtExtkRAdAaPCsrlO1LR1ofL5BFWUkVNRhbZFPtjW6pPuPlKxqdUfIU8uQEVdAwausg+rmpY+VFRk7TUt1TT5sUkZTvRTy9QeJq0mQ8cqv6TtLdFnJnzvnIFdxSYwKyyLBmt8ziiiNlMgiUrnmRQoBXN3DyTERKdZJxUK9lUYvgQf71+AY9UWMHZ2E5+hpAwoVXUNaFLgS10DFkUrwbxwOUQGfEizTiqU9TFx5TZ4Xr+IRl36wdzGTpRj0zOUtbdk5ZpiIVRCLOl26n2kYu9aGAPnrsCrh/fQtM9waGprQ11TU2QPES1dPZF5o6Wri5ptusHG2VUEkIzMLVGjdVeUrC7N3FRJSlWrIwJrvm9eotuYqbI26uiK7yhSp2235AycAN/3Yh4iWlLvI5VGnfvAysFZzOHkUVcWpNDW0xeBH3UNDTTtPkCse/Xkofj/Y2xmoXAfqcwf1Q+7T14QJ+Uta1UW64wN9cUFA2JlZoz4+AT83r9MckBI0TqpbPl9EP49dwM9GlRFnbJF5UrI0chYurihqa6GZSN7oqK7a3LJh7GdGiv1gociVBN7R69mOProFbqVd0dpB9kFMhNd2QhkOvdZ3qEuDj94iaramqhfRHYhyUxPG60+X5zSk2i+BWKoqYHlHqVxzS8AnVwcYKerg9DYWJEtlDQydXbporj40R9VrcxRycJMXNSjUd0Oemro4eoEG90fZJ4ayvRRZv3yLGQSDR8+XJSmoowBylSws7MTF6dpVL8idDGczq3pInLqTKCMBIlokNDGjRvRrVs3uWAUZZnfu3cvOQuEsa+hv0Gad5YGrFGfhOYkGjt2rMieSfq7oiBObujzUZCHgh+U6ZQZWd2f+me0HwWJIiMjRf+NjpUZKfuXdJs+lxm50JTZPl929feSFDQ0wJwSRfEoJBR98+cT2UVUFkv38/9pum+upYl2jvaobinLtu+ezxH59fXwMToa9awtUdJEuqoM9VztYaStiTfBoZhSQxYcowCR9udz0SpONuL/9cOPgehSoiAcjGTnJXaGeqjjIt05fpLh7RrC1d4KIeGRIvhDDPV0ofH5/VW0vW/TmnC2Mccrn08Y06kpdFJlEyuTnqoa3D8PSDT8POhTn4IAkA3Bbva5DLlYr6qenEVEWwtp6aK8rjTzPFQt6Yb1EwfD0+sNZg7sLL7DUr5PirZTBomVqTGGtm+cfK4qJcraWlWlLM59+IQ2+RxQ1NRI9KupXHASyjpy0tfDMPcCsNCRnaOm3kcKkyeMwd79B0U57FbNZf0hI0MDkeFCfh32Cw4eOQafD774fdxo8f4q2kcqq8cPxO7TV2BsoIfGlWX9ITNDWQZMetvpc1izjKz/YpJi3k8pHNi4HHuPnkTfTm1RpYKsPaafv0fpPba2MBf/D2pX8UD9GlXE+7tr9WJcvH4bVcqXQfP60gRcUhrcpBpcrM0QGBaJVpVKiHX6Olqir0eoFF1heyuxJJWlo+2OlqZYNrg9qhVVPJgku9k7OGDr7n9x9fIlDBk+EqamZoiNiYXh5+tB9P5aWlmhWPESaNC4iXifKYuI1tE5SofOXVGshOz1SaFihfLYsHol7j94gOmTJyafK1HwkpQsURwL583BlWvX0K1zR5QpVUos5JOfH+55PkCJ4tIEj5WNvpGV+a2smsnHU3IHZQ5JxcrK6quVw7JCJTFl/QAJ3bhxQ2TLHDhwQNynDBgaUUb1mL/1WApG0ag2SrGnkhl0n94YT09P0Qk/ceKEqPFcvnx5MVcRjXKj1NUxY8aI+YwyIjP7p3xeihxShJBGSFDqP5X5oLmFUpf7+Nr7Qh16+pDTSE8KONH+1Iak58jM+1m2bNmv7pcadY7ogkXRETuhppX7o8txsZmfayon2eVLW3otN6OJZ/OSbhWVV39bGUJjvh3IzU10NaS9SP+96IJFXtEy4R7yknj/D8hLAq/dQF7y/nLeKQVGWUk1jp4VE6fm9ISeSecwn14+hqGhgRKfJxQW+Qpn+DUfOnRIlAK+evWqOC+j0sI0gIiy8imDQRE6n6NzzvTKcmXkPJYCRWZmZmjbti1MTU3FwK5169aJLA0qscWYFP73v/+Jv/mkagdUypvWUTbStx7r7OwsSqhRP4oGJ1J/j7J3qANOfUAaOZk0lxBVt6DRlDQvD5XgppJ1GZGZ/VM+L/W9aL8iRYqIAA8FcGke2My8LzTnLg0MJFRuicqRbN68Ofk5MvN+Xrhw4av7KfquPFmrKvRzuIxSRpm7ZWzum9zCrmkD5CWn+qQ/8XVuVHuTrMJMXvFgwVrkJW5rNyGvUHlwFnmJqlPeGqQTc+MY8pLQWv2Qlxgja+f5OYH6P5aO+XJVn+/5m/cwUGJbQkNC4Opomytec3r9PYo/rF+/HpUrywYMf68snxXSySml2dNJcVLWEI16pJp4XyufkYRKFlCZgpSlBdIbGZn6sTSCzdHRMTly9ueff4rbNFkolUMg1CmmkZI0YpKCJRS8oQ5yRmVm/5TPS6PKqAwD7UfvA10MSJqML6NcXFxErWkKPhUrVkwcMzQ0NPk5MvN+0gfna/sxxhhjjP1QKNNHmfMGZfLYdCGYsg3onJBQtgGVM6aL3ekFiQiN46IyxpQlQBkGdH6YmfNYuuA8adIkceGaOlM0x8ycOXMwePDgTLWf/dySSmfT6GoqnUGBS8qgyWg2GgVDU2a6UL8kvVJnqR9LQc2k8mvVqlVLLr9dv359EdAhNKiPyrxREJQCo5SpZ26e8bkiM7N/yuelzzPNIXTnzh1YWlrCzc0NmUVBJhocSKXmqFIHZfukfI7MvJ/f2o8xxhhj7EdC56ZJ2X/KOn5u9u+//4rYDPUzqUpF6vNXGhiY2X5floJEFLyoW7euyHahjkNSkIhGMVWvXl2c7H4LBS9STqJEddGfP38uTnxTogDLvn375B6bstYztSEpYkYlEKhUQBIKklBZD2JjY6Pw+FQeIGVZg5TPmZH9kx6b8nnpBD9lvUHaj7J5MvK8Se8lBaSoM5Qk6bUpOpai9yhlnfmUbWOMMcYYY9+PAi8ZKelEF52TzieT0CAgGv1F87nQOaYiNNiI5vekn5TZQJnzlMmQVK7zWygTgi6yM5ZVu3btEiXnKAhCA/MoSERBmy5duoi5cCiL5lsoGygl6pfQQEOa4zYlCjpRbfWUatX6Uh7H1tZWLCR1kIkqP9BCqF+o6PiUjZQ6My/pOTOyv6L2UcCXKlskyczzJr0GGviYNPgx5WvLaBuS+nkZKUPJGGOMMcZ+DEWLFhUVz9JDg58yK0tBIpr4ksq9UeZKysjakCFDxMjIjASJFKGTczpuShnJSvqe4ysK1qT3nFltX2aeN3VGUHa1gTHGGGPspyBRJlHqC9V0/kalf1OjEgWp55dMyk6nbYqCRHQ+SFkKSRePaYCWh4eHuChPGfWMKRtNtkt9DgoU0d/rsGHDkvsd/fr1E1lwK1asyNKxKdiRuj+TmYoPWTk+le5OHcRN7zmz2r7MPC8FvL42l5Cy3yPGGGOMsbyMZhdQ5gwDqrk7kUick9PgppRJKinLKcd/nqdK6UEiqqlOHYbUChYsKFL1s4qyi1Jn6mSn7z1+VvfP7H4pM4Kyqw2MMcYYYyz7vH37Vq4+dXoTw9OgoLCwMLl1VEY4aZsiKTPCk0aK0dxCNFdRRoNE9BxU6u7Zs2cIDw+X20alByibibH0UPUCyiBq2LCh6Pul7vPRHEFZRdlFqTOMstP3Hj+r+2fn61L2e8QYY4wxxvKuY8eOiWx1RUEi2kYDlRRty/YgEUWrqOYxdYxTZhJRJ5TS7hljjDHGGPuRM4noPDgjk5jSXEKvX7+WW0f3KahE9aMzispJ+fn5Zeixnz59EiUGqDb1x48fxcVmChRR6Tua14VqVDOWkf6eoprs3OdjjDHGGGM5LZcn++SYV69eiUGGmZWlHnSDBg0wY8YMMaFuUqfh/fv3YlQiZ7kwxhhjjDH2ZXL6EydOICAgQNyn8+ft27ejXr16yfMLeXt7Y/PmzckZP+/evUtTpvjw4cMZnmdyyZIlYh6ku3fvokyZMvjtt99w7949MY8MlbijQBFjX0PZQjTv6d9//y0XJLp9+zamTZvGfT7GGGOMMcYkRuWeCxcunFz6mW6nXGhe2k2bNom+ZmZlKZPozz//FBOJOjk5ISEhQZRHo5GJ+fLlw+zZs7NySMYYY4wxxr5boooKEpWYSUTHz4w+ffpg3bp1qFu3Lrp3744LFy6IC+2XLl1KfsyNGzfQtWtXvHz5UmQM0fxGQUFBItBD85Zs2LBBnHNPnTo1Q8/5+PFjtG7dWtzW0NAQQSbSsWNHcR5/69atDAec2M+JAkPUwWzatCn09fUREhIi5smhvy0aFNirV6+cbiJjjDHGGPtJ/axzEpUtWxZDhgzB0aNHxRyiKZN16PydKl1UqVJFxGgkCRLRRJt37tzBzp07RaeWOq2DBg1Chw4d0q2tzhhjjDHG2M+GysqdO3cOq1atEpk9BQoUwNy5c+VO3B0cHNC5c2dxMZ6sWbMGhw4dEhlINOno0KFDRRBJW1s7Q89JZcIo2ESsra3F/DKptzP2LTQZ7pMnT7B161bxk/7+5syZIzqjqUvQMcYYY4wxxpQfJKKFkneon5id88xmKUjUokUL7N27V3RWaWGMMcYYY+xnmpMoMyhgM2LEiHS3U0k4KjeXUqNGjcTyveiCPmUwUUYR1ad++vQpSpQo8d3HZT82+jtZu3atCGhSkJIxxhhjjLHcggYsKXPQkkouHxDl5uaW7cfMUg+aRjXyCETGGGOMMcZyn4kTJ4py0ElzItGcRNu2bROVAPbs2QNTU9OcbiLL5WJiYsQ8WIwxxhhjjLEfX5aCRLVr1xaZRIwxxhhjjOUqNOpL2UsuR5lJFhYWyfcpi4lKRNN8SHXq1MnRtrG8gSa+DQsLE3MQMcYYY4wxlhvnJFLm8rPJUrk5JycnUbbiv//+E+lNmpqactvHjRuXXe1jjDHGGGOMZULr1q3Rr18/uYlMGcuMjx8/onjx4vDw8ED79u1F/09V9cv4Qjs7Oy47zhhjjDHG2M8cJLp06ZKoZU6T4KaeCJdwkIgxxhhjjOWIXDgnkdSsrKzg6+ub081geVhAQAC8vb3h6uqKmzdviiUlmiSX56ZljDHGGGM5gRJ9lJnso4LcacGCBTh58qQYDEiLg4NDzgaJqFwFY4wxxhhjLPfp06cP+vbtK0pEOzs753RzWB5UtGhR7vMxxhhjjDGWi7Rr1w6JiYnYsWMHhg0bhiJFiiQHjCpUqCCX+S9JkIgxxhhjjLHcKFFFVSzKPH5ut2XLFrx48UJkgVCZMAMDA7ntS5YsQdWqVXOsfYwxxhhjjDGWVaoqKmJRFtVcOg+tvb09Ro0aJZbg4GAcOXIEBw4cQNOmTUWAqGHDhiJgVL9+fRgaGmbq2BwkYowxxhhj7AdSs2ZNODo6prud5pNhjDHGGGOMMZY3GRkZiblDaUlISMDly5dFwGjatGno3LmzGBQ4YcIEUV0iIzhIxBhjjDHGfhw8JxFatmyZ078FxhhjjDHGGFMKSvRRZrKPSu5MJEoXZRFVrlxZLLNnz8br169FwCgwMDDDx+AgEWOMMcYYY4wxxhhjjDHGWB5HJccHDx6cqX04SMQUCvTxh6pmRK5/d2IiQpCX6BtrIy9RyWOh8/jEROQlEbHxyEs87I1zugk/rnexyFPi8lZ7Ax6/Rl7ie/8T8oqIhNz3PZaooiIWZR4/t+vRowdOnDiR7vaNGzdmuOwAY+zH5PnbMujoy89Xllst2e2JvOSmWyHkJYYnGyMvCbfQQV5yekQZ5CXFA98gr4gvWB55SlQo8hK/an2Rl1hq5v55Q1NS9807nzX10DDkNiqJiWJR5vF/NlkKEv3666/pbtPS0oKLiwtatGgBMzOz72kbY4wxxhhjLJO6du2KGjVqyK3z8/PD2rVrUaBAARQqlLcuYDLpvXnzBn/99Ve6g4h0dXVRrFgxMTGutnbeGgTFGGOMMcYYy4Yg0Z07d3Dy5EmYmprCzc1NdBQePHiAgIAAUftu/fr1GD16NM6fPw93d/esPAVjjDHGGGOZRoO+lDnwKy8MKksvS6hXr14oVaoUDA0NJW8Ty1siIyNx+vRp3Lp1Cw4ODiK4GBYWhnv37ont5cqVw+LFi2FnZ4erV69CX18/p5vMGGOMMcZ+FokJskWZx//JZCkXjwJD/fv3x7t370Qg6Ny5c+J23759Ubp0aTHyrE2bNhg5cmT2t5gxxhhjjDGWaTTAizL+79+/z+8e+yrKNlNXV8fy5cvFxLc0QJCCQdTPo0DjgAED8OrVKxgZGeGPP/7gd5MxxhhjjDElCw0NRUjI16deefv2rThnlyRItGfPHsyYMUOutADdnjlzJv7991/RoZg8eTJu3LiRlcMzxhhjjDGWJQmJiUpf8ipfX194enpCT08vp5vCcjlvb2/4+/uLYFDKOSotLCwwYcIE0eczNjbGqFGjuM/HGGOMMcYkpZKYoPQlN6JM/nnz5iXfp/LQqQds0fyzq1atkqbcXGBgID59+gRzc3O59bSOSs4lzU2ko5O3JhRkjDHGGGMsr5s4cSKuXLkity4qKgp3794VZcKKFy+eY21jeQP194KDgxEbGwsNDQ25bR8/fuQ+H2OMMcYYYzmMsoqon5cdshQkatq0KTp06ID58+ejbNmySExMxM2bN0V5OdpG9u7dm3ybMcYYY4wxKVCejzJzffJCHhHNHxMXFye3jgZvDR06FK1atYKqapaKCbCfCJUXp7mrqIT4pEmTULBgQYSHh+PEiRNi7tnff/9dPI77fIwxxhhjTHI8J1G2y1KQaOXKlRg8eDAaNGiAhARZ+hV1Njt27IilS5eK+5aWlnLpT4wxxhhjjDHl6969O7/N7Luoqanh4MGDYh5aGhSYhEoV0sBACjhGR0ejYcOGaN26Nb/bjDHGGGOM/WxBIpqgdPPmzVi4cCGePHki6lTT6DKqUZ2kWbNm2dlOxhhjjDHGvikhUbYoizKPnZ38/PxgYGAgSkAnobLQlFHEJaFZRhQuXBhnz57F69ev4eXlJf6eChUqJH4S+tviABFjjDHGGJMczROrzLliE3Nvpy8+Pj65xBxVj0h5P2mdZEGiJBQUShkYYowxxhhjjOWsR48eoUuXLrh48aLc+mvXrmHJkiUiQ4SxjHJychILY4wxxhhjLGfNmTNHLKnXpfTbb79JEySiOYj27NkjOp40IjG1DRs2ZOWwjDHGGGOMse+0fPly9OzZE9ra2nLrqVT0qFGjRCUAyghh7GtoRCL16+7fvy/mI0opf/78yfMSMcYYY4wxxpSP5pd1dXX95uPc3d2lCRKNHTtWjEKsVasWTExMsnIIxhhjjDHGsh0NZqJFWZR57Ozi4+OD0qVLK9ymr68Pb29vDhKxb/6dU1Dx4cOHqFKlCnR1deW2U1kLxhhjjDHGckRigmxR5vFzITc3N7EoQ5aCRDSi7Pjx46LDwBhjjDHGGMs9SpYsiY0bN6Jbt25QVVVNXn/r1i3cu3cvSyPL2M/l5cuXuH37tpiLyMzMLKebwxhjjDHGGEvH+/fvRcU3Ozs7VKpUCVmR5XJzpUqVytITMsYYY4wxpiwJibJFmcfP7QYMGIAVK1aIbKK2bdvC1NRUzFO0bt06sc3a2jqnm8hyOervOTo6coCIMcYYY4zlOiqJiVBRYraPSi6uHjF+/HiR8V+9enVx//r166hduzZCQ0PF/V69emHt2rWZPu6XoYWZUKNGDRw+fDgruzLGGGOMMcaUiDI/Ll++LIJEq1atwoQJE3DhwgUxoemCBQv4vWfflC9fPjEnEWUSMcYYY4wxxnLemzdvsGfPHrnqbuPGjUPx4sVFJQCq/LZjxw7cvXtXmkwiW1tbdOnSRQSKaLIkFRUVue3UOMYYY4wxxnJC7h33JR17e3uROcRYVnz48AFFixaFh4cHOnfuDCsrK7ntVMqia9eu/OYyxhhjjDHp/aRzEp05c0YEiNTU1MT98PBwnD9/HgcPHoSzs7NYWrdujStXrqBEiRLKDxJRjTvqNFBUSlFkioNEjDHGGGOM5Rw/Pz8YGBhAS0sreV1AQAB0dHTEwtjX0N/K27dvRck56nimVqxYMQ4SMcYYY4wxJiF/f3/Rx0s552x8fLzcPEQ0WJD6gpmVpSDRjRs3srIbY4wxxhhjSsVzEkHMP0RZ/zSwK6Vr165hyZIlYqQZY19DAwK5z8cYY4wxxnKlnzSTyMXFBevXrxeBIcom2r9/P8qUKQM9Pb3kx7x69QqNGjWSZk6i7PDgwQMMHDgwW4/57NkzMTmTIrNnz8axY8eyfOyv7Z/ea/laezLrW8fKzufKioTYaATd3I6AKxsQF+6fZnvgtc3wO71Y/IwLDxDrgm5sFeto8b+wUtL2xkeFIOT2DoTe3YOE2Mg024OurEPQpdUIvf9f8vbEhHiEPzmBoMtrEfH8nKTtjfz4Bl57F8D71N9IiI9Ls/35jll4unUa3p/bhsT4eLHu5X9/4emWyQh98xBSC3nliee758Hn8r9ptsVFhuHJlsl4um06fK8dSF5P62iJCvCRuLXAjdNHsGLSCFw6sjfdx9y7dAb/LJyeqX2U5fJ/27Bl5mg8uX4hzbanNy9hzfj+2Dx9FF7cvZ48+fTprWvw97SRuHvmiKRtpX9ca/+3EJNGDsYjz/RrotJjDv67U9ymn8N6dcLkUUPw4uljCVsLREZGYNGsKZg2Zjjee79J93FzJo3F1Qtnv7lO2T4GBGHYHyvx64I1CAoNS7N9xpptaP3rLLGcvHYn3XVSeer9AYMW/43pf/+HmNi032VDl2xB22nLxPLg1TuER0Un36dl0e7jkrb39scATLx0F38/eik+R6n5RUZj3o2HGHrmBm59lP1vI8+DQjH6/G1I7UpEMBb4vcXpsECF271jo7DU/x2mf3wFn9hose5UWCD++PQGB0IzP9KI5V7Lly9Hz549oa2tLbeeJjilTsOTJ09yrG3s57JlyxasWLEiW4+5a9cuLF68WOG2Dh06wNfXN8vH/tr+6b2Wr7Uns751rOx8rqzYuXgmlo7ui4fX0p6DJiQkiG1Jy8H1//vmPsoUE/QBr/fMwJt9c+B3fa/C/+NxkSHwObMOr/+dgbBXsnOiN/v/EPtFfXoFqR07fR69ho7B+n9k58SKXLx6A0PGThK3g4JD0LrHwOSFfgdSeeP1HOP7d8XvQ3or3P7gzk2xfdqIAbh2/nTy+sf372Du+OFiW3Dgl3MnZaPf//Kl/8PgAf1x9cqVNNtjY2Mxf95c9O/TCydPfDnfPHXyBAb164ue3aQtr3n75AGsmzAAJzcvV7g9KjwUB5bPw9aZo/H2iadYFxcbgyPrFmHHvAn4+Eba+eSCQkIxcsYCDJ3yBz76pf299hs/A20GjsGvsxYlb6ffydJNO9Dhl/GYs3yDpO19/dYbg36dgN+mz0VERNrrQSQ0LAyd+g7GO58P4v6h46fQY9BwTP9jEaKjZefRUrnt+Qh9fp2EecvWiv51SvS5b913GNoPGIlFa/5O/q4LDA7B+NkL0abfcBw7Kz9oSJkiwsPRt2sHsfh9+qTwMefPnMKoIf2xddN6cf+Dz3vx+MG9u2PPjq2Q2vYdO9Cnb1/8999/abYdOXIEbdu1S15u3rwpFrrdo2fP77rOnFVL1m9Fr18n48L1tP3N/SfOolW/kWL7jXuya4H0OFo3dNIcydvKsl/Dhg0RHByM8uXLo3379li4cCF69+4t9//s0qVLqFWrlvKCRBMnThRLytvpLRkRGRkpShhkJycnJ0ydOlXhNjrZDwkJyfKxv7Z/eq/la+3JLPonRJNTZXW7sgXf2oHEuBio61sgQEHAR8ehNHTzV4GKmiYCr8hOALTtS4l12nbFJW9vyPXNUNU2ErdD7+xO217HctB2Ko+EqBCEPZCNtg27txcxvo+g7VAaGuYukrWV/sk/2zETutb5EeXvDZ/zO9I8xqxELZgXr4kw7yf4eEPWXlP3KlDTMUBMiLQX/uKjI/Bizx8wdC6OoMdX4Hf/jNx2VQ1NWJSqB7Oi1eB35wSCnskyE2ldQlwM4qPDJW3vhzcvseN/8+BWrhIObFiOlw/vpXlMbEw0bpw+iie3r2d4H2V5dPUcrh/5F66lKmDHvIkIDZT//VrYOaN8w9YoULoi/pk5BjFRkbh39igeX7+AIh7VcXDVnwj88F6y9u78ex08795CsVJlMWZAL/EPK7XHnvdw+/oVeD2VXbQsUrQ4mrfrhMJFi2PMIGmD30vmTEdYaCgc8rlg/JB+Ch9z4tB/ePPSCx/eeX91nRT6zfgfLEyMoK6ujuHzV6fZfuXeY3RpXBPdmtSCq4Ntuuuk0m32GhR1tsMrX3/8seNwmu33Xnqja51KYrEyMYKmunryfVtTY5gb6kvW1ojYOEy4dBelLExw/t1HHHwp/7lJSExEv5NXEZeQgOYu9rDT103etuXxK7wMThu0UyYK+qwP/IDi2nr4N/QTHkdHyG2PSIjH2A9eMFFTRx19ExioqeNVTBT2hHxCOV0DnAkPwt1IaduszP+byl5yOx8fH+jrK/680Hpvb2m/q1jeQH8X1JdbuXJl8u30FnpMRgQGBmap3MXX1K1bV8yTpAgFQRWda2TU1/ZP77V8rT2ZFRQUhE/pXFTLyHZlK161NkwsbeDvk/Y7hOYprtSkjVgs7Byhq2/4zX2UifpBxsVqw6hwFQQ/u4TgR2kH+b3aPRXxUWEwdq8FTRPZOZGxWw1AVVUEkKTk4/sRw3+bihpVPLDunx04df5SmsfQxeHNO/fi9v0H4n5UdDSCQ0LQrX0rsaSeK1qZjM3M0bB1R3g9UTwg0crWTmwvV7UmFkwag0B/P/j5+mBEtzZwyl9QbNPW/XLupGy7dmwXAR+PipUwoG9vMYdDSlv+3oT79+6hZq3aGD1yBEKCg3Hzxg0M6NMbFSpWQvsOHSElW9ciKFq1Ll4/VDzI7vjGpYiODIdzsTLYMn2kWHdy8woE+ryDjUsh/D1lmKTtHTZ1PjQ1NGBuaoy+42ek2d66YW10bdkIcbFxGDVzoVg3a+k6bNl3BC3r1UTNimUlbW+PQSPg6uKMoJAQTJo9X+FjFi1fg09+/ggNC4effwAmTJ+DOjWqQk1NFbMWLJGsrfQ/qcsvY1C+VDFcu3MfKzfLB5Hpc9+tTXO0a9oAx89dxLZ9h8T6lr1/gV9AELq0aoqCLs6StVdDUxNtOnZGYGCAGICZGgWOfhs9Ah6Vq2Lnti04e+oEDAwMxT4NmjTD2hXLcOeWdNWrKOCzcMEC1KheHb9PmoTnz5/LbS9cuDC6dukilvfv36NgwYKilBfdr1e3Lkb9+quYx1EqOw4cxcFT51G1fCn0HDVJ/H2m5FbABd3bNEWlMiXQeeg4EUTM52An1l25Ld11K0kziZS55EKampo4e/asKP388eNHzJw5E3379k3efvXqVVFRwtraWnnl5i5cuKDwdm57oxwcHJBb5Lb2KFO03wtYNZwEFVVVRL65gcT4WKioaSRv17ZxEz+1zF0QdEv2T03burD4GXJ/P/TyV5GsrXRxJyE6DPpF6ovbASfnpXmMlm0x8VNVUw8xn2T/JKLe3oRZ3fFQ1f5S+1EKMcEfoW1qA2uP5ogrUUtkDNnV6CT3GJNCHuKnmpYuQl7KvviN8pdGhM8LSC3M+zFM3SrDskwD6Fo54+P1QzAvViN5u6q6pthOYsNDkoNYtC7omSwII6UH1y6ifsdeqNa0rbjQ7nntAvK5yQcuj21bj0Zd+2H5xGEZ3kdZnt64iAa9hqJg2coI9PWB190bKFGjQfJ2E2tbsZD7548jMiwUYUEBKF2nCcrWa44Xt68iPDQo+THKduX8GYyZOhu29o6fA0GPUchd9vlKsvufTejQoy9uX5ON6nMpUEgsZP+ubZDSnRtXsXHvUWhoaOD0kYOIiAiHru6XtN3oqCjcuHIRNeo2/Oo6qbz/5I+JfSaK77JyXUYofMy+M1fgaG2BmmWLf3Wdsr3zC4SduTEGNa+FkPBItJryP/zWuancYyi7aM/5myjmYo+G5YtBQ10NzSqVFNv+Pn4JLauWlqy9DwKCUcPOCq0LOKKIqRH+efIKTVzskrc/DQyBmooKJpQvKrff8dc+qONojSeB0l5cuhcVjqaGZqijbwotFVXciQpFYa0vF19uRYaiqLYeOhl/mXzeKyESZXQMUEPPBIHxcQhKSJvdxfKmkiVLYuPGjejWrRtUVVXlalbfu3cP7u7uOdo+ljvRgDjq57m5uaFq1apf7fPRY3KKkZFsoFdukdvao0yFSnvg41vFGTZ0obJU9fri9rWj/6Fxz1++uY8yqWnpwahARXE7JsgXarqyoFWS6MD3iI8Mg32DoXLrDfOXQ4S3LAgjpfNXrqNnp7bo1q4VLMxMcfLcJdSq+mV+AbJ283b0694Rv4ybnLzO5+Mn7DlwBI3q1pQ0SGRoZIxq9Rph8wrFmW3mltZiO7l56ZzIGqKf9Vu0RYc+gyC1M6dPY/LUaSharDgeeN7Hndu3ULlK1eTtFADu1LkL6jdsiEMHD4oskl07t2PUmLHo3rOn5O21csoPHX1DeJ5XnEVP/btanfuLx107tFM2sPTmJfSdtxZaunp4fPUcgj99gJFF5i8SZsWDpy+w8U/ZAOkyTdIGzetXk30WC7o4YdU/e8Ttf/Ydwb+r5qNgPidIKTw8AvRRGTmoH2JiYlCjads0j3nm9RL6+npwcrAX9/X19BATE4vNO/6Fvp4uAoOCJGvvw2de8ChTEv06t0WdqhUxZsZ8DOreIXk7fe6b15dlDNAwptdv3+HV23fw8f2EM7s2QmrUj67fqCn270k7GJtcu3IJbTt2EYuVtY3IKqpeq47Yh7x5/VJkFknl9JkzItDTtk0bETw+c+YMXF1dk7c7OzuLhQbP0LkPzQVDS7NmzcT2c+fOie+PrFyQz4qTF65h5pghKFPMDZ5PX+DaXU/UrlwheXt+JwexkK3/HUFEZBTsrC3FMne5LHOL5X3Ozs7YsEFxBmaVKlXEkhUZDhLRB0XR7e9BEc0JEybgxIkTqFmzJubOnZtuObcFCxaIE3CaOJXKAHh4eIh9KRCzbNkyFChQQJRco7Jw69atE/vNnz8fO3fuFDW1qU5fZmVmf0WvJXV7MoJeK5UQMDQ0FF82tWvXxvTpX0pcZabMCJVEoHZTKTx6z9LLQEqZKpvVbCs6HaUAEVHV0kdCbBTUUgSJSEJMOILv7IFxmS9toZOZqPeeMCjaGJJJiIOKumwSZ3EiraI4oS4+IgCRr6/BsFRbJCYmQEVLT9yPC3kPnXyVoSlRNlF8VATUdWSBKTVt2XubWmxECLz2zEdcRAgKdPhdknZ9LZMoqb3qukaIj0mbvh3h+xJvjq0VJfwKdpCVS8gpkeFhsLCVnfwZGJvindczue3+vj5ITEiEtWO+DO+jTFHhYdA1NBa39Y1MEBWRduQ/ZRpdPbgTjkVKwMjcEqXrNMWacf1w89h/MDAxg30B6S7sUFaOkbGJuG1iaobwcPn2Ht63G41atkF0pPzf9bpli3Hy0H607CBtaYekE1tiZGKCiLAwuSDR1g2r0bXvIFw8ffKr66QQGxcHHS3N5O8yCqikNrFPB/j4B+LsjfsYs3gdVvw2ROE6KYRGRMHEQPZeGurpICombUBiRq9WCAqLwJ4LN7F4z3H82k4WAPV6/wmmBnrQ05Z9d0shPCYORlqyvwUTbU2EpyqPFxgdi0Imhph29T7UVVTRp2h+GGpq4J5fEEaVKYKld59CShGJ8SJLiBiqquN5gvx3b0hCPKzUNUU5OlM1dbQzsoS7lh7+C/HD775eoMIVkyykG2WoTDTmS5njvnLnmDJ5AwYMEGWxSpcujbZt28LU1FTMU0TnpLRNqo4sy1vo4ocy+nyU/dK1a1dR5nDQoEHo0aOHwsdR34X6T0nL5MmTxYWX//3vf2LkLmUvUSYclVx79+4dhg0bhoiICIwdO1bMt9W4ceb7E5ndX9FrSdmejKLXSllLnp6e4ieVCenTp0+m2h4VFYUpU6bg5MmTsLS0FO8XlR5RVn8vo4I++YrzEj0j2flqTgt/6ykqFRg4l5JbHx8ZAh1rV7w/sUL0SS3Kt4amkWWOtTM0NAymxrL3zNzUBGGpMl38AwLx4ZMfShX7EuQ3MTLErN9GIzwiAguXr4WDrQ0qlS+D3OLskQPYs3kdHJxd4OxaECcP/AtjUzNMHzkQdk750GXAMGhqSXNuFxoaCmMTU3Hb1MwMYanKNHft3h1dOnTA8qVLULBQYdjZ2SHA3x/58rmgb68eKFO2HAYMGozcokaHPtgybQTUNbVQsmYj8ZmLi4kRASKiZ0z9xHBIFcKmgZNJdLS1ERsbBw0N+cuNn/wDsWDNZsz/TTawjdp85vJNzFiyFl1bNUbdKl8udCtTSFgYTD5/1uh6oiKrN27BjN/GiJJ0RFtbC1eO/4erN2/Dwc4O7XsNgFQoYGn6eTCCuYlxmswR8uadD4b+PhMhYeHYuXIhXr59hzLF3TFiyhzxuxjVvwfyOcquX+S0sNAQGKe8PhAm+yzevnkdi/6YAw11dfQeIE3/lISGhMDURNYeM3PzdCte0eCrjimuq9I8MCtWroSzk5O4ZiwV+vs1Nf7y9xAWnjZba+WWXfhn7yE0qFFZBDV/WJTpo8wyq4l5odeXvXJsTiJCAZ/ixYtj6dKloo4j3U+vnNuOHTtEYIgCKH/99ZeYdPfPP/8U0TE6OU5dco0m5KUPMQV6WrRogb17MzdvSGb3V/RaslICjl7r33//LUaAUoeIOiwZLeeQ5PDhw+I9mjFjhghcjRkzRowcVYSCWBR8S1qynPmkpoH4yGBx0T8uzA+qWl8uqhKapyjwykYYlWoDNZ0vo7iifR5Cy7IAVNIJ1CgDZThRwCoxIU78VCQ2yFuUmTMs2UY8ntonFnUNaJjnR8jNLZK1V9PQDBEfX4vbkb4voWlgluYxahrasChdD0aupeFzcRdyEnWuIj7IaiBH+DyHppFFmsdo6JvCsnR9aJlY49MdaecYSc3c2hZvnj0St18/eQhTK/kMm0ObVuLB9YtYOLI3PrzxwtGt6765jzKZWNni/QvZPD3vnj0U5TtScyhUFJVadIL3E098ePkM1w7tgm3+QijfqDWioyLgdU+69G0bO3s8fSQbjfnkoSesbb9kYpBNK/+HDcv/wrI/Z+PYgb24de2yWF++UlV07NlXZBKFBEs3UktLWxsfP/iIch7er1/BxMw8eVtMdLQonzdv8njxc8u6laJEXup1b1+9lKStdAJNGTnRMbEIDgsX5c9S8yheGC1rVsSUAZ3w7M37dNdJwcbMGI/e+IgLMY9ev4eVcdqszFqliqBV1TIY2aa+mL8oyboj59G5jixjUirWejoiW4g8DgiBta6O3HYbPW08DQxFCXMTEUyaf/MR9nl54/anQAw9fQOvQsKx4Jbse0IKlmqaonwc8YqNhIW6/EANuv8yJhKltPURnpCA3SGfcCsqFOoqKqitbwJLNQ1cjAiWrL1MuczMzHD58mURJFq1apU4H6SskDlz5oiBV4xJiWr8UzCF/v7GjRuHsM8XhBSVc9u6dasIlvz2228iE+7+/fvib5hGe9MkvalLrlFfhkb30qBBc3NzUTYmMzK7v6LXkpUScPRa16xZI+aUpTIh1I4bNzJ3fkb70PtDgwPbtWuH5s2bK3xvs62/l0EX9++AR8OWyA2Cn1xA6MubsK7WLc02DQNzRPu9ho51AXGbgkU5ycHOFvcfyUov3/V8BHtb+WD+0nV/i/mIaO6hZ16vMPevFdDS0kLzhnXRqXVzdG7THE+eSzsPzbe4FCqCZh27wfe9t5ijyMrWHi+ePEKFarXw9uULHNghXZ/a3sEeDz1lc/d43rsHO3v5Psna1atRtVo1dOrSFV5eL/Do0UPY2zvg+bNnqFe/AY4eOSzK1eUWNFdRmfotULFZR3ieP4GIkCAYmJqJuYjoXNv35TMxWFAqCfEJCAkNEyUQKYiROkD0/NVbjJj+J+aNHwYjA/3kIGdUdAxqVyqPXybNE5k6UqBMvZev38r6e+99oKsjP3/jzTv3cOz0OXTqOwRnLlzGqImyDClDAwPUrVENp89fROECXzJNlI2Cv/cfywaf3X34OM13Q9J72b1tC5QrURQrNm2Hg401Hj59jlJFi4hSY0N/n4XcwtbeAY8fyj6LDz3viesFYr2dA9p37gZDIyMc3r9PsvbQ/8T7n78b7t29K0rJKUoKOHnqFGrUqCE3sIb+h1OpucsK5jlTWnttrXH/kWyQ8t2HT2Fv86VKRBKPUsUxsGs7HDt3GW/fS1cKj0mDrvVTkP1bS0anA8pSJlFK9E9nz549uHjxIgIC0k5Kl17KU2qVKlVKznChiZdevnwpShuk91g6+SUULKLRVmXLlhWdYEU1oGmSJhopmXQ8OmnOjMzur+i1mHyORmcWvT6qH0gowLNp0yZUr149w/vT74XanvQF1qlTJ/F66EJBauPHj8fIkbIatkkjy7LScdAvWAO+h6dBVV0bOvYlRUAl7MkpaFkVhoaxLfxOLxbZOzQfkZquCUwqyE7Uw56dgVGp1pCatn1J+J+YC8THQbdQbVlbHhyEboGaUNXUReCZxdAwy4fgaxugbuIkStNpO1VA1OvrgKoa1HSkGxmnrmsIbTM7eK4YgtjQAORrKft9vTq4DM6NByE2LBAv9y2SZWUFvId5Cdnr8T65CQEPz0NVQxuxof6wrihNZ03friCigz/iweqRiA70QeHus5GYkIDXR1bCudFAhH/wgvfJjWJdpO8rGOYrIfbz2rsQwc9vIcrPG1YVmsKsaMb/5r9HyWp1sGv5fDy7cwOf3r/FtL8PIOCjDy4d2osmPQaiSpPW8P/wTjz23cvnKFiyHGyc86fZRypl6jXHksEdcPXgbkSGBqPNqGl48/gefF+9QLkGLXH10C5RZi4uOhof376Elo6eGGH28PJZUerD58UTaGhKl43RulN3/DqgpwgOmZiZibJzZ44dhra2Djyq1cDg0b8hJjpKBJKeP34EOwcnrFmyAPdv3xCTXlL9ck0J29uuW290a14P+gaGqFS9lsjG3LVlI0qUKSdK4I36XZbZefHMKdH5MTE3T7OOMpAka2/dKijXdTji4uIxtKMs5X3Gmm0Y2KYRzIwN0WXifIRHRuH+s1cY26ON2K5onRSM9HTg7myLKsNmwTcwFP/7Rfa/e9SK7fhzQHuRQdR3wQbExyfg9vM3WDWyu9geGxePc/eeYnpPaS84FTY1hH9UDLodvYR3YZFYXqucWD/7+gOML+cOZ0N92OnrYL+XN/wio9Eonx08rM1h/jnbySskDNXspOucl9UxwKagD2Iuoo9xMVhgUwAB8bE4ERYosoZKaRtgZ/BHnAwPxIe4GHQztoamiioeRkcgPCEeb2OjUfpzFmheR/FSZU4blAemJBKok5uZbHbGFGWqUL+OAhGp5+/Inz8/fv89Y9nrHTt2FNURki6qJNX0V4T6U1SVgVDWEGXHUFmXVq1a4fr1tGWJqY8zb948lClTRiwUMMmMzO6v6LVkFb1Wms+IUECMArs6OvIDEr7V56MAE/WJaaEg2uPHj8VtZfT3yJG/V+D6MdnE3mGB/mjYYzC2L5yGtsMmitKWdCHt3oWTaNBt4Ff3kUKU32t4H1oEfaeSeL1nhphriOYnen9yJWxr9xeBIW2LfAj0PIn4qFAYuMjet/cnVyPU6zrCvR8itng9mBSrI0l7a1bxwNhpc1C3dRd4vXqDM/9tR1BwCBavWo/Jo4ehZeP6KOFeRDz28fMX4vFXbtzGH/9bJQaaUoDo3IG0c9cqS2hIMGaNHiKCPeP7d0Xv4eOgrqGBO1cvokXnnjh1aB+O79uF2JhYeD19CD19A1Rv0AR7t6zHkX934I3XczRtL7vmIYXOXbqifZvWcHR0EhfPqOzcubNnREZRoyZNRMBt37//4s6d23j88CG0NLXQsXMX9OzWBS9ePMeL58/h5CRdxvWzm5dxZttqvHv2CBsmDkaPGUtx6/h/MLNzhJNbSahraOLWsX3QMTBCoO87qKqro1Lzzlg5sjv0Tcxg41oE2nrSndf1aNsU5Zt3h7q6Gjq3kJXgXrBmC9o0qg1HW2s07PELnOxs0HvMNFFybs7YX9C3Qwss3bQDBvp6MDLQg6am/AAnZWY91apWGR51m4o5vWb8NlasHztlpsgecrS3w9RxvyZn+DVtIPuepjmJbt/zhOfDx9i/Tboybk72ttDW0kTVll3x9r0Pti77U6ynLKGFU8bh5Rtv/Dr9D/H9++TFK/zSqzOsLMxQpkRRbN69HwFBIahTVdrBdkP798b1q5cxfuRQ9Oo/COU9KmH5koUYPWESPCpVwYzfx6N9s4Z48/oVduw/gpvXr2LFXwsRFxeHxw8foFmrtCUAlYWSAihAfOTIEZEVTNnFlM1MWbp0XZVQIkDVKlWSSzhTEsGWf/5BbEwMPB88gJGhfDlTZerephka9xiCBWv+RnxCgig7d/bKTQQEBaFlg9pYtmk7Tly4isioKBEg0tHWwlOv1xg3ZzGevXyDVv1GYtnMCbC2+DIYNq9SoYpPSsz2UcnFmUR0napRo0bi7ze9jEias0iSIBF9aCiTp1atWlkOhKQs6UOSTioz8lhRVufzfbqtaD/qGFM2D5Vaow4OdSrq15fVSM6IzO6fmdfyLVQOhEor0Gg2akNmT+Kp7VT2bsiQIeJLljo/o0aNUvhYOhmi5XvpuVSGpqmTKIWmaZ5frNO0cIXq56wh49LtkRAnG+FMQYsk+oVqQcNQ+pIn+m6NoGXtThPkQMNYNopI06pI8jxKhuW+nLCqasteg17BWtC0LIjEmEhoSFRqLkn+Vr8i7O0jaOibiIARMSkiq+urqqkD81L1oKKmBm0TW+hYOor1RgXKQNdG9rvQMpa2dIJbz3kIffsYOuZ20DQ0FwEsk8Ky9moamMJCtFcdOpZO0DaR/f5N3avCuKCsRAatl4q2jq4I8rx67AnHgkU+l3ALR5FysvbmK1JMLETf0Dj5dup9pGJmY49Ra/biw6tnopycmro6TCxtoaklu6jgWLi4qF9NwSGHwsWga2CIyi06waV4GTGHkY1LQZGNJJWS5Spg497D8HnnjRKlZRfZ8xcsDLXPpdGq1JR1vEuW80BQgD+sbGxRvnI15C9YSARqipUqC+1MXDD5Xg2bt0Yht6IIDQ5G8TKy9hYvVQbmllbiH3GtBk3EOppXib7rKUU+9Tqq0S6V8b3aoZ5HKfF+liwo+16qU74k9D6PiOtYvzoSkYhCTvYo4Gib7jqpLB/WDVceecHKxACudrJRT009ZIFiHU0NdK1TCZrqaiiazx72FrLPVUR0DBYP6SRpnf0kq+qUx91PQXA00BWZRaSW/ZfRWvOqlsLdT4HQ11BHYVNZ2r+ToSyTlrKLylqlzfxUFi1VVSywccXz6Ejk09SGoZo6ohISUFxbNlqTMoamWbrgSXQEzNU1YKch+98/zyo/vOOiYK2uCfsU/58ZYz83Ondr0KABHj58KCo36KaaYJ5GYOemPh8FeF6/fi1Kt2VGZvfPzj4fTSxME5LTa6MAEV2Moj5gZtpOJQErVKggRjJTCTxbW1ul9fdI4TIVYfW5BLORmax/UaJqneT/0bHRUegwaoo4P/3aPlLQ0DeDfZMvfWBtM1kfydD1y8VS+0bDEfH+MVTUNKFrU+Dz9grQc5SVDtIyke48if62zh/YgRt37sOtUAFYmpuJyiT1asgGrRYrUkgshEoNlS9dEu8/+KJb+1bQ09VB+dIlRKaDVLS0tNGwdUexELPP58run8/3XQu7Q11dQ5zHFypaAkafS70t+nsPHt27DXvnfLCxl/1OpOBetBiOnjyNl14vULacrM/p4pJfZCmSUaPHoG69emIS8GLFi8PKStZH3b3vPzzw9IR70aKwtEw7Yl9ZzO2d4NFM9t4mfb7sCriJfh5pN3Y2Xj+4g5jIcNEn1NbVR5GKNTBg4d8IDfSDs7t8eUVlG9S1LSqVKSGuP5UtLittXq18KZgYyv4m/5oyJvm9NjGWvYZurZuIx3745I/KZWT9Aan8MW0irt64DSMjQxQpKMsKalyvtvgbtjA3Q/NG9cQ6R3tbFMgv+/5qWKcmqleuiLIli8PMVLr+P9mxYiGu3r4LRztbETQizerWTP4+6NamuagyUcDFCQU+z/G0au4UXL51F9paWihbXNq5KJu3bovGzVuI264FC4mykjVq1UkO0u0+eBx3b99CgUKFYWFpCU0NTbTp2FkMDi3s5g6bVNVHlImuu144f16Uf6WKTnp6emJdxYqy60HE0dFRbtA9lZfr2qWLODeifejxUnEvmB8X92zEs1dvUKGk7H9Vfid72FrJKvhUKlsS9tZW0NfXRfkSRUW5Ofq77t6mqViIgZ585SeWtwwcOFAMKlq7dq1ILKEkE0qkyUpQKFuCRDSi7Pjx41meCEkKVAaAaj1TzXVKq6dJnaTc/3vQCT+NUKNOAj33qVOnMtVhoLbv3r1btJ3+SderVw9Nm8pPDq4MGsbyaZmapl9O+rRtFf9T0raWjYbKCRqm8oGIlHMMaduVyNBrlIqKqhoMnOTrnBq5yE781DS1YepWOc0+Bo45Nym1qoYWjFy+vId0YmvkIpt8XkPPWGF7jQvIj3qUEgVS3Mp9mRhWW1cP+d1l7U2pcBmPdPeRkoGpuVgU3acgEC2p2bgUEktOoOwgWpI4OH+Z3ymJqZm5WEjx0jn3t0AoYyilgm5pawynfD1fWyeFMm6yixpJqJxcksZVZR31lBStk4qamioqF5Uvz1CjpKy9WpoaaFappMIMpFKu0l1ESElLTQ3lreUDPRVsvnz2NFRV0w0ESRkgSqKnqoYSOrKgENFWVUVhLV25+ym3E2sNTbH8SBISZYsyj58X0PwLSXO8pM4CGTx4cLZ0JNiPi6oi3L59G15eXqJyQ25FpTQoG4fKkNPAOkVlYpS5//egi0s0ry59VqkqBfXXkkrqZQRlctEoUhq8SdlBVFZSUZAoOzm7lYAz5PtJhct+OR/W0tFFvlTn0Ir2kQLN5WpU4MtFviT6Tin6KKpq0LOX7zPpOxVHTtHX00ONyl/6GxTcq1gubTWQqhVlQQ5baytRbi4n0EXfavUapVlv+HmuEUcXV7Gkpm9oiHJVpKkYkRrNM0RLEvtUg3GLl0h7HkqBIcta0gWHktCgvtQD+6ycXeWC1PmKpf3bsHBwFktOKOkm3wdNChaRhjUU95vdCriIRWp0fcIj1WerWqW02Talin/pB1atKM2cSYpQllXVCvJ95JqVZe0xMjRA8/q10uxDgYEqCr4/pFC7nmxe2ZTKeXz5G9DT10elqtWS71vZ2KC+jfKvWaaHznNSVm+iZIiUCRGUOZySq6urWHKKnbWlWJKkLDlX0q2QWFKXI2xeTxZU/KFQpo8ys30Sc2cmEf29UiIILZQUQuWLKahJf6cULKKsd8MsZrepJNIwsUyysLAQo6wowppVlJ1DgY+kE3G6TaNnKCjyrcdSWr+xsbE4sabRV3TfyclJjEzw9fWVy7zx8fERE3n6+/uLSBuVK8iMjOyf3muhx6duz7dQLWqaY4kmQaPntrGxEf/AFL22lBRtp9rY9I+BJivOKOpg0O/ArsMKkaGS28VEKHfi1ezmXCJnLtJnVU6M3P8eQxp9uTieF/hFyEZT5RW18uXei0R5XeF3Z5GXxPvnrdrGL/6RlbnJK16fUzxham4UkRCPtm8fIDg4OMsno9kl6Rzm0av3MFBiW2iC2yLOtrniNaeHzgGTRjbSqGga7ESBIiobRnNoUhk6ypxgLD0vXrwQJSzob+Z70Hw9lHWUFGiiCaGtrKwUlsZI/Vhqg4uLizgfpUAKldSifhl99qgPmDRyl/ahfhAFSCgbiC4Cp5xE/Vsyun96r4XalbI9GUFzz9LxqEIH9TNpcB9J/dpSS72duvJUHof6e6mzvb71Xbn45H3o6OeNMqNLdsvmi8grbo7IW32+a5HSZkV8LzeL3H+dIqX1t32QlwyzSzulRG4Vr592/uPcTDUqFHnJe720gztzM0tdWbWSvELDVzbXdF5Ac46ZFq+WK/o/Secxfs/vKTWDNiQ0FOauxXPFa/4WOk9esWKFGPw0evRoMW+RZJlENNfN4cOH0aZN1ucy0NbWlhupRSe6Dx48EOnycg1UVxdpfykfm3KEFAVjKEBEqLOROohCQRZCHQo6fuvW8vPfUKZN6k5A0nNmZH9F7Ut5Uk/tyczzJs3nRJ2hlK8z6bV9rQ2pXzsF8xhjjDHGfiZ00TQLY6AydfzcjjILaF7Nbdu2oUmTJqKMFf3cunWrmJ+EAkWMfU2+fPnEQDjKJKJATVbRwL6UqL9CGW7Tp8vm80tCQUtan3reoyQ0UC9psF7qQYU0KC6p30T9QkXHpyBK6rl0k54zI/sral9S3ysp4JWZ500qY0N92aQAUcrXltE2UJ9RyuwnxhhjjLFcgSeiBfVLz549K0rP7dmzB+XKlRPVxLIqS0EiOommmncUKKIUu9TZBuPGjctSY6gjQBOAKSuTQdHxKQMn9Ui29J4zq+3LzPNSx4Sif5k5Vl7L9mCMMcYYY8pDk9cnDSqii9CU6UCo/ACV1bp161aagVmMpURz3FDNfQ8PD3Tu3FlkzKRE2TZdu3bN0pvWuHHjNJlsGc2CyerxKfsm5XxCX3vOrLYvM89LI1K/Nq+Tst8jxhhjjDGWN3l7e2Pjxo2iTHFYWBi6desm+neFCn1fJnOWgkQXL14UnYa7d++KJbWsBokou6hwYeWVi/re42d1/8zu97XRYMp+jxhjjDHG8jKqHq3MCtK5szq1vIiIiOSy0JSl8Pz58zTbGfsayn6hcmo0WfP58+fTbKc5rbIaJKLsotQZRtnpe4+f1f2z83Up+z1ijDHGGMvTftI5idasWSPml6WMofnz54tqEZkps/w1WToKzZvDGGOMMcYYy92o49C9e3eRzUBzij59+hQlSkg/iTzLW2hAIPf5GGOMMcYYy13Z/lSd7MCBA2JJz2+//ZbpuYmyJ9TEGGOMMcZYLkAzBilz2qDcPyMRxKSlNKdMUtkq6iTQXCaUkU71qmmeFMYYY4wxxhjLi1QSE6GixGwflVw6D22rVq3E1D/f4u7unuljq2ems0koCpV0Oz2ZjVQxxhhjjDHGsgfNI5PSiBEjxMLYt+qb09yoDg4OIrj4tXlS6TH9+/fnN5QxxhhjjDGJuLm5iUUZMhwkunDhgsLbjDHGGGOM5RYJiYliUebxGfsRhYSEiH4edTyrVq361T6fsjqnjDHGGGOMfdNPOCfR+fPnxXyhTk5O33zsmzdv4OXlhRo1amR/kOjMmTMKbzPGGGOMMcYYy9so8MN9PsYYY4wxxnKf4OBglClTBo0aNULXrl1RqVIl6OnpJW8PCwvDxYsXRZnxgwcPYs2aNZk6Ps9JxBhjjDHGfqw5iZR8fMYYY4wxxhhjOeQnzCRq0qQJ7t+/j7lz56JNmzYIDw+Hra0tDA0NRUWA9+/fQ1dXF927d8fdu3dhb28vTZAoMTERb9++FelLcXFxctsyk8rEGGOMMcYYYyz3oX7ey5cv4ePjg4SEL51l6oyWLl06R9vGGGOMMcbYz8TGxgaLFi3CnDlzcOnSJXh6eiIoKAjGxsZwd3cX2UU6OjpZOnaWgkSvX79G+/btcfXq1XQDSIwxxhhjjEktIVG2KPP4jP0Mzp07J0pZ0KDA1CpUqIArV67kSLsYY4wxxthP7ifMJEpJW1sbtWrVEkt2Uc3KTsOHD0fBggXx6dMncT80NBQnT55EoUKFsHz58mxrHGOMMcYYY4wxacXHx6Nz584YOnQoTpw4IYJCfn5+2Lx5MywtLbF48WL+lTDGGGOMMZaDcxTduHFDzCl6/fp1kVH0PbKUSXT+/Hncu3cP5ubmctEr6jR06tQJAwYM+K5GMcYYY4wxliWJlNWuxPeOM4nYT4BKzKmqqmLUqFGiegSVnTMzMxOBo8DAQCxdulQEjhhjjDHGGJOaSmKCWJR5/Nzq8ePHIoHn+PHjcuWg6dy9Zs2aYjAXlZ6TJJPI399fTIxETE1N8fHjR3G7SJEiohQdY4wxxhhjjLG8ifp7VPM8dX+PcJ+PMcYYY4wx6QUEBKB69eqIiYnBpk2bcO3aNTx9+lRkElHyDgWKqlWrJnfurtRMopRowlKKUI0ePRpr1qyBs7Pz9x6SMcYYY4yxLElAoliURZnHZiy3SDnHrJOTEyIiIrB9+3ZRPWL16tXc52OMMcYYYzmHMmhSZNEo5fi50N69e8V0P1QOmgJCKZUtWxYdOnRAvXr1sHv3bgwcOFD5QaL27dsn3541axYaNWqEefPmQVdXF1u3bs3KIRljjDHGGGOM5QJUVrxu3britqamJhYsWICuXbsiNjYW9vb2omPKGGOMMcYYkzaTyM3NLU2AKImKiorYTo+TpNzctm3bkm+XK1cO3t7euH//Pnx8fNCsWbOsHJIxxhhjjLHvRgkQyl4Y+9G5urpi+vTpyfe7desGPz8/0ed7/vy5GMHIGGOMMcZYjvhJO31ly5bFrl27RHk5Re7cuSOy/+lxmZWlTKIWLVqI9KYkWlpaKFq0aFYOxXKp6LAgqGpEIbeLDst8ZDQnRYW7IC9JSMidX4rpiYiNR14SE5c701fTE5fH/h70NLM0DiJHJMbnrb/dvCYxPm/97YZHxCKviEjgv13GfkRU23zt2rWYO3du8jpDQ0Pu8/1AOrrqwtBQF3mBZbfSyEtCdi1AXlK+QlXkJQkvg5CXDPJojbwkIVQFeUXs6bxVyUi7eGXkJVbPbyMviSjfBnmJqr4F8or4RO2cbgL7rEaNGmjdujXKly+P4sWLo2DBgjAyMkJISIgYyHX79m10794d9evXR2Zl6QoalRegutSMMcYYY4zlJhTPVvbC2I+OJsM9fPhwTjeDMcYYY4yxtBITlL/kUitXrsTJkydRpUoVUVaOMv0p479ChQo4duwYNmzYkKXjZimTqHbt2iKTqFOnTll6UsYYY4wxxhhjuVPhwoURFhaGx48fi9uMMcYYY4yx3KFWrVpiyU5ZChI5OTmJ1KX//vtPTIZEk5mmNG7cuOxqH2OMMcYYY4wxCX38+FGUsPDw8ED79u1F/y/lBLl2dnbo2rUr/04YY4wxxhj7AWQpSHTp0iWUKFFC1LqjJTUOEjHGGGOMsZyg7HlGc+kcpoxlKypd4e3tDVdXV9y8eVMsKRUrVoyDRIwxxhhjLEeoJCaIRZnHz6v++usvREdHY/To0coPEh04cADW1tYKt3348CErh2SMMcYYY4wxlgsUKlRI1DQ3NTVNsy02NhahoaE50i7GGGOMMcZY+kJCQhAVFYXMylKQyMbGBonpDKP82jbGGGOMMcaUKQGJYlHm8Rn70VHm0PDhw3HlypVMbWOMMcYYY0zpKNNHmdk+ibkzk2jfvn3YuXPnVx/j6emJJk2aSBMkSk9kZCR0dHSy85CMMcYYY4wxxnIJ6vPp6urmdDMYY4wxxhj7qTx69Ajnz59HkSJF0n1MRERElo6dqSDRlClTFN4mCQkJYlQZTXDKGGOMMcZYTuA5iRjLOpqHaM2aNeInLan7fFRq7vDhw6hWrRq/zYwxxhhjLAc7fcrMJEpEbtS4cWNs3rwZR44cSfcxM2bMUH65uRMnTii8TTQ0NODs7Iw///wz041gjDHGGGOMMZbzNcypn0dzDiXdTomqRlSvXh2///57jrWRMcYYY4yxn1GxYsVgbGyMCxcuoEqVKgofo6qqCjU1NeUGiagBpEWLFti7d2+mn4wxxhhjjDFlSkhMFIsyj8/Yj8rNzU30+Z4+fYq1a9di7ty5Od0kxhhjjDHG5CXGAwnxyj1+LkVzEn0tCDRhwoQsHTdLcxJxgIgxxhhjjDHGfkwFCxbkABFjjDHGGGO5jI2NjVKOm6UgEWOMMcYYY7lRfIJsUebxGWOMMcYYY4zljMSEBLEo8/g/Gw4SMcYYY4wxxhhjjDHGGGOM5XJ//fUX5s2bl+6cRDRvUeXKlUXpOQcHhwwdk4NEjDHGGGPsh8FzEjHGGGOMMcbYDyxByXMSJeTeOYlIrVq1sHr1aiQkJKBTp06wtraGr68vtm7ditjYWHTv3h07duxAnTp1cOfOHejo6HzzmBwkYowxxhhjjDHGGGOMMcYYy+XCwsJExtCNGzegpaWVvP7XX39FpUqVUKZMGQwbNkz83L9/P9q1a/fNY3KQiDHGGGOM/VCZRPGJiUo9PmOMMcYYY4yxHPKTZxJdv34dNWrUkAsQEU1NTZFlRMEj2l6/fn14eXll6JiqSmorY4wxxhhjjDHGGGOMMcYYyyZ6eno4c+YMoqOj5dbHxMTg1KlTYjvx8fFBvnz5cneQ6ObNm2jevHm2HvPBgwciQqbIqFGjsGfPniwf+2v7p/davtaezPrWsaRow9fER4Ug+NoGBF1cgdjgd2m2B19dj4BTfyDo8mrEhfiIdYkJ8Qh7eBABZxYi5NY2pbdRrr1hnxB++x+E392BhKjgNNvDrq5B6JVViLi3GwlRIWJd5MODCL28UiyxH59I2t6wN/fxYusEvDm4EPFR4Wm2P/t7FJ6uHwrvI/9DfExUhvZRpqDH5+G1dRzen1qNxPg4uW1xkSF4tuEXPN80Ah/ObkBiYsI391G2k9vX448B7XB4w9I0294+fYBpXRpiTu+WOLltnVgXERaCTTPHYm7fNsnrpEL1Ro+sWYAVwzrjxuHdabY/uHgSC3s3xfKhHXH39CGxLj4uFnsXT8OyIR1wfud6SdsbGRmBmeNHoF/7Zrh4+ni6j5v926/YvmG1uE0/O9Sviv4dmuPm5YsSthbw//QJI/r1RO/2LfHg3p10fweDunfCySOy93fVXwvQvFYVsWxavULS9j55/Q4tRs1E27Fz8NrnY5rtQ/9YhXLdRopl35kr6a6TykXP52g8YRH6zF+PoLCINNvbTFmKCoOni+Xm01cIjYhKvk/LlI17JW3v8Tc+6HXiCubefIjYeNl3VUpewWEYfu4m2h6+gIvvPyWvv/0pAF2PXYbUjkQGYHLQa+wJ90uz7U5MGEYHeCUvt6PDvrlPXpWQ+GVeIuUsOf0KGcs7li1bhlmzZmXrMdetW4fJkycr3FalShW8e5e2L5JRX9s/vdfytfZk1reOJUUbvubm7Tto0roduvUZAD9//zTbaV2/IcNQsUYdLF0pO6/r3ncAyletiX0HDkJq47u1wPDWdfDc867C7WM7NxXbafF99xaxsTHi9q/tG2Lt3MmIi42VvM2+IeHovW4/2i3bBU/vtOd2JCEhEZ1X/ovjD2Qjgxcfu4oaczbij8OXJG3r7HU7ULv/BCzbcTDD2w+cv4aGQyahbOfh+BgQBKmdu/sYDUfPRd95qxGs4FxU0fZ3nwLQY9YKePSfhI1HzknSzqCgIPTp2w+NmzTFxUtpf6/DR4xABY+KYqlWoybi4+Oxes0acX/0mLGQ2uu379Cm5wC06NIHT54rHrFOFzIbtO2KC1eui/sPHj9F88690ah9d9y6e1/iFsv6dFO2HUODqaux+czNNNuXHb6EimOXoN6UVdhx8ct3yK5L99Bo+hqxLTJGmu+Iw+euoE7P4Rg05U9ERMqu9yR58eYdyrTsDY92/TBsxiJEfb5g/LV9pPAxMARdpy9Dk7F/4PbTVxneTr+Xtr8vxoFLtyRr68L5f6BJg/pYuXxZmm2JiYmYOul3NKhTG/PmzBbr9uzehWqVPJKXa1evStZWmmtm9O9TUbd5W+zauz/N9sPHT6FcjfpiGTB8tFi3dde/4j7tc+jYSfwoEuPjlb7kZm3atEFoaChKly6NmTNninPB2bNni/JyHz9+RIcOHUQG0f3799G0adPcHSSiDxr9I8tO7u7uOHz4sMJt9Fz0ZZNVX9s/vdfytfZk9/slRRu+JtxzP9SNHaDtVAGhCgI+em6NYFC6IzStiiD0nizYFvH0OGIDXsOgWAvoFqwNKUU+OggN62JQN3FC5JNjabbruDWFrltTqOqaINpLdiIYHxkA3aLNxaJm4ihZWymY5n10KSzKtYC6jiF8r+xI8xj7+r/AvuEvgIoK/G8fzNA+yhIXEYyPF7bAslJHxIUHIdBT/p+QmpYe7BuOgG3dgYjyf4uQZ1e+uY8yvXniiatH96JZv5F4cPUcntyUv7Br6ZAPPSfNR9thE3H12D54P3+MC/u2Q0tHF62HjMOL+7fw8oHiYIIyeJ47Ct9Xz1G3x1Cc3b4GgR/kL2TkK1YG7cfNRd2ew3B03SJEhoXizqmDiAwNRqP+o3Hr+D588k57kqYsW9eugKqqGnoMHIZ5k8YhKjIyzWOuXzqPiPAw+H30FfdrN2qGqQuWoXOfgZg5YSSk9OfMKSjsXhRtOnfDxJG/KHzM7q2boaunh+CgQHH/g897/PLrOMz5axkaNGshaXt/mbcSbWpXQq1yJTBqUdqApde7D1g5YTDWTPwFVUu5p7tOCvQ/9JclmzGoWU1YmxphzlZZkC2loPAIrBrZQyyFHKyhq6WZfL9uGXeUyC/dd29wTCz+uvsU/Yq6Iig6Bv96vZXbHkvBwjPXUcrCBNM9iqOYubHsdSYmYp/XO8R9xzlHVryJi8LZqGC01zWHZ2w47sXIDw4ooK6DwYa2YjFUVUMBDZ1v7sMYY9nx3f89fTBFevbsmW5AJC4uTvSLsupr+6f3Wr7Wnsyi43+tzydFG75m4NAR6NOjGwoWyI8pM2QXy1LqPWAIdLS1sXTRn2jVvJlYN3HsaDRp1AD+AbLzJin1nzgb+YsUQ2RE+v/fhs5cJBZTC0uoq2uI2wN+n4PgAD+cPZB2QJayTdt3FiUcrdDJoyiG/XNU4WM2XboHA21NBEbILvy2K++GUQ0qwjswVLJ2nrh6B+dueWJy/07YsP8E7j9/9c3tz968R99pf6Ff64ZYM3koTAz1IaX4+AQMXbQRv7SqD3MjA/yx9UCGtneethS2ZiZYPqoXGlYoKUlbZ8+ZC1tbWwwcOABDfvklzed+2NChWLVyJebMngVHRweoqamhaZOm4v6r168htRETp6Fu9apo1bQhhoyZqPAxS9dugpmpCYJDZH+nY6bMQtsWTdCrc3uxv9T2XnuAx+8+YkKb2lh04DzefJL/jmrlURQrB7bGuFY18fuWIwiLisadl+8wZuMBDG1cRWzTUldTejsjo6Ixas7/MLp3J7rcgyWb5b+X7KzMsXbWOPxv0kj4+AVg83/HvrmPFH5fsxPF8jugZ6PqGPjnugxv33jkPPR0tBAQIk2/5Py5szhz+hTGT5yIHdu24v49+UEFJ44fw+NHjzBt5iycO3NGBISqV6+BpStWikVf3wDFS5SAVDZt24lPfv6Y8OtwTJ79B/z8A+S2BwYFoUWThlizZAF+HTpIrKtTo5q4P3bELxj52yQRaGJ5n6GhIa5du4bGjRtj+/btGDt2LP755x/UrVtXlJozMTERGUS3bt2Crq7uz1lujiZtyk1yQ3ukaENc6AfoFqgFbftSUFHTQGJcjNx2dQMraBjbQ9OiINQNrMW6aB9PGBRvCQ2zfFDXt4BURMcvIQ6aNsWg5VgBCZFpRzCpGVhBzdAGakYOUE3RtsjHRxD9+rKktSljgj9C28IJhq7lYVWlMyLePU7zGB1LZ+haF4CBSxkkxEZnaB9lifR9DsOCFaHvVAIWFVoj4v0jue0qqmrQsXKBrk1B6DkUFe391j7K9OLeLVRr0RmFSnugZtseIuiTEgWDHAq6I3/xMnAqXAwxUZFwKVoKV4/uw8YZYxAa6I/n925I1t5XnrdQtW0P5C9VAaXrtcCbR/InMbqGxrAt4AbXUh4wtbZHfFwMEuJi4VysNJyLlYGVk6u4L5W7N6+hzy+jUKlGbZStWAUvnz+V204XQk4c3IcmrTskrzO3tEIh92KoXLMu9PSk7Tw+efQAfX8ZgYbNWkLfwAhhobJMwiQhwUF45fUcZSp4yK1fvng+lvwxByFBQZJ+l4VFRqJLo5oY2KYhvH0VZ4L8umgdFv6zD1EpRropWqdsbz8FooC9FRp7lMDELk1x7XHaUYYh4VEYvmwr1h0+D/qqVlNTRYn8DmK5/9IbjSsUl6y9jwKCUdPeCh7W5ujtlh93Psn/bp8EhsBKVxvdi7igsIkhDDU1xPp/X3ijjasDVKlHJqEnsZGorW0MN009NNQxxdNY+dGxeqpqcFbXhrGqOgxV1aGvqvbNffIqSvpS9sIYyzkqKiq5oo+Vm9ojRRv8AwJgYWGOFk2bYPyvI3Hn3v005U1u3LqN+bNnoHTJErCxthLrC7jmh72dHXKCc8EiMDIzT3e7v68PVs2YgHMH9ojBdvQ+uhQuCteiJVC8QhVER6Ud3KRsTz7445fa5dCidGHoaKgjItW5WmB4JN4GBKOss03yOhtjA7hamkjazsv3HmNQu8aoVroo+rasj6v3n3xz+6EL19GrRT20rFkRJQu6QENd2imyX/l+gns+OzSqWBITu7XElYfPv7n9U1AIvP0CMLNfO5Qq6AxLE0NJ2nrt+nX8NmE8mjRujAIFCsA7VYYjXfgrUaI47t33ROdOncQ6a2srFCnihpzwzscH/Xt0Rrf2rREalvbi/vsPviKbxb1wweR1FPhqUq82GtauAXU16adLv/rkDYY0qoxq7i7oVLUUrj+XHxBmbWKI4s62YntxZxvoamrgwI1HYp8GpQuLbVJ89z988QrVypVA/SrlMWFAN1y+4ym3XVtLCyWLFEDZooVRvFB+FHZx+uY+UvB8+Ra/dmiM1tXLw1BPByHhkd/cHhQWjufeH1C52Je/E2WjoE+ffgNQuUpV9OjVJ01WUFxsHEqXLYvyFSqgeMkSIsPUzNwcxYqXgI6OLtzc3aGtrS1Ze69cu4FRQwaiZtXKaNmkEW6n+l9Mdvz7H8ZMno5nL16K+xbmZihRzF0Ei2ytrMSAmB8CBc+VveRy5ubmmDdvHu7du4dPnz6JrKEFCxbA0tIyS+eHOXo2GxERgXbt2sHU1BStW7dOd9QUlVKjSZeaNWsmImGDBg3Czp07YWdnJ/45UuQsdWk1+oczYMAAmJmZoXr16vD1lY1Qz6jM7q/otWSl1Bu91jp16qBFixbiWG3btkV4eMYi6FlpA6X8hoSEyC1Z8vmkWtxU00RifNoLj/ERgYh4dgp67k3Efcp2oUBR4NnFiPA6D8kkxgOqX05CktqdWnyID+L8X0DToZy4r1OkMXQK1YeaniWinh6XrrlxMVDVkP3TUdVQ/N7GhgfhydrB8Dm7EealG2doH+W1NzbFc2sjIVY+YEgifJ7h6bpBIovIqFCVDO2jLLEx0dDS0RG36WdsVNpU7NtnjuD3drUQERIsAkSuJcpi5u6z+GXhehSrXAtxMdK1Ny4mGpraslEA9DM2Om17L+zeiNkdasK+UFHoG5uhZO2muHZwF+Z2riO2WzkXkKy9MdHR0P48aoF+pu5s/7t1E9p27ZVmv4UzfkejCsXQRsE2ZVLBl3+iNNoiKlVq/qY1K9F74FC5df2HjsT0+YtRr3FTTBwlv02ZaLSjZorOtapq2u+yv0b3w58jeqF4AWeM+Wt9uuukEB0bCz1t2aSK2poaiIlL+z9/07g+mN+/HTQ11PHnziPJ6+95vYWLjYVYL5WY+ATofB4dSD+jUp2jhMfGwUZPBwNOX8cvZ2/gRXAoQmJi4RMRiaJmsqwiKcUmJkJLRfa3Sz9j0hkJfyYqCFW0DDO1D2OMfY/Xr1+jYsWKoo8yderUr5ZS69atGypUqCAeu3r1apEdQ/0/WkelM1KXVvPz80PDhg1Fn432zezo2Mzur+i1ZKXUG73WPn36oFy5crCwsMD48eMzvG9m25Bd/b2oqGjofT6no6yF1BlXYeHhKFjAFQOGjkD9Zq1w/NRp5HYTl/2NXmOnIjgwAMd2bhbr3r18jqEta+HI9k2o2byd0tsQGhUjSsXRQqWr6Fw0qX+qq6WBqFj5i3mrzt7C4Nqy/mlOioqJgd7nC6O6OtqIjI755vbQiEhxjaJm3/HoM20xghUEE7Lb8ev3Ub7f76J8XHRMLHQ/n4vqaGkgJtV7q2g7lT4u4mSHLtOXovGYebh4X37Am7LQd1HSZOR6unoKqzEQqhpTt46sj5eT1FS/ZNRoamqkuQhNWUTDB8j368YPH4yS1RvArVJtjBrcT5J2Dli+S5SJO3H3KaKob6KlKdZT9YKomLQXzmPj4jFh82FMal9X9BHDIqMREhmNOpNW4JfV/0pSbi4qOga6n69X6NFnKUp+/pEke46dFdurlCme4X2yE33G6bNGC33W6VssqV9NfcDU3xGKti/dcxzD2zWElKKioqCrJ/vfRj8jU33W6tavj7OnT6N8qZJ48ew5KlaunLxt04b1aN+xk7TtjY6Gru7n362uDiJTXb9qWLcWNq1YghGD+mPMpGkIDZOVGJ+z4C+4lqyAxg3qQefz3wb7sUREfP+AzxwNEt25cwdjxozBixcvEBAQgJMnFZeYohNQeuyUKVPw5MkTnD59GidOnBABkEmTJmHOnDnJj0sKNG3btg2PHz+Gp6cnVqxYgUsK6rh+TWb3V/RaslJSj/ahoNe4cePw/PlzaGpq4q+//srQvllpA9UrNDIySl4cHByQFSpqWogL+4SEuGjERwRAVUs2QVYSmqeI5h+izCFVddnJjpquKVS1jaDn3hiRXheR8Hl+BGVTUVVHYlwUEmOjkBAZLAJcqcX6v0DMu9vQLlQPKp8voFF7KbtIw9odiTHStJVoGFog0tdLzNMT/vYhNI1ko/JSUtfWh0PjkTAv0xQ+Z9ZnaB9l0TS2Rvi7h+J2+FtPaJp8GeWWRMvMHg6NRopMIr8b/2ZoH2WxsHfE87uyTKBnt6/BwsEpzWMKlvJA76mLoKGphRsnZTW1dfQNYGJhhcuHdsOxUFHJ2mtm64hX92XtfXn/Bszt0ra3eI2G6DxpIbyf3If30wc4/c8quFWuJcrQxcfF4ck16YKy9k7OuHP9ivgeun/rBuwcneW2b1u/ChOH9cfM8SOxd9vfuHRG9n+gffe+mLNsLXZuWovAgLQ175XFwNAQL549RUR4ON68fglT8y+jT6OjorBl7Sr0bNcCKxb9ib/mzcLzJ49hZWMLt2IlULdxU0RFSpeJoa6uJjqugSFhePfRH6qfv6tScrGzRqlC+dGubhV8DAhOd50U7MxN4PnynehwX374Ao6WpmkeU9jRBmUKOqNZxZLwDfxyEYsyi7rUqQgp2evr4vbnshM3PwbA0UA+RdvBQA+vQ8LRxz0/qtlZYum9Z/jPyxtHXr9Hm0Pn8SwoFBMuKZ4DQRms1DTw+HMm0KPYCFiryTq9KdE5wc3oMJTU1M/wPnmRcucjki2MsYyjMhfUt6K+yqpVq8RcGwo/uwkJePjwIXbt2oXz589j5MiRInjz5s0blC1bFhs2bEh+XFL/Ztq0acifP7+oud6lSxcxgjIzMru/otfyrRJx6b3Wq1evYseOHWKQ4NGjR0UfNyMy24bs6u9ZWVrg2QsvcQHtvucDmJnK/x83NTFBQGAgOrZrgxFDBmH871OQ21GmUcFipVCmSk0E+cnmFjS3scOwmYtQsW4j7FixUOlt0NVUx5IuDcWipa4OPS0NvPgYiLDoGLzxD4Gp3peLeeHRMVh//g5aLdmBxSeuYdb+83jmK19qSCr57axx8a6s+sOF2w+Q397mm9tpoXPPmUO6QV9HBxv3K7/EeLkiLlg9pg/GdW4GB0sz3HvxBrFxcbjo+RRO1vJZZoq225ob462vP/o0rYX+zWpj3IqtSm+zaIuDvcgmoiAvfS/RAOnULly8iJIlS0Jd4owsRSiw6eP7EQGBQQgLj5Br04ePn7Dhnx2o0aw9VqzfjJETp4nMIio3t2rhHKz/33z8Pnv+d5UKzajRLWuKMnFlXR3gYmWGS49lZRLpp4u1mdxjQyKiMGDFbnSvVRbFnGR/3/SY4IhITO/cUASQdl+6p/Q2uzjY4tq9h+J7/vyNe3B1TPu3sOTvXXj74SNG9Gif4X2ym5aGuvis0aKrpQVDPV08efMe4ZHR8PL5CAtjA7nHp95uoKuNFftOoPGYPzBny35M27gHj15lfY7BjKLEgyuXZVMOXL54ES4u+eW2r1+7BiVLlcKS5StgY2uL3Tt3fMmevX4d5cqXh5RcnJ1w6apsXq8LV67BNZ/89RUTY2ORNVS/dg3Y2lgnl3fs0qEtNq9ehkNHj+O5lyzDKM8T2T7xSlwSkNu9ffsWnTt3FgOq9PT0YGxsjPbt2+Ply6z9jnP0v4mHh4c44SeVK1eGj4/PVx9LkzGREiVKiBdNL75mzZri5Dg1Cu5QVo2NjY1Y6tWrl6m2ZXZ/Ra+FashmBR2Hjke6d++OjRs3ZmiSqay0gUatUQcsCY0sy0rHQbdgLQSeXSTGA+h9nl8o/MkxaFq5iTJzIdc2inWB5xZDVdcMxh69oVugJkJubBZBG1UtA6hoZqxGYnag7KDQi/+jS2bQKSLLbIp8eBBarjWhqqmLiNtboapnjrArq8W8RTqFGyDs6hokJsQhMTpMzFkkFTUtXRjkK4WHy3uKL6p8bWSjBF/9OwvOLScgNtQfXjsn09U/MZ+PdfVu6e4jBW0LZ6iqaeLxip7ivkvHuSJr7O3++XBsPhYR75/g3bH/IZE6sxHBsG88UuE+UilWqSYOrV+KCS2rQlNHF+N6DIbf+7c4s+tvtBk6ATdPHsLB9UtEcIXm96nXpR8CfN/jf6N6I/DjB7h7VBOLVMrUb4mlg9vh/M4NMLNzglPR0nh57zrePXuIKq2749zO9bh5ZI/IbkpMTICxpTWs8xXAngWTcPvEfkSFhaBuD8Vz7ShDhx79MLBTSzEvUeVadWFhZY1D/+6AtrYuajVsgrnL1omRZvdv38Cje3dQrFRZzJ8yATcu0zxF4TA2MYW+gTRlHUifIcPQqVl9MRqu58AhYnTTqr8WoGK1GnAvXhLrdvwrHndk/z5oaGrC3tEJw/v2gNfzZ/D98B6DRoyBlIa0a4yi7YeITtWS0f3FuqF/rMLE3u1gaWqM6n3HIywyCh/8A7Fs7ACxXdE6KdAIsQbli6JQjwkiC2rH5IFifccZK7F1Yn/4h4Sh0fiFiItPQEBIGHZMltVQplFld728saig/AmwsrkY6cNAUx11954CdVk31JH9Xx55/hYWVC0NWz0dFDIxxKQr90RW0dAShVDD3hJlrWQdzAmX76J/UVfJ2ltCUx97Ivww2P85tFRUMMPYGX7xsTgQGYAe+rKBAg9iI8RcRGqfB0co2ocxxrJb8+bN4eQkG9RSpEgRUaGB+nGKUOWIpL4IBYgo24Yye6tWrSqCKor6bPPnzxfBD+qvFSpUKFNty+z+il5LVlH/ji5OEaoiQW3JyOjezLYhu/p7dNG3Q5vWKFyyLOLi4rF5/Wqxvkuvvti4eoXILurfuycGDh2J+Pg4tGwm6y9RsGjnv/vE7UePn+CPWdMhlT9G9Yfn9cu4dOwgGnXsgWbd+mHuiL4Yu3C1yBj649cBonRQVEQ4pq3dgbcvnuLPMYOQmJCIIP9P6P972nmXspuaqiqK2cvKwpBf6pRHwwX/gBLEh9atINYtOHoFdd1d4GZrjp2D24p1e289hr6WJuxNDLD96gMsPHYV/mEReOsfjD2/KD8Dqk2dKqjaewz+PnASlmbGWDJmAO489cLhCzcwvlc7hdsp+LJsx0H0nf4XQsMjsXeB4rlrspOxvh6MXb8MYK1duigKdhqFuPh47Jk5QqzrOGUJtk75BQa6Omm2a2tqon1tDwyYv1ZkGvVqXANSoDmH2rRpKz5XXbt2FRf/VqxciSKFC4tKN2T9+g0YMXxY8j4HDh7E5MlTRGk6Kp118YJ0AwNHDOyD0jUbiUy4mRNHi3UTZsxD7y7t4WBrg4PbZdeD1m/ZgYL588Hc1ATOjg7oP3K8+AwUyJ8v3Qov2Sl/ikBQx2qlREbQ0kMXRfDHo6AjLj56ibuvfDCoYSXM3XMaZx+8wMO3su/YnaO7ol3lEmg8fS3O3H8hgrYjmin/OoCNhRmKFsiHfLXbicEAR9cuEOvbDZ+EHYum4d6TFxi/YCUKOTti079H0K99M/Tv0FzhPspE/eYSrl8GsI5s3wi1hs0Uv99hbRuI7TTPV63S7ihTKF+a7VRp4sBc2d/OnnPXRcUMZxvlT0vRrEVL1K9dE7t2bBcZurPmzsPDB544sH8/xowbj4KFCmHOzJk4fvQo/P390bV7D7HfwQP7UTeT15mzQ6+unVCnWRtMm/snSpUoBrfChXD4+Cl8/PQJ3Tu1FxlDu/87iI+f/FCrehXY29pg5vxF2HvgsMg6okw/GvTB8r6wsDBUqVJFXPf/888/xWACigOsXbtWxARo4FV659u5MkiUui7e10YOpH5syvuK9qMT5r///hudOnUSo6tOnTqVqdJvmd0/M68lI6PDqJ4g1Z6lUWLUlozIShsohTkpjfl7aFkVgVnd30QQRU1bdkFX27E8VDVkHR2jCr3ENkJzFhFNc1eY1hmPhKgQqOmZS3JSkNxe+zLQsCxMaUXJbdR0rggVdVlKvH6FPsmPVfmc+SQCQ1TPUdsYKurSjra2rdkL5mWaQU1bD2qasvZaV+smfqrpGsKx8QhAVQ2aBhbiMentIxWnlr8hJtgX6nqmUFWX/b5pbiSiZe4I+4YjoKKmDk0jy+Qyc4r2kYK6hibGr9uLgA/vYGJpAzV1dbHUaNNVbC9UtiIsHZygoaUNc1sH8fi42Bj0nDQfBibmMLaQLkuL6JuY4ddNRxD86QNMrO3F58Ymf2ERMCIlajYSacSaOnowtXUQ3wuUWeRaphJCAz6JeYrotUglf6Ei2HfhFoIC/GFjJ7sgUbFaLRE0StpOHJzzoUa9RjAwMkLHXv3RrF1HMSeQrYMjpFS5ei0x0S6NkrW0ks2f1rxtB+gbGor3skhR2Zw41rb03gPaOjoYOvY3UVbP2tZWBLWk1LNZHTSpWk60zcxINjpreKdmyRMBL/0cBHKysYTB55GoitZJZVqPlujXuAaM9HTEaDEytUdz8ZPWrRrZA5rqanC2NofO5/IP9K+MytDlhD+rlMa7sAiY62hBS032N/tLiS81sqdUKAaf8Ehoq6nBRFvWXrPPZUoWVikNJ0P5rFplUldRwXRjZ3xKiIW5qoYIBGkmqqKRzpc5CuzVteDy+f9cevv8COIpi1qJo1GVeWzGfkTK7vPRJL3u7u64fv06nj17lqm2ZXb/7OzzUZmooUOHiovABw8exMyZM/HqlWxUe3a2Ibv6e2TsqOHo2qkD9PX0YGgoO++YNH6seA1kUL8+aNOyOWJiYmFvJxus2L9vL3Ro1zp5kmUpdRoyGtGfSzMbm8qyRroMHZucMTR05iJoamrBis71NTRFWWRap66uAQtbe+joSvd/PEltt3y4Pqk3ouLiYfX5PKJDeXcY6WrJBZRsjfXFfR1NDbGPm53sop+GmjTFYgz1dXFzy2K88f0EZxtL8XdZwMEW5o1rpbudsuDPrpkDr3e+sDE3EWWwpDarf3sMalUXxvq60P/8/FN7t/nq9vFdmqNHQ1lgxkaiksIeFSrgged9BIeEJM/p1aJ5C+ikeM/GjhmNggW/nJdWqlgR69auFbelvLZC2jRrhJpVKorMFZr/hPTv3hlWFuaiOk7JorK5kiaMHAIdbS2xbtua/+HVG28kJCbAxUnaPh+xNNLHzT9H4F1AMJwsTMR7RhlDSYGkwY0qoX2VEl8eb6wvAhdnZw7Eq4+BsDMzEp8/KayYOhpv3vvC3MRIlG8k04fJ+kcFnOxx4Z9lyY+1NjdNdx8p1SlbFA///gMR0THJn5tOdSrBSF833e1JQSZ7CzPR107qDyqTvr4+zl68jHfe3nBwdBTfVc75XNCpi+x6UPUaNXHb8wF83r+Hnb29eDyp4FERdetlbnqR7ODkYA/PK2dF5h4FWolHudKI/Vw+kzKGGtWrAzMzE9jZyDLgunVsh6YN6kFfXw/5nBwl/35QFhqMTosyj5+b7d27V5QsPnfuHDQ0vnwX0cACGly1e/du9O7dO1PHzPm8VCWh4A6VW7O3txcn/UlZSFLt/z0oU4rKxlGZBZqLafHixaL2dG5HGTgpqel8OYFSN7RWvI+6FlT1cyaKraopf9JP5eSSbxukba+agbTBgNQ0DeXT4bXN7MVPVTUN6Fjlz9A+Ukpd4o5KzBEKWOlYuWRoH6nQiQAFgJJQWbmk+/pGJmJJiQJFDgXdkVPU1DVgavOlvdp6BmIhRuZWYklN18BILDmBJnTUsfvy/WCiYAJhXT19sRA7x7Ql9KRkYGgkliRUTi41kxQlVlxcpZvjSRELE/nfK5WTS1I0xWiur62Tkr2F/OepoL2svepqaiiRP+3IZl1tTehqSxt8S8nuc0cmSb7PAbgkNC+RIlIGiJKoqqjAKkXJOAoCWaa4b5xiPr709mGMsbyESo936NBBjP5t3LhxhgfXZdf+36NYsWKoW7euCAxRxhTNS7tmzRrkdlS+JiWahyglSwv5vp2zo/QXf5PY5XNNd52Wtg5cCsuXjFa0LicY6Woj5dmdrYl8mSZiluL8xNxAVyxSo6BPyvNOCvqkDPyk3p7U73J1kK60uCL2FvLnlQVTtSf1dimDQykZGBiIJYm1tXwfL2WAiND3GC05xcxU/hzfySFtiTMKGqXk7Ci7RpBTNGhwWooS2Ia62mIhtqZGYkmN+iyuNtJfZ3G0lf/9F8on+26loFvJIgUytI/UKCCUFBQidqk+W6m3JzEzku9vKRtlyjo5f6moQBnMtCShAQ6pBzlktXJUdqCBH0kBoqQSc0koc4iWlBzsbMXCfizv379HpUqV5AJEhAbuUIbRu3eZL9eokihF4c900CiDpJFQdJuimelFNL/2WEqfTBq9lPJxqff/2vEz0tastC+99qTnxo0bYu6lAwcOfPU5lNUGKj9A5RbMm8xJzvDIzaLDcqYGc1Y5l5GVDMgrEhLy1mjpX1rnfKcuM4IilT/RZXaqkz/ngo5ZoaeZo9PuZYqzl/JrsmenhCBZ3f684vnfsjI3ecWj43mnTnREQjx6+D9FcHCw5CPE0zuH2XvrBfT0015Qyy7hYaFoUTp/rnjNjOV2SV3NpD7U1/olqR+bso9H22ihfVPeTi3lPlnxtf3Tey1fa096/ve//4lqFRMnypfb+taxsqMNSd+VH1+/SM4Gyu2OvEs7oXxuVuW08ks7ZSfDClWRlySEKp7XLLdK9JBl0uUVaqFZL6MptZgz25GXaBevjLwkzufb2a25SUT5L1mBeYFeVN65nhkSGgoLF7dc0f9JOo/xO7lFzG2ltOcJj4B57c654jUrQnN40mAniiOkDGpSVZwKFSqI5BOabzMzcvQKWsqTWLpNZdYogpty0dbWVvjYlMGalCfy6Z0Yp3d8SnNN7zkzsv+32ke3aTLSrD5v6jYoOlZG2sAYY4wxxhhjUko9yI76JcuWLUvTn6lRo0aax6bs49H6pD5Nytup0T6Kjk9lvBU9Z0b3V9S+1O35nudN/Tqz2gbGGGOMMZYz4uPjcebMGTF9DZUxViaax5POD6ka2eTJk7Fy5UpMnTpV3I+NjUWbNpkPmuaqcnNlypRBVJSsZnCS7KyVmN7xUydTpfecWW1fZp6XOib//fdfpo/FGGOMMcaA+IREsSiLMo/N2M9g4MCB6Nevn9L6M4qOr0h6z5nV9mXmeb8V0FH2e8QYY4wxlpfltjmJwsLC0KBBA7x58wblypXD2bNnUa9ePWzevFkpA3ko+eTChQuYM2cO9u3bJ8rLURnEli1bYvz48d9MRMn1QSJCUbDcfPys7p+Z/b71x6Ps94gxxhhjjDHGlIGCHcrsz3zv8bO6f3a+LmW/R4wxxhhjLPvMmjVLBIju3Lkj5od7/PgxSpYsiS1btqBr165KeaupDB49Ly3ZgXPSGWOMMcbYD4MytROUuOTgdJ6MMcYYY4wxxijTR9lLJmzduhWdO3cWASJSuHBhkUlE6/MKHp7EGGOMMcYYY4wxxhhjjDH2WUhICFLS0tISS0rh4eF49eoV3Nzc5Na7u7uLTKLsQnNWLliwIEOPHTx4MEaMGJGp43OQiDHGGGOM/TDiE2WLMo/PGGOMMcYYYyyHJCTIFmUeH4CDg4Pc6smTJ2PKlCkKA0kmJiZy6ymrKHWQ6Xt4eHjg119/zdBjy5Qpk+njc5CIMcYYY4wxxhhjjDHGGGPss7dv34q5f5KkziIiOjo64mdoaKjcerqftC07lC5dWizKwkEixhhjjDH2w0iaO0iZx2eMMcYYY4wxljMS4+PFoszjEwoQpQwSKWJsbAwzMzNRci6lly9fwtXVFXmFak43gDHGGGOMMcYYY4wxxhhjLK9p2rQpdu7cifjPwaXg4GAcOnRIrM8rOJOIMcYYY4z9MOITEsWizOMzxhhjjDHGGMvJOYmUl0mETM53NHXqVJQvXx6NGzdG7dq1sX37dtja2mLw4MHIKziTiDHGGGOMMSWLjY0VJQhS16rOiGfPnsHLy0sp7WKMMcYYY4wxlnWOjo64e/cuqlevLvp83bp1w+XLl6Gnp4e8gjOJGGOMMcbYDyM3zkm0adMmDBs2DBoaGqL0QK9evfC///0Pampq39x3y5Yt6NKlC/Lnz4/nz59nsdWMMcYYY4wx9oOgLCKlZhLFZ3oXKysrjB8/HnkVZxIxxhhjjDGmJLdu3ULPnj2xZMkSfPz4EXfu3MGOHTswf/78b+5LQaGxY8eiU6dO/PthjDHGGGOMMaYUHCRijDHGGGNMSVavXg13d3eRDUSKFCmCHj16YMWKFV/dLyYmBh06dMDs2bORL18+/v0wxhhjjDHGGFMKDhIxxhhjjLEfRnyi8pfMuH79Ojw8POTWVapUSdSqDggISHc/yiBydXVF165ds/pWMMYYY4wxxtgPJzEhQenLz4bnJGKMMcYYYyyTQkJC5O5raWmJJTV/f3+Ym5vLrUu67+fnB1NT0zT7HDp0CHv27BGTnzLGGGOMMcYYY8rEmUSMMcYYY+yHkZCYqPSFODg4wMjIKHmhsnCKqKmpidJxKUVHR4uf6urqCoNP3bt3F5OefvjwAY8fPxaBptjYWHE7LCxMKe8bY4wxxhhjjOUJCfHKX34ynEnEFP9h6OhCVUMn1787iXnsQ6upwx85ZdJQVVHq8X92R599Ql4SHZd30oP7Xj2GvCTKPxh5yetzb5GXeIbIAgh5QXRi3vmcZbe3b9/C0NAw+b6iLKKkYJKPj4/cOgr+qKqqwsbGJs3jQ0NDYWZmhkWLFoklKeOIgkctWrTA0qVLUbt27Wx/PYyxn0tAojZiE3N/f4/c9PZGXlLLQBd5SbxbTeQlak8vIC95F5m3zpXs4uQHtuRmqroGyEsS42KRl6g7FUJekteyIFTi885nLS+1lWUdX7FmjDHGGGM/jISERLEo8/iEAkQpg0TpqVGjBpYtW4a4uLjkzKGDBw+iQoUK0NGRXaCl7CBvb2/kz58fdnZ2ImMopYkTJ2Lbtm1p1jPGGGOMMcbYTydRydk+iXkrKeFnDLQyxhhjjDGWZwwePBgqKiro2rUrLl26hLlz52L37t2YMmVK8mNOnDiBIkWK4N27dznaVsYYY4wxxhhjPx8OEjHGGGOMsR8GJfrEK3HJbJKSubk5Lly4IMrLDRw4EKdOncJ///2HevXqJT/GwMAAhQoVgoaGhsJjWFhYiCwjxhhjjDHGGPvZJSYkKH352XC5OcYYY4wxxpTI1dUVW7ZsSXc7zTH0tVJyw4YNEwtjjDHGGGOMMZbdOEjEGGOMMcZ+GAmJiWJR5vEZY4wxxhhjjOUQyvRR5pxECT9fJhGXm2OMMcYYY4wxxhhjjDHGGPsJcSYRY4wxxhj7YcQnJopFmcdnjDHGGGOMMZZDKItIqZlE8fjZcCYRY4wxxhhjjDHGGGOMMcbYT4gziRhjjDHG2A8jISFRLMo8PmOMMcYYY4yxnJEYHy8WZR7/Z8OZRIwxxhhjjDHGGGOMMcYYYz8hziRijDHGGGM/DBrzFa/EZJ+fb0wZY4wxxhhjjOUiCQmyRZnH/8lwJhFjjDHGGGOMMcYYY4wxxthPiDOJGGOMMcbYDyMhMVEsyjw+Y4wxxhhjjLEckhAvW5R5/J8MZxIxxhhjjDHGGGOMMcYYY4z9hDiTiDHGGGOM/TDiExPFoszjM8YYY4wxxhjLGYkJ8WJR5vF/NpxJxBhjjDHGGGOMMcYYY4wx9hPK9UGimJgYfPz48bseGxsbC19f3+9qR0BAACIiIhRuy47jf8u3nkOKNjDGGGOM5XYJCYmIV+JCx2eMZa/Q0FAEBwd/12PDwsIQFBT0Xe3w8fFBfLzikaPZcfxv+dZzSNEGxhhjjLHcLjEhQenLzybXB4nu3buHXr16fddjX758iVGjRn1XO6ZNm4ZDhw4p3PbkyRN07twZyvSt55CiDV8TG+yDT4enw/e/CYjyeZBm+6cjM+GzYwg+HpiE6A+PxDq/4/PEuqT1Uorxe4GAE3MQcHo+4oLfp9nud/h3+B2cgMCzixAX4iPWxUcGIujSSvgd/A2h9/ZI2l7/+6dxd2E3PFo7EjHBn9Jsv/1HB9ya1RJPNo5FbLis4+i5bABuzmyOwEeXIDWfCztx58/OePbPZMRHywdXY8OCRLtuzWmN59tnICEuRqyndbRE+L6UvL1b5k/GiEblsW76aCSkujDw/N5NDKhWBL/ULYXNf8j+TgN8fTCnf3sMq19WrEuUsPRQbHQ0Nk8dhpltquDoukVptt86vg8TG5TA9NaVcWbbarEuKiIMa8f0xrSWFbF30VRIKSwoAGt+7Y45Harj2oHtabaf/HspJjcqibkdauDOyQNi3e75E8Q6WqY2K4voiDDJ2uvn/RIrB7fG4u518ezauTTb//1jHGY1L43/9W6El3evinUbRnfHzGalxfJXz/qQ0rXXPqi2eBtqL9kBz/d+abb7hISh88aDcJu5DmP/k72eftuOocD0NVh2/jakdvDpG1Rf9x/abj+O9yHhabY/8QtCm+3HUWLZTqy88VCsq7xmL0os24XOu07CLzxS0vb+G/oJPXweY6rfK0QoSDe/GxWGIb7P0PbdA1yMCEZUQgJm+b9Gl/ePMNv/tbgvpaNxflgY/Rr7Yj8iIdX30qP4MMyOfpm8PI4Px/34UHF7fvQrHIj9lGYfxhiTysaNG7FkyZLveuzBgwexdu3a72pHy5YtRaBIkW3btmH+/PlQpm89hxRt+JpjRw6hUuniaFynBl56vUizfcpv41CycH60bFQP77y9RcCtgIN18jJsUD9J27uoSw3MaVEGbzxvpNkWFxODPbNHYUGHKtg2eRCiw2XnmzcObMXirrWwZUJvRIQEQmpP3/uh2uTVKDFmCU56pn2P+678F7YD5qDchGW49OS1WFdj6prkdWceeknSzri4OPQbMBAuBQph9NhxCh+zbPkKFHYvBnMrG7z38cGzZ89QvVZt5HMtiClTp0FKIWERaDF0MlwadsOizWn78juOnoVJ5Vao1HV48rphc5aJdbTP8u37JW3v0oXz4e5kg6F9eyrc/s77Ldo0qouqpYvhwF7Z6+nfrZPYh5YK7gXTDXgrw23PhyhVuxmKVm+ES9dvKXxMSGgYilSuj6Onz4v77tUawqxwWZSp2wJXbkrfL4mKiUXXeevg2vN3TNtyMM32U3cew6LdKLj3+9J/PnbzoVhHS80xCyRr65pdB+FSryPq9x2NTwFpBwq4NugM4wpNcPb6neR1znU6iHW0PPKSfVdIZfKiVXCq3BQdh/6GqOhouW0HTl2AUfGasCpXH91GTkZMTCz8A4PFuqQlOlp2jUgK9DkZMmggihR0xZjRvypMRujWpTMKuDijf78+SEjRv5s9aybat20DKYWGhaFlt37IV6oy/ly6Kt3H9RwyCqN+ny5u+/kHoF7rznAoVgG/zfxDwtayvCbXB4myQ8GCBbF58+acbsYPLfTuv9B3bwSTKv0Rcmtnmu3mdUbDquV8GJbpgFBP2YVgs5rDxTr6qW1XXNL2hj/YD4PSHaFXpBHCHqQ94TOrOxGm9SZB28kDkS9kF1bD7u6BurEDTOv+Bn33ZpK1lYIo789sQcEuM2BatDp8zqe90F582AYUH7kZBvlK4uM12ftbpM8i2FTtIHkdTQpiBXieQZHeC6FlZodPN4/IbVfXM0LJX7eJNqvr6CPw4QWxntaZFq1BQ8Albe+TW1fh+/YVJm34D5HhYbh38bTcdpeipbDoyA3M3H4cvm9e4YXnbVw5ug+uxctg2pZDYp9H1y9K1t47J/+DroERBi/biWfXL8D31TO57SVqNsbvey7hl+W7cPv4fwgPChDBFyvnAhixdj98Xz+Hz4vHkrX34q71cC1dCX3//BsXdq1HVFio3PbqHfpiwu6L6DlvLU5ulF3waT50ilg3cuMRFChbGVq6+pK199SGxajctjfa/b4Yx1bPS7O96bBpGLXtPJqPmomT62Qdg84zVuLX7efRfd4GFPKoCSlNO3IZf7asgd8bVMT0o5fTbJ944ALcrM1wcUQnTGtUWaz7q3UtLGxZEzHx0n7WYuMT8NeV+1jbojpaFnHG0mtpBxSMOHIJLQo740LvFuhVqrBYd7x7E1zo3RzVnGyw6a7837syBcTH4lREEOZZuMBJQxuHwwPktlMAaH7AW3QxtMImm8Lw0DHE3egwqEMFS6xcxT4nIqS7uPQ2IQqfEmLRW9MOcUjE0wT5AH0hVT2M0nQSSz5VHeRX1YG7qr64P0DTHgGJsfBJlO/E5VXKzCJKWhhjuU/79u2/e2AgSx8Nipo9dTLW/P0Pevbtjz/nzEpz0frKpYs4eOIMevcfhEV/zIGamhruPHoulrG/TUKDRk0kfYsHrTmEMo3aIUHBOc/bh7cQGx2Fvkv3wKlYWdw4uBWRocG4sf8fdJu3EQ7upXFtn/TXEKbuPoVRTapg46A2+G3b8TTbl/RsiqeLRmJJzyaYsuuUWHd4fHexbm6n+pi7L+0gJ2U4cPCQuHh6/uxpPHjwEFeuygZPJbl1+zYWLFqMndu34tWLZ7Cxtsb6jZvQonlzXLpwDg8ePsS9+/chlVW7DqJogXw4vfYPbPrvBHw+yZ/XtapdBW+Pb4Gqikryunkj+4p1h5fPwoJNuyUdGNh38FAcPnsJUVFRCrf/788/0Kp9J6zbuhML58wUlWX+WrUO1x48w77jZ1Czbn3x+ZPKuOl/4I/JY7Fq/gyMnjpX4WMWrFiHmlU8EBcfJ+5fP7oHr2+exW/DB2He0jWQ2vazN2Cir4czf4zEyduP8eiN/ACB6sUK4uXGWTAz1EteF5eQgPHtG4r1h6YPkaSdwaFhWLJlD46v/RM1y5fCXwqCnHf3rMHADs0RlyIwWMDJHt6nd4qlkLMDpPLw2UucunQdF3ethZGBPrbtl/8ea1i9It5dPogHR7cjJCwcZ6/dEp+t+lU9xHpatLQ0JWvv0SNHRNWok2fO4cXzZ7h0Uf7azpHDh2FgYIDLV68jPCwc58/LvmO9XrxAbEwMoqKkHcS4dvN2FC6QH2f378Q/u/fB+33agS2nL1yCpblZcoBuw9adqFbJA9dO7Mfp85fwxjvtQPm8KDEhEYnxCcpbEn6+Pl+eChJRybesPPZ7SrEFBgZm6WSAyiCkV54uoyXzMvN6s7MNWREfEQAdxzLQNHWCqrY+EmLlT2ZU1DWhqqENVXUtaFoWklsX+foadFwqQSoiaKKiCg1TJ2hZuyExNu2Xuoq6lmibipo6NMxdxbrYoDfQK1QXqpq6Yr1UYoI+Qte2AHQsHGFRphEifNOODlPT1Ia6th409E2hqqmVvE7KdiahTCDjQh7QMraERan6iPggPwJORUUFalo6UNcxgLquoXifRXu1dKCiKv1X0punD+FRvzlMLK1RqWErvHkqf+FaVVUV2rp60Dc2hZ6hETS1dVCici1cPbYPo1tUQVR4GF4/8ZSsve+fP0KZ+i1hbGmD4jUb4X2qgI+aujq0PrdXx8AIqmrq0NYzgI6+AfSMTMQ2WqRsb7nGbWFm5wiXkh745C2fKaauoQktHT0YmlpCx9BItk5Ttu7BheMoXqMRpBT0wRtFazSCjasbDMwsxcUCufZ+bpuBuVVy8EpDS1us8zx9ECXqtpSsrUklr8o5WqNGAQeEKhhxdfOtL0bWLAMTXW1oqcs6itoa6tDSkP67wScsAoXMjeFqaoR2RfOLrKGU/CKiEBkbjy4lCsBASwMaarLvA10NdXHfQk9H3JbKq9golNc2gKW6JurqmsArRv5/xcvYSOTT0BbBIR1VNaipqKCwpi6846JF9tGb2Gi8SLWPMn1IjEZRNX0YqqijuKqBuJ8SXfTQVFFFOOKhBzVoqKgmr6P7uipqMFeRrkPGGGOK0AjdjPaBUj/2e0qxUZ8vK/z9/UVfM7OSSubRa/je8nFZbUOmn8fPDw5OTnBzL4o27TumySSytrGFtY0NKpcpgX93bcfjR7KMYD19fbGcPnEctes1gJQ0tXWhqq743MGmgDvCAv3wV7faePfkPj48fwi/t17IV6oiTGwcUKpBW/hKOLAqiXdACJqXK4KSzjYwN9BFaKT8/3NtTXXoa2vCxthQ/CQ6mhritp62JqoWdpaknVTFpVPHDrCztUX7dm1x7558wOf0mbPo06snihUtCn19fdEHbNa0Cf5a8j8UK1FKPObOnbuQyt2nXujRvB6cbK3QqGp5PHjxSm67uroa9HV15NZp0fuqqwMzIwNULOkmXoNUNDU1oaOjm+72J48eoH2XbihQqDAKuxfFh/fvoKWtLT5rB/7djZbtOkBKIWFhqFOtMiqVKy3up8y2II+fe8HMxBj2ttbJ63R1dKCvpwddHW3UqFQeUrv38h261CoPe3MTtK5SCvdfvpPbrqamCn0dLahA/ve+YM9xlBw0AysPyTKilO3JK29UL1cS+R1s0atVQ9x98jzNY/R0daDxua/3Zb+3KNSoK7qOnSUy6aRy99FTtGlYG/Y2lujWurG4nxIFL/X1dGFuaiyCSCWKFBDrj1+4ioK1WmP49AUiU1Eq9+/fQ7v2HWBra4s2bduJ+ykZGRnC0MgQZubmIlhk9PmaxapVKzB4yC+Q2l3Ph+jeoQ0c7W3RtH4deD56kiYzau/BY+jYpsWX12BoAFNjI5ibmojb9JljLM8Gifz8/ODh4SEygkqWLInIyMhMPTYrpdhoxEadOnWQL18+FCpUSBwjo0Ge5s2bw97eHlZWVujZs2emgkx0stW0aVOULl0aBQoUQKlSpfDpU9ryYtnVhujoaISEhMgtWfPlH6eKihpFYtK2K+A1Il9fh0HRLyPIEuNiEBfiCw0jW0gqAyd40R8eIj4iENoOZcT7p6pliOBr6+F/dCoiXpyFVBITE5JPSFVUVMX91GJC/HFzRjP4nN8K8+K1kaOofUntVVVT2N7Q1/dxY3oThL66D0MXWSchp1Cd0aTglKq6WpoTWnLhwC70r1JI/B3YOrvCLn9B/LHvIlacfQSnwkXl/v6V397EL+1VUxMjHFI7tn4xJjYsCccixaFjYIjCHtVx98xhTKjrLoIZxlZ20rU3MVH83Sa3V8H7u33Wr5jWohzKNmgtt/7B+eMoXLFmjn03UIBN0d/DmuEdsKRnfVRs/aX8Q2xMND6+egZrF1kQXAqJoPc2xXevgsdY6Oui3/bjKD1vE1Zc+FJ+ICdQKbOk0Zn0M1FBppGNvi6a/3NUlKT77/Gr5OCR+/92YPGV+yIDSSrUvqROIQWAUv8lxCYmigBL3w9PMPDDU9yLCoORmjr+siqAPXZFUUXHSMJvhqT2yqiqIE17k9yKD0VxtS/ZefGJiTgW74/KasbQ+vxZzes4k4ixvOnWrVui/+Ps7Izu3btn+rFZKcXm5eUl+o20VKtWTQRvMuLNmzeir0n9RDMzM/z111+Zel4qmUftpr6mo6Mj2rZtq/CcIzvakF39PWofDZ5KD1342/DPDjx/9xETp86Qu8jnee8uXFxdxYXv3IIGUfVevB3j/7uDkvVbiXNA0S+QO29VfkWG4Igo2PSfLZbw6Bi5cwc1VVWF2atUXq7s+KUY2rBi8rrbr95j99UHGNu8GqSQ8u+BAiypS5vRCPuXr14hX/4CqF23PrzfvYNHhQp4+fwpPrx7CycnJ0mDLjS4SpVOkJLbm7HPm69/ICb8tR4Lfu2P3ERcn0h6/9W+vP/0e7ly8QLKV5RuEC5JGUih9zn19ad1/+zEwB6d0ux34eoNnL18Db/07iZJO5tOWipKxe29dEe8VypJfxNqij9rqdUv7YbHa6bhxOxh2HPxNl77+kv0Wcv83+6j/Rtwf986lCicH8u37YNU5Nqrpri99Jjx85aiT4cWsDQzFQGjt5cO4M6hLeJvZ8/R0znTXnX1NN9lFTwq4s6t2zA3MYKf3ycULVYMJ44fQ+3adURgNkf61Mnfvepp3l/KNOrfQ/76d6smDbFk9XoYOrujgEs+mJuZ4keg1CyieNnys8kTVwPoBHjfvn0iWOLi4oITJ05ky2O/ZtOmTTA2NhZBp+vXr+Pt27cZPuGnLzXaj7KXnj59KupjZwbNoXTgwAExKqx27dpYuHBhpvbPTBtmz54NIyOj5MXBIWtpqKraBoj++ExkFFFghbJtUop6fx8RLy7AsHQ7uWyRyDc3oG1fElISgYu4GMSH+yM26K3IaEot8tUVxAW/E5lDYh8VFaioaUCvSEMYVRqASC/pyotpGVsh/P0zxEWEIPDRBeiYO6Z5jIaBKUqO3ga72j3w5mj6dUmloG3hhODnN5AQG4OAh+ehY+GU5jH6Dm4oNXo7jApWwIeLacsTSokCPnfOnUBcXCxunj4CO5eCaR5TsUELUXJO18AQV4/9l7w+2P8Tzv+3HQVKlJWsvVb5XPHgwgnExcbg4cVT4n5qtboMxMTdF+Dj9RSvH9zGqb+XoWzD1vht53nx+Xt48aR07XWm9h4X8wp53bkKM7u0fw+tfp2JsVvP4Mq+LQgNkM2r8+bRXbEvZRpJycDUAq/uXUOIv6/IKtI1NE7zmO5zN2Dgin04tFRW45c8On8MBT1qSNpWunAQGx+PVwHB8PTxg2aq0WNEV1Mdo2qWxY6ezbDhatryblKiANCjT4EIjIzG8RfeyGdiILfdUk8bQdHRWNSwIhY3rIQ1t2Sjd810tHC1X0uMq1oSM85KV6/cQV0Lt6JCEZ2YgAuRwXDU0JLfrqGFwIRYzDTPh86GVjgQ/qWTSPvsC/ODm5Z0WXsWKpqixBwFfWi+IUsFWUHUqXiTEAlHFe0v7Yz7iJKqBrBRlX99jDEmNcqsefjwoaiqcP78+eTqCt/72K+ZMWOGGFBHfcc1a9bg1Sv57IL0TJ06VcxfRP0tasesWbPw4cOHTD039fUeP34snpsyoqj/lxkZbUN29ffMLSzwyssLH319cfbUSdja2qV7we1/C/9EuQpfAhibN65H2w5pLxDnBnGx0biyax0c3cvAzN5ZnAfGRIbj0fmjMHdKe56d3Yx0tfFs8Six6Glpiuyhy0/f4F1ACN4HhsBYL+1FyEPjuuPCtH4YuemwuH/s3jP8fe4OZnesJ84PpVCkSBHs339AZLH9t/8g3IoUkdvu5ibLvLl86QI8PCpgz7//Jm97/fo1du3ajQrlpcseccvviH2nLyMyKhpHL95A4XwOGcrgGDFvBeaO6C0uEOcmLq4FcPzwIQT4++PB/XuwsbMX68+cOIaKVapKGoAjVB7s3sMneOb1CtExMXKl7t6+98HS9Vtg4VYBsxYvR8f+I/Dm3XvsOnAEB46dxqwJoyRr747f+opScU0qFEMRRxvsv3IPMbFxOHjtPoo4fslySk9SdpGFsYH4vMbEKT+QXMDJDudu3ENoeAR2Hz8HN9eMDZqj7CJTI0NYmBojWoJs0yRuBVxw8NRFMdcQBXvcCuST2x4ZFYU+42aged3qqFruy7VAyi4yMzaCsaEBomOkay99lx06eFB8lx3Yv1/cT2nViuWoUq0aHj19hnwu+fHP5s3YtHEjenTrikKuLrhw/jwm/z5Rsva6FSqAfYeOITIyCkdOnkGRQvL/p1b/vRWVG7ZC9aZtsf6fnVjz91b8Pns+xgwdCK9bF/Ce/odfvCJZe1nekieCRJRNQxkx9I+DTja+VoIgM4/9GjpxqVWrlojM0sl0xYpfTnK/hjoWFNjR0NCArq6uGJGW0c5GkqJFi4pUR1K3bl3RlszITBvGjx8vOllJS0aDYakZFGuGwIur8PHgFBgUayrWBd/chphPsjIEwTe2IdLrEj7sGopPh79cWI14eQk6TtKnFusVro/AswsRfGUt9ArLyh7Q7YRo2ejBsLs7EfHsFD4dGI/g6xs/79MAwVfWiP207aXLflHV0IJluSa4u7AbvE+uh03V9mL9w1Wy1NbooI+4NasF7vzRAW+PrIJZMdmF6qdbJuHdyY3w2jMPr/Yvlqy92qY2Igh0a05rBD2+BMuyjcQovIdrZJOAhry8i1uzW+HO/I7wu30ExoVln637/+sLv1tH8WjtCPhc2CFZe4uUqyyynQZWd0ew30eUql4XH157YdUkWXvP7duGwbWKYUSjcvB6cAeFSlfAR+83GFCtCCa0q41KjVqL+YmkUqpuc7zyvIlJjUrB1NYBdgXc8eTqOZE9RI6sWYApTcthVrvqiI+NhVW+AihYvhrO/LMKM9tVw7unnrAvRNlP0qjcugcu7tqAmW0qo0StxqLk3eV9W3DzqKyW8tbpIzCjRQX82a0eLJ1coWcsG9Vy/cB2lGnQClKr3mUw9swZg6V9mqBap4Hif8nRlXPw+r5swuNVv7TFvLaVsGpIGxSpLAsikzvH/kWxmtLW2Se/1iqHxiv2oPPGgxhTW/Zd2vufI/gQEi5u07oeWw6j/rKdaF5MdgI59fAlDNh2DIvP3ETbdV+CnspG5e66lywosoTmnL+DgeXcxfqWW4+Kn3RRo2+ZImi17Ri67TmN9u754RsWgeLLdsFj1b+YduYWmhZKG2RUFiozV1xbHx3fPcSFiGA00jMT60f4yko8mKhpoLy2IQb7PsOyoHeorWsi1rd/9wBdfR7BVl0TNXXTBhmVxVlFW5zUzYl5iaDEOBRW1UNgYix2x34pt0tBJBdV3eSOuGdCGB4mhGNd7HvMjn6J56nmMcqraNCXcrOJcvoVMvZjqlGjBrS0tKCtrS0G+32tBFxmHvs11M+qX7++uE3ZRHSsjKC+VYMGsj4EVW+gi0re3t6Zeu6qVatCR0dHvA7qr2Wlz5eRNmRXf49GLg8ePhLVypfGr8OGYPiYcWJ9s/q1k0det2pcH0Wc7eDt/RbDR48R6yLCw/HogSeKl5S+esD6ER1xZc9GbP29H05tWCTWrR3aLnn7oi418Gf7ytA1MhXl5fSMzVC4Uh0s6FAVNw9uR/nmXSVpJ5WKSyodN6FlDfRYthsevy3HuObVxbqxW47i6nPZ763a5NXIP3Q+ak5bi+ZlZfM3jtlyFP9cuAv7gXPFdim0aN5MZAqZW9mIAEG1alXFvES/jhkrtjdq2ABBQcEoVMQd+w8cQOOGDXHj5k3x+MrVamDUqBEoWFBWZkoK/do0xvYjZ2BVsz1qli8pys79c+g0/rdVlmFx5e4jmFRuhRsPn4mfb3w+YvaarThw7ipcGnYX64JDZefXUti7czuqlyuBU8eOoLiLLKD11/y5OHlUFhgcNHwUZk4aj2qli6Fnv4Hie4Rs37wJrdp3hNSmjh6KRp16o2qzDpgyeqhY13vEeDx69gL2Ntbw9byMt7fP4deBfbB+8Vw42Npg6IRpWPn3NpgWKot67XpI0k4dLU0R5KEMl441yuHKo5ew6jAazlbmKJnfAcduPsT0f2QDrJ+9+yiyjm4+fy1+3nnxVmyj205dJ6CEiz0K2Fkqvc0UOGlXvwbsa7bD0i17MbSLrI9csdPg5MfU7DEC89fvQItffsf4havxyOs1jCs0gUG5Rti8/zj6t5Ndo5NCKfdCcLa3gUnJ2rh6+wE6N28A/8BgNO4lu76y/+R5/PPfUdTv/guMitfEjoPHceDU/9m7C/CmrjYO4P+6C+7u7u4yGLYNxpjClLkAG3Nh+8YYMB9sjLExYQJswxnu7i6lpS2l7m6R73lPllKnQHObtP/f8+Rpcm9ucpImN/fc97zv2aOu+7QfjAsBQRg/YrBm7R05ajTCQkNRt5bpPPKAgYNw7OhRTH95mlrff+BA/PHbErRu0RybNm5Az169sHDRD7jofxnHT51Gr9598NY772rW3scfvBcr1m1AjVad0ad7FzRuUB/LVq7FFwt+UOt3rVmOkNMHsHH5r5g4YZwqTTfytsF488M5aNKlH0JCw9CmZcGB0bZIsn8tfalo7Ixazr53E44cOYIZM2bkjKx6++230bRpUzzyyCMlvm/Xrl0xZcqUG8oqkmwkGZ21cOFCdTAt5ermzZuH8ePHF7jvmTNnch5/zZo1eOutt9RINCl1J2UEli1bhu4lHCUjr0E6CUuWLFElFF577TV18C8Bq+JeQ2m1QcoPSFCs5j1fwd4pb03ekpdGM8UejfpsydNXtw26TFlpupOdnZqbSN1Hl1VoJk9JZacm3WIZrGsjVkztdVTLjNJeMzt7lUVU2Gu8UQ07ty2VsmjCkJ2pAkjyOgxZGSpDyj5X1oVk8phLJNg5OMD+Ft7nUm2vQa/aJvMl2Ttee1/1WRk5nxE7RydV6utGPTvS1FG61ZIJ8p7qsjLVXDN6nU6VEnNydlHz/ZjXZ6anqfmJiiu7cT3RqVml0l6DXg+DQa+ybiTDSIJD0nYpk5GbXpcNh1zv+Y26lcnac7dX2iH7AQcHR/XeGvV6OLm65fk+ykTC8hpuRabOUCrtlffUXr5fDg6qXfLZds5XI/xW2zv5oOnERWnsyzKydSogk3tZ7hIbMuJNSrsJWeZ2E/P8ZMTmnavpZsvOqcfS6eCaa74AWS9r1b7YaERatg6O9vY5cyrdjPN/nSm19kr2Te6ybJK5I+XozDIMBjjJ5/sWRkQeik4tlc+DXNfDCMf/2ittlTXm1yOvTZer8J8j7PK81pKQ92N2VpA6VvL29kZZMh/DvLfmGFw9rpXUK20yF937YzpbxWsmKi+knyVz80j/TUjZb1nWsmXLEt93z549KnAi2UElJQEU2UYeS6pHTJ48WVV1kKBLfuZMI3l86WtJFpD0F8+ePYsXX3wR/v7+ar6Ckr7eL774Aj/99JMacS99W6lkcfr06WJfQ2m0wbyvvBAUCq+b2IflLzsn/U0Jdpmvy+vJXVZO7i8jtM0nsW/GgkM3FoAzMx23mfpDcgzs4OSc55gtKyNNHT/n73fk78fcqOkRf+BW5T52y8zWwUn6c/Z2SMvMViWHJYvBTJbJb7qQ33F3lxs73nca/0rp9EkK+V9LANGcVSLrZc5kDw+PW8occfDbUyrt1en06ljJyclUYjot49o5AA83V5Vhkp2rbGL+OYtKKrR+3xveRso1ZmaY5nmW98rdw0NNKyDvZe4snfzfx4z0dLj+9328WXUy8s7NU1LmU4vm/21mZpZ6b3O3Tz4fclteQ2paWs42DvYOcLuJeVJ0x26uclCR/VS9AXrp6zk5qralZlzrr8v3Sqc3qL6UXL+Z8wBODVuVSjuFZMW5ubrkXJd2C2m7BMHk8yzXnZ1uvv9v998cvKXT3gy4ubqqz3ZGrrl0XV2c1WdGXoNcl4H6NyulRjuL7cvktgzEL2yqEhmwcjPcUyNKpb3ynpr2ZdfadyOvoSSSkpNRtVlHq+j/mI9jrnz9CrzdLFcNIyk9E/Vf+MQqXrNWtJ+5+gbJB1oyg8wqVaqkJj68kfvKQWrNmtdPHc1N5vSRg/THH39cZfZMnDhRHcwUJvfjy3xCUt7tmWeeUV/Yd999t8QBIjMJEm3btk0FvCQj6Omnn1adgeJeQ2m34WblDp7kDqyYg0IF7q9x4CLPc+c7MM3dXrsi2mvarmwS8PJ3VCTgYmqPHRxcCh4I5g4YWVV7ZYL3Qtrr4Fy2k+flPoCR99TceZTAkDk4lHu9q7t2ZaSu114JXpgDQtLRLapE260EiEqzvbnbIcG3wtxqgKg025v7/SyqXWXZ3vz7MtdCgj7mkwxCytIVVppOK/mDELkDRPnXy2vzcC67z23+9oj88/bkDwa5alTmpSh55qmys1OBn6LaKq/NWdOZk7Rhzvix5OMTUenK3/mWfk1RJzKKuq8ER6T/dyMk0DJ16lTcf//9GD16NEaNGlXkCarcj//mm2+qANN9992H6tWrY/ny5SUOEJnJc3366acqKCUBnh49eqg+X3GvobTbcDPynxg1B4jyX899/1sJEN2Kwo7Pci9zds076MfsVgJEpSX3sZtLrmO7wgJANxoUslifpJD/de6Ahqwv6lxOmRzjO+ZrW74gkIuzk7qUBdkPOeZ7rwqb0yv/9/FWA0Sl2SeRDLP8cu/XPdwL//6VaT/VQQJY9jmvR7KOcnO2t1eBl7JQYN/7X4Ao//VbDWparr2uOZ9tz0J+Z6XcXFm63r6sqGOSmw0Qle6+rLD+f8lfg62y9LxBxgpYPsLqM4kKIyMowsLC8iyTHbhMfnij2xUWSS3qsbR4XqlJnTsbqrTaUFK3mkmktVvJJCoLt5JJRJbNJCoLt5JJVBZs7cTorWQSae1WMonKwq1kEpWFW8kkKgu3kkmkNWvMJHp71VGLZxJ9eGcXq3jNROVZcnKymrcn/8mY6w3+K2w7GWWb/yRKUY+lxfP+9ddfebKhSqsNJXGrmURl4WYzicpKaWQSaelWMonKwq1kEpWFm8kkKks3m0lUFkojk0hLt5JJVBZuJZOoLNxKJlFZuJVMIq1ZYyZR8BfTLJ5J1GDKZ1bxmrVi9ZlEhbl48aLKlslNDr4lvf5Gt8ud/ny9x9LieVesWJEnG6q02kBERERUETCTiKh8kH6RVETIrVu3bipj5ka3k0ybqKioEj2WFs+bvz9XWm0gIiIiqgiYSVT6bDJI1K5dO5WKr9V2Wj/vDz/8UOptICIiIiIishWTJk1SF622K+vntdRjERERERGVyyAREREREVFRk31bsjymPD4RERERERGVDaNeD4Neb9HHr2jKfkZGIiIiIiIiIiIiIiIi0hwziYiIiIio3NAbLZtJJI9PREREREREZcNoNMBoMFj08SsaZhIRERERERERERERERFVQAwSERERERERERERERERVUAsN0dERERE5YaUmrNouTkLPjYREREREREVz6g3qIulGC342NaKmUREREREREREREREREQVEDOJiIiIiKjcYCYRERERERFR+cVMotLHTCIiIiIiIiIiIiIiIqIKiJlERERERFRu6AxGOFhw3iB5fCIiIiIiIiobRoMRRoMF5yQyVLw+HzOJiIiIiIiIiIiIiIiIKiBmEhERERFRucE5iYiIiIiIiMovg96gLpZ8/IqGmUREREREREREREREREQVEDOJiIiIiKjcMBiMKpvIko9PREREREREZcOoN6iLJR+/omEmERERERERERERERERUQXETCIqlEGXBdg5WP27o89Khy0x6Gxr9LGdvR1sCQd3W5YlR+ZbgqMNfX6Nej1siT5LB1uiy7at9zfLhr5rWUbra6veaFQXSz4+EZGt8/z3K3i5ucAWvDnhddgS/Uon2JL0JR/BlrjVbwhbUj1oEWyJsf+9Zd2EcsvOqzJsSdrOlbAlWWPbwpa4OnvAVhicrC+rhplEpY+ZRERERERERERERERERBUQM4mIiIiIqFxlPVoy89HWsiqJiIiIiIjKE6PRAKPBgnMSGa0ve8rSmElERERERERERERERERUATGTiIiIiIjKDWYSERERERERlV+ck6j0MZOIiIiIiIiIiIiIiIioAmImERERERGVG8wkIiIiIiIiKr+YSVT6mElERERERERERERERERUATGTiIiIiIjKDb3RAL3BYNHHJyIiIiIiorJhMBjUxZKPX9Ewk4iIiIiIiIiIiIiIiKgCYiYREREREZUbBoNRzUtkyccnIiIiIiKissE5iUofM4mIiIiIiIiIiIiIiIgqIGYSEREREVG5IVlE9hbM9rFklhIRERERERGVJJNIb9HHr2iYSURERERERERERERERFQBMZOIiIiIiMoNnQGws2C2jzw+ERERERERlQ2jwaAulnz8ioaZRERERERERERERERERBUQM4mIiIiIqNzgnERERERERETlPJPIgvMGGZlJZH2Sk5Nx8uTJW7pvamoqjh8/fkvt8PPzQ1RUVKHrSuPxr+d6z2HJ115SRqN8QbMLXWfITIE+IwlGQ8FJxQxZaTBkpkJrRoNOtbmo9sol93pDdjoMmcnqUtTrtCSDLgtGY+Hlc7JTE9QlN11aIrJT4tR2ZcGQnVnocnlPpV269KQ8y2WZXAr7jGghMyO90OU6XTYSY6KQlpK3vfK/yEjT/nNrlp2ZUeTy5NhoZKWnFViXkZYCa2tvZnoqkuOikZ117fOSkZKslpkvBo1/nOX5dFmFf2+kbSnxMdDrru0D0pISctoq67SWrTdAd52Do/i0DCRlmF5TYnomopLTkJal/X5MZOr0Re7LhKyLTctAWrZO3ZbrcemZxW5jSZlF/E6Y6Y1GJOh1yMp1v/Qy2o+J7CLaqzMakGLU5VzkftJ2uZ5xnddIRGRpV69eRVBQ0C3dNywsDAEBAbfUjsOHDyMzs/Bj2NJ4/Ou53nNY8rXfqPQijiNkeWRCClIzs0q8jSWlpxd+jK/X6xEREZFzSUoyHevL9djYWJQlORbN/O84KL/Y5DREJqZAl2+ybrldFu+vyNLpoS/ieD0uJR3RSakw5CsDm5GtQ7YFJxwvTkZWdpHHlTGJKYiIS0Rmdnaxy7SUXsh3ySwiPkld8n8ektML73tZkk6nQ3YR71FiUhIiIqPUJS0tvchl1vT+ZmXr1HsbnZics0w+A+b3PDZJ23MB6RlF/08jomMQExef538hy2Lj854j0pp874valyWmZSAqMQXZumuf3eT0TPX7Ieu0lpZW8PyJmewv5Nxq7vc3MjISUVGRKCtF/baZSXsjo6JvaBsiqy83d/HiRbz11lu3dN/o6Gj88ccft9SOb775Brt27Sp0XWBgIKZPnw5Lut5zWPK1l0RWzGVErXpdXdIC9xdYH79vEWI3fYSoVa8iPfiQWqZPS0DM5tmIXvcuEo/8Bi1lhp9GwtZZ6pIdU7BDlbR/ARL3fI2EbbORHXtZLUs5/gcS98xXl6yoC5q2N2Lfcpz89F6c/vIhpEcV7BheWDwV5xY8hdNfP4LMuDC1zP/P93Bm/hNI9DsIrQWv/Qon5k7AmW8mIzvl2sGK0KUl4eyCZ3H668dw9rtnocswBS7Usi8fKfT1WZJBr8cXUx7GlGGdMPupCQWCGVcunsWMh0bi1Tv64POXJkGv0yHssh+mj+mFl0d1z1mmlfTkJHz55F2YcWd3/P6/qQXWn927FZ8/cQc+GNcbf3/6jlomQaM5D92GmeP7YtH0x9Rr1kpceAg+eWgwZt3dC5t//LzA+oOrfsP8p+7CrHE9sevPhWrZ5sWfq2Vy+eKR4dAVEWCyhJBzJ/DxPb0xa3xPHPl3eYH1GxfNxbwn78RH43ri1PZ1atlfH7+a096FL90PLa07F4j2c35Bh7m/Yqf/1QLrL0TGoc+XS9Hv62X4ZPsRteydf/dh0Pzl+H7/aWjtxxN+6Ll4Nfr9vBYXYxMLrN9/NQq9f1qDUUs3Ydk50773gZXbMfLPjRi8ZD0CE6510rSwMCkcD0dfxPMx/ogrZHDA9vQEPBR9EVPiAnA8MxVJBh2mxAbg0ZhLmBp7Wd3WinRelusi8JXuCpbowgoEiy4b07FYF6Yu3+muIsiYgWhkqdvf6kLwpy5cBY3KSyaRpS9EVLpWrlyJJUuW3NJ9ZcDc7t27b6kdL7zwguo/FWb9+vVYvHgxLOl6z2HJ115SctK/77sL0ej5T/DkwpUF1q85egF93v0OzV76DC//+m+JtrGUt995B/Xq10frNm0QfOVKnnUSbOvZs6e6dOjYEfPnz1cBQvPtjp06qYCR1g74XUHDp2ej4dMfY8nOYwXWP/jFH+j1+nzUf3IW/txzQi3bcPwiGj09G42e/hhfr9uraXtXHDmPZq98pS47LxTsxz2+aCX6/u8HtHrtaxz471h18a7jaD79KzR9+SusOqptv/rT5RtR7/5X0GTS6zgXbOo353b3+/PRfvJ7WHfwVLHLtPLSd3+hwaPvoutLsxFZyHFwr5c/RdtnZuLsFdNnNSohGd2mzEHTx99Hn1c+Q3xK0Se+S9Pm7TtRp0V71G7eDivXri+w/rlpr6HbwNvQfdAw/PrnsiKXaSkhJQ29p85BvYfewCOf/FRg/f7zl9HzpY/R86XZOcu2nriIDk//Ty2fNPdHzdr6zifzUbPLEDTtNxqBIaEFAt5dRt6PNkPvRuvBYxEWGY2A4KtqWesh49By0J0Ivlrws25ph/yvoslLn6LxC5/gl50FB66/t2wrer/zHRq98An+OnBGLfvy333o+c4CvPbbRs3amZGRgTtHDkfrJg1x//hx6v3M7fjRo2jXoqm6PD7pIdXnunD+PAb37YWu7dpAaxL46dSrP+o2b4MXXn61wHoJ1D702JOo37Ideg8aVqJtbJbelElkqQssmKVkraw+SFQaGjZsiDlz5qAi0uq1p5xbj8oDnke1kTOQemFzgfVVBk1B9Ts+RpWhryHtsunANeXMGrhUb4Hqd81FpT5PQksZATvh028KvHtMRrr/tgLrfQe+gkpD3oBH+7uRGXbi2vIBU+E7+DW41GqnWVsN2RmIPbUV7af+jvojX0Tk/r8K3Kfd84vR4eWlqN3/QcSc2KSWtXzsC9Tq/wC0lhEbivSYK+j0+j+o1nkEoo+ZOodmTh6+6PjKn+j06l/wbdYDCedNnwdZVqXDEM3be/bgLnj5VsG3uy6gbtMWOLbT9P6ZNW7TEZ9vOIKvt56Ch7cvAk4fw8k929Br5DjMWPIvfKpUx6m9BT9DlnJs80q06TsUH/57UgWMQi7kPdHfcfAovLtiP95fcxgRgX5Iio3C6V0b0X30BHyw7jic3dwQFqBdZ2z/Pz/jtsem4Y1/DuDiwR1ITYjLs77/fU/i9eV78dryPTi+aYVaNuaFd9Wy5xasQOt+w+Hs5q5Ze7cv+QaTZi7EtF+3YM+yHwqMMrxzygeqbS/9uB57/zKdxJn00UK17IH356HDkDugpa92Hce25+7B6ifuxOeFnEh49999eLJXO5x5bRI+GNHbtM24Qfhy7ECURQbR3xcCsfvh0fhocDd8fyzv51De6ze3H8acId1x4NE78EiH5mr5xgdGqNtTe7TFX+cDNWtvhC4LgboM/FatJUa4V8bG9LwB7xSDHouTI/Fx5Yb4qVoL9HD1woWsdNRzcMGXVRqjp4sXtqRrN3pPgj4usMfLTg1R084FF415Two0t/fAC0711aW+nSsa27mp+8ntKY714QFHXDVqP2qPiKi0jBgxAo888kiFfEO1fO2/7z2Jsd1bI2zB6yqj5VRw3kDKhF7t4P/VywiaNx3HAkMRk5R63W0sQQI827dvR+jVq3jj9dfx7bff5lnfpEkTFSiSy70TJuD++++Hi4uLun01JES9n8uWFxwwZGkf/7Mdq994GKc+n4rP1hQM/G1453Fc/vY17PrwGSzedlQt+27TQWye8QT85k3Hkl3aVBEx++zffTj4/mSsf+UhzFm7p8D6FVPux8W5L+Lnp8Zh4X8Dln7efQInZz6DPe88hu93mF6DFtIysvDHtoMIXDIb855/EJ//nbffJ3Z+9hreeGDkdZdpwT8sGhdCIhD+60d4ZGhP/LT5QIH7BP74Ph4Y2DXn9v4LgWjXsDYOfv4KRnRtjV+2mgboWtrMuZ9hz6a1OLZnK2Z+UnBgoNizaR2CzhzDU489XOwyrfyx4zDG9GiP6GWfqGDa0Ut5A8kD2jdH0C8foW7VSnmWv3Hf7Tj//Qys+98LmrQzOjYeG3fuReSxbXh/2jP4evHvedY7ODgg9PBmtf7piffg95Xr0aJJw5xlD40djX93aBs8FnNW78Lyqffj1NwXVfAnf7/6i0dGqd+Kwx89i3kbTZ/tt8cNwsY3tP0dX792Ndq2a4+g8Cj4VqqEndvznttZv3YNnnz6WWzfs1+9hv1796Btu3Y4e+ky2rTV7ryg2Q8//4rJj05CZKAfzpy7AP8A08BKs9+WLkdUTAzC/M8j4OzxEm1DZHNBIkmVO3ToUE4a+I3c92ZLrslIIik5EBeX96Tm9Ujq4YkTJ3D69OkbLo9jLhsnf+U1pKTcWHmo0nrtN8qYlQanSvVh7+oFB/fKqoRcYXRJ4XCr10Vdz4oLgnvzQTDqCi/pYLG2GnSwc3Q2tdWrurpdFH1KNJxrmnb89k7uSNj5uSm7KE67bJfMhCi412oKB2c3+DTtioy4vCNHzCXcspJjkZkQCZdKNVCWMmJC4N2oE+zsHeDTrIe6nZ9Bn42spGhVXs7Zt2zbGx4UgLa9Bqjr7fsMRkRQwcyyrIwMxIZfRUZqCnyr1UCP4Xci6NxJfPbiRDg6OSE8yF+z9kZduYwW3fvDzs4OLXsOQPSVgj/wmWmpiA0Lgb29PVzdPVG5dj0kRIQhLuwKMlNT4F2lmmbtjb4SiObd+sHJ2QWN2ndTbcgvPSUJsVeD4Vmpap7lRzf+gw6Dx0BLqYlxqNuyPTx8KqNyrfpIS8obGBBpifGIuRoE3+q18yw/sn45Ot12l2ZtlbIero4OqOXtgSZVfVXZufz8YxIwoVPzMistl1t4ShpaVvGFp7MT+tWrgeCkvL9vUWkZ8HByRI861ZCVawSXwWhEZGo6QpPSUNfLQ7P2huoz0d7ZAw52dujq4onQfKU7g3UZaOXsjpoOztD991vf3sUDehjxbnwwko16hGj4+xZrzEJjezd1vYmdO+KMhZfOiDJmoZKdk3pdQjKOEqFDJgzwsisfU1Uyk4jIdkkWj/RliipZVNx9b7bkWkJCgurz3WgZFulnHTx4UFV8uFHmsnFSrubo0aMler2WeO0lJZlAUgJIygZdCo/B0HZNYG9vh9vaN4VfeMFSu1IyKCgqHi6OjvBwdS7RNqXNPyAAffv2VYGf22+/HX4XLxb5f5T3TwZXmvvyoaGh6n/UoH59i7dTzhdIhohc5Hp8Sjo6NqqNat4eqFfFFwmphX8uL4RG4e5ebdX1htUr4WJotHqfa/p6WbS9UuJYAn0xyWmqzJynqwtq+niiZe2qyMwuvFqB3P9yVBwaVPVVt2tX8sKlyDhcCI9B/So+Fi8tJ6Xi4pJTcSUqFh2b1oenmyuGd22LS6GFTydQllIzMlUps8TUdPiFRmFQ++ZwcLDH7V1a42Lo9UtbDenQQr3mUTO+VY9x8ao25bCys3Vo1qQx6tWpA3c3twLZGD4+3ug/fAzqtGiHfzdvLXKZpcnnQN5feY/kvbmtSyvVr769axv4leC9cnVyxNertqvsoxe/WapJm/2Dr6BP105wdXHByMH9cCGg4Lkotd+KiEJIWCQa1s3bRw2NiMTtA/totC9LURe5LvuIro3roIqXOxpVr6R+R/KLTUmDf0Qs6ll4P1CcS35+GDz0NnX9tuHD1e3c7n/oIWzdsgn33T0W3t4+Rf6WaOXiJX8MGzIYjo6OuG3wQPj55z0XdejIMTzzxGPIzFVG8Xrb2CqD3mDxS0VjE2cDgoODMWDAABUhl4M1Cb5Urly5xPeVgz4p1bZly5Yb6ixIqrkcVMoPnJeXlzrIvB4JzvTv319tI0GmWrVqYfPmzXB2di7R80rZOBm1JO2X55YD1JKWDriZ1y5tzF13uyRBuMLkCYZJ0MWuYPwxI/QUdAlh8GwzQt22d3RF4sGfkR0bCOfqzeHb92n1A21xdvZ521tEIC8j+KBqo3O1Zuq2Z6f71F9dYhjS/TbDqbKpI2H55jrAqDcFsmS+HrtC3ltdSjzOf/8cHFw90XTCeyhLqr3/Bd5k7ia5nV9amB/8/3wfLlXqwLVqPZQl+b6Yy8XpsrNh71Cwvcd2bMAfn85A807d4VWpCtw8PPHyPFOJxL+/mQMnF1fN2mtfgvbu+ftn7Fr6AzoNvQOOzi6o1aQl/l34CY5uWqECTG6ePmXSXn0R7d3w3Ryc2bkegx56LmeZfEf9j+xVmUaayrU/kHmH7B0K/kz++eFUhF48jbEvf5izTAKIKXExKiCnFXs7uzzlwQoblFDFww13LlqFy7GJuLdTC8wcZfkOQlEc7O2h+69evc5ghD3y7u/lloujA4b/vgGx6Zl4sVsbPNGpBRIysnDXss0quPTtCO3a7wC7nOCP/DUHVXJLM+rxZPQlZMKI571roa+rD171rZdTiu5itnZ1n+X9NHfHJVBll+/9NTthSEJ7+2snkPyNadikj0UdO1d4oOD3k4hIK9JfWb58uTrZVa1aNezYseOG7iul2iTw8uGH136fr+fYsWMYPnw4mjZtqh6rpIMDT506pQIP9evXV32w++67D59/Xvjo+aLKxi1duhTx8fFq/hlPT88S9/lu9LWXRn/vyYUrcCIoHF8/Nkb9npsHpshcEo4OBfsm8zcdwPwNB3B/3w5wdnQo0TalTU6GmYNoWVlZcHAs/NTHsmXLcMcdd+SZi/j2ESNQp3ZtVX7Q0iSg1uuNb9T1M19MRe6jORk0U9h7te7oeZy5EoHXxw1St+/o1hoTv/hT3f+jB2+3aHsvhEdj/FfLUNXLHTvfejTPXEQysKcwt83+Rf3ff356nLo9umNz3P3ln6rvv+hxy2bhbz56Fi/M+x3dWjTErMfvzpn/ROZDks+ltVm8+QA+XbENd/fpiNs6tcyZt0kCco6F9KPy83Rzwe+vPqqu/7z1IE5dLjjA1BJy90N0Or0arJjb/E9nq0tAYBAee/ZFjLhtSKHLLO2pL3/DYb8gfP3sfeq7JW01zz9Ukv3S4I4tEfDTh+r+d//vO1y8GoEWdWtatM2ODo7I/q8/nZWVDUfHgp8DKS838N7HUbt6NTz38L1qmZyTfP3jr/Dg2FEFAkeWkJ6lU6VGxbHZz+fZlxX1fXv0m79xLDAMXz+q7cDQon8rCr6/jZs0xar1pvJ3r06bAlc306C8siLty9PefOcrZL+65M9lmPz8S6hVsya2rF153W2IzKzvV7EQMqpr7969aqSWHIwX12G4kfsW56effsKgQYNUVo9cShq8kBrSnTt3VgEaCfjIAf+KFaYSSiUlWUT79+9Xo8OmTp2Kr776qkTb3cxrnzVrFnx8fHIu9erd3AlOR8+qSA86gKzoSyqLyN7JFYbsdBXUEGn+u5EVeQFujXvDkGkaPe7gXQPuzQagyrDXoU+Lg/G/5ZamgixGo5qLSOYmsnfxVMtVe/+bwyHtwkYY9Vlwqt5cLVfrM1NgyEiCPiUScChZ0K80OPtWR3pEAFLDLyHq0CqVVWSe20eo99jODq0mz0etfg8gbOevpvXpyTBkpkOfmQJ9hnaTKrrVbIyEC/uRHh2MqMNr4V6rWZ72ShaRS6VaaP3UN/Bt1h1Rh1ar5dmpCTDostT99FnalTxq0Ko99q3/GxHBl7F79VI0bNVOzdmTlpyYk0XUqmtvvLdkvQoQHd22Xq1PjImC3/FD2LN6KVp3u34AubTUbdEOh9YuRVRwAE5sXYM6zdtAl5WpsodERloKuo+8By99vwqx4SEIPntMBYy6DB+LqT+sUUGj8/u1K49Xp0VbHFz9G8IDLuDyiQOoWq8RsjLSc+Z+kiyioY++hGe++QcntqxCcqxpRF/AsX1o0LZzgQ6GpVWp2xBH//0LQaePqKCPm6c3MtNSoMvOyskiuueNuZj8xW/Y+P0nOdud3LoabfoP17St8rskgQuZi0jmJqrk7qKWJ6Rn5kwOXM3DDbNG98XayXdh+yVTVl9KZhYSM7KQmqVDvIaTgtbydINfXCJORcbhl9OX0KqaaTRpfIbpxFU1d1cV2Fg4qi9+GtMf6/yvqJMNEghbNeE2TOneFl8cMtWr1kJDR1cczkrBFV0G1qfFo4mjKRhsnmeogaOrCh7NqtxQBYj2ZJj2cQl6HYKyM/B3agw6OmuX+VTLzgVnDSkqo+ikIVmVkpP3L8N4bSSnZA1FG7PVOvPtenaueNSxDrztHHDeqM3vsKXJ59+S2UT5J98motIhAQ/pd505cwYxUiolLKxU7lscKcv9ySefqL7Xhg0bShwkkj7UjBkzcODAAfj7++Pvv/9WwaIbUalSJdVnPHfuHBo0aFDiPuONvvbS6O/9Ne0BVRZoRMfm6NyoNn7acQx+YTH4++BZtG9QU2UYpWRk5QQ9Hh3QBXs+eEplDB25HFroNpbWqmVLVW5O3qcF332Hzp06qeWxsbF57rd02TKMG2cKXkjgTQajHti/H5MmTcLnn31m8XZ6u7uq8nFycXdxVllBv+06jj3ng1QJLMnUSUrLyAluLNpyCNtOB+DRwV1VKT/x+q//YsXrk7Drw6fVnERpuUaQl7a2dWvgwpwXsOedx9VJXznWkLmIZG4iyX4SCakZOb+VUUmp2PTaJPww+U68uWyLWj5n3V61bPXUB/DOX5btl4zp1RFBv83B8nefRf3qVXDycgiOXQrG/FXb0LGJ6bsQmyu7PT45FSnpmUhMSVeZOEUts5TnxwxQJeQ+eXws2jesgzWHzuB8SAS+37AXnRrXNbU3+VrfPjoxBelZ2SpDRrKQhGTKnAoMxecrtmFIxxbQQqVKvli17l9s2rZDnZCWfkpiUpL6Tql2xsQiPCISZ89fVBkxRS2ztOVvP6lKyI3p2R6dm9bHj5v2qUDP0l1H1echMztb/a+FXm9Q76UEOOSvTq9X6+T6pbAoxKWkqu+spbVs0hA79h/GyXN+mPfzn+jctpVaHhtvKmst77GPtyeOrvsTkx+4G7O/+RHpGRl44IU30K9bJzRtWA8pqZafm8rdxUn9TsjF280FTWtUxq+7jmO/3xVEJqbC18NV/T5IwFO1PyUN3z85FhvefATvLTcNak/LzFbLM7KzEf3f/s3SOnbqjN+X/IoA/0tY+vsStO/YSQVUkv8bUCHvr2T97ti2DatX/IM+ffuZsqYiI9X95K8MRNBKl44d8P1Pv+D8hYtYvW492rVto+ZVkqxYtb5TR7Rv2waHd29T1w8cOlzoNuWB0WCw+KWisYnwYcuWLVVWjZDMnOJGP93IfYsjB4+tWrXKyTRo0aJkP65yoN62rSntW34YW7duXeBA9Hqkk+DubpqDQ7aXkWolcTOv/Y033sC0adNybss2N9Nx8Gx3JxIP/QKjLgPene5Ry1LPbYRL3Y5wrtIQ6cEHoU+NQcbVY3DwrI4qg6fBq+0YJB5eopa7Neiuyr9pxaP1KKSeXaOyXDzajVXL0s6sgnvrMbBz8UBmqKlEX0bgXjhVbQLPDvcgaf8CtZNw8KwGjzbazTti7+CEOoMfRdCqT+DsXRUN73hFLQ/4639oMWkuspPjcP6HF9VrcalUE3WHPqHWX1n3FZKvmE6opkcFod7wZzRpr7NXFVTrMhIBSz9QAa2qnYerQFbAXzPRYtJspF69gIDlM1WGhmu1+mgwyjRC79Lv7yArMRqJ/kdRq+8E1Ohp6qxZWtP2XdCoTUd89fLjquxcGynhFnoF//7yLR55azaObF2H5V/PgrOLCxq27oCO/YchLjIMMx8bC+/KVTD++TdQp4lp7hQtdBg0EpdPHMRPbz+DbiPuRrV6jXDpyF5cOX8SQyY+i93LF2P/yt/h4uaOpl16o16rDnD18MKKz2eoDKPaTVup5VrpPe5h/DP3DZzZuQEDH3oWLu6eOLxuGZxd3dBhyBis/3YWLh3eDVcPT7QdMAJeVaqr7c7t2Yzed2tfm/q2x6bi7zlvICMlESOfe1Mt2/nHQrTsORD123TG7++/iJirgfDwqYR+95q+a8L/yB6Mf137ue/eH9Ebb67do0a+zb2jn1r21rq9eHd4T9Twcld/X1uzWwWOnu3bQa2ft/sklh43pclfiIrDLxYecWrmaG+P6b3a443th1Hd3Q2zh3RTy1/YsB9L7hqofjOn9WyLlzcfVCPPZA6imLQMlUUkJyHqenvg1V7toRVfB0eMdKuE2QlX0cjRFRO9TJ9NuT2zckN42jtgmFslfJQQAnd7BzzhZTrh9XLcZbjY2WOgqw96unpr1t5a9i6oa3TFX7pINLJ3Q1M7N1VGbq8+AaMcTSUmLxvT0dLuWuDKz5iGbfo4lTV+tDZGAAC/7ElEQVRVy84ZLey1C2oREeXXoYPpd0p+D2rWrKn6JbVr1y7xfW+1z1elShX1WDfa5/Pw8FD9N3ks+VtS0nczD0SUNkiAqiTVJ270tZdWf8/snp5tsedCMO77aikeGdAJjatXxuZT/jgTEompo/qouSckICRl5ga3bYLODWujY4NaBbaxNAmIvTxtGiZOmoTmzZtj4XemEe73P/AANm00jQqXAF/zZs3UwE4hATvJKpJy0rLNJ3PnQmvv33sbnvruH3Uyde7Do9Syuat24s5ubdC1aV38vusEgqLjseLgWTSpWQWb33sCL47qg2e+W6FOvt7dq50mJ67NPr73Nkz/YyOcHR3x1URTxZCpv23A3PuHqWyjQR/9pAZo1vL1wiujequSg5P6dsDEb/9RZdSe/u94UAvOTo744OG7MPmzn1Gnqi++n2aa++T+md9h0+yX1fXnv/4Ne8+aSjGdCQrFp0/fW+gyLdSq7I3Hh/XCA3N+QofGdfDw0B4qcDHp01+wboapbz/uw+9xNTYBW05cxMtjB6sgU6+XP4WPuyseHNQNI7tpczJ47v/ew9NTXlFl576eO0ste/P9mXjx6clo0awpRt/zAKKiY9C0cSN8+tH7an1hy7Q0vm9n7Dp9CRNmfo9JQ3uiWZ3q2HbiAg77BeO1CcMRFBmLIa+bMkR7vvQxVr//LP49fBbfrt2Jyl4eeGpUP9SrZvl9mZenB9547nE89NIbaNaoAX6YO0Mtv+eZ6dj25/e4eDkYwx58Gk5Ojiog9Nk7r8Dv8hXsOXxcXcS0yRPx8pOToKV37x6MZxatQkJaBmY/YBpM+dm6Pbi9Q3P0aFYPE79ejksRsWo/8dJ/c+cu238aH/6zXV0f9fEvOPSR5c9hDRoyFNu2bMZD907A6DvuRLfuPXDqxAn8tXwpPpg5CxfOn8e94+5EjRo1MffzL1Gvfn3Ex8VhcN9eanv5+90Pi9G3v2kKA0t75KEH8PRLL+Pehx/Hs5MfQ62aNbBi9VpEREbimcmP48F7x+PFV17HyLET0K1LZ1VeTuTfhqgwdsYbnTRHY0eOHFGjtNauXatuv/3226ocQGETcxZ1365du2LKlCk3VG5u586d6jlk5FViYiLeeustLFiwAOPHjy9wXxmdZH58KRPw4IMP4uOPP1aZPW+++abK6DF3PkryeqVk3HvvvYdmzZph5syZePHFF1V2UnGvobReu3Qa5IC6+thPYO9UtmmUJZGVXHC+EGvWqOu1SSVtgZ29BuX/StEzo1rClsSmaTfipDRk6SreSAqtPLLnWlaSLUiLMo1csxXn/j4LW3IgTrsydbcq02jA57pgdazk7a1dUKy4Y5gBczbA0c1yAS9deip2vnq7VbxmovJi3rx5qty39GHE0KFD1TIJpJT0vnv27LnhcnOzZ89WpdokiCL9KckqunTpEurWNY3Yz23RokU5jz937lxVMk7Kep89exbfffedKlPm6lqyEsTSXnnu//3vf2pA4uuvv65KlO/bt6/Y11Aar928r7z67WtqtLctcJrwOmyJfuWnsCVZGo3YLy1u9bUp+15aDBmWz+IoTY79tQmClQb9vn9gS5zblV3p75uRtnMlbEnG2FdhS7z1tlPFISkpGdUbNrOK/o/5OObQA7erkvSWkpKVje6/b7CK16wVq88kkn9Ex44dc25L4KOoEWVF3VdGBkmQ5UZIoEYCRFLvWUaJvfvuu6hRo/Boa+7H79evnyoP98cff6gySfK3pAEiMylzJ6PIfv31Vzz66KOYOHGiKl9Q3GsozddORERERESkFclskfJrZtJ/kQydG7lvnTp11NwCN+Lll03ZAz///DNGjx6NyZMnFxnoyf34ElSSCg5Sorx69epqQF5JA0RmUs5MAkyBgYH4/vvvVQUJ6fMV9xpK87UTEREREdlMJlFRc/bIvDu5SUBm8ODBN7yd1G7Mf0Bf1GNp8bwS7MmdEVRabSgpZhJZFjOJLIuZRJbFTCLLYSaRZTGTyHKYSVQxRpURlYWrV6/iwoULeZb5+vqqSgk3ul1mZmZOWe7rPZYWzytzGeXOCCqtNpQEM4ksj5lElsVMIstiJpHlMJPIsphJZDnMJPJGRWCTw43Cw8NVObfcJE3/eoGSwrYrLFhT1GNp8bxff/11noyg0moDERERUUUg45+M/02YbanHJyLLO3r0qOob5dauXbvrBkoK207me01LSyvRY2nxvAMHDsyTEVRabSAiIiKqCIx6o7pY8vErGpsMEskkkjcyv9Ctbqf18xZXS/pWXwMREREREZG1u/POO9VFq+3K+nkt9VhEREREROUySEREREREVBiDwagulmLJxyYiIiIiIqLiGQwGGPQGiz5+RWNf1g0gIiIiIiIiIiIiIiIi7TGTiIiIiIjK15xEFpw3iHMSERERERERlR2Zg9ai89AaKl71CGYSERERERERERERERERVUDMJCIiIiKicoOjyoiIiIiIiMovgx4w2FtwHlo9KhxmEhEREREREREREREREVVAzCQiIiIionLDYDCqiyUfn4iIiIiIiMqGUW+A0d5g0cevaJhJREREREREREREREREVAExk4iIiIiIyg2jwXSx5OMTERERERFR2TDqjTBacE4io77iVY9gJhEREREREREREREREVEFxEwiIiIiIio3jEajuljy8YmIiIiIiKhsGPRGGCyYSWRgJhERERERERERERERERFVBMwkIiIiIqJyw2AwqoslH5+IiIiIiIjKhlFvgNHeYNHH18L+/ftx8uRJDBs2DI0bN0ZZYpCIiIiIiIiIiIiIiIjIwrZv346pU6fC0dERR48exfLlyxkkIiIiIiIqLUaDUV0sxZKPTURERERERMUzGC1cPcJo2T6fvb09fvrpJ7Rs2RJubm6wBswkokJlJyfAzinD6t+dzOQ42JLMjGzYEjs7O9gSJwfbaq+taVbVA7YkU6P04NJg5+AAW2LnYA9bEpelhy2Jz7adz24WbKetRER0zdjw3nB0tY1juyO9n4Mt2bPqE9iSz7cHwJaMrFETtuTLledgS/YOtJ1+ycl2D8CWVHW3rVOw65o+DFvyTHYibIldZgpshX2G7bTVVgwYMED9zciwnnPvtrWHIiIiIiIqjoUzieTxiYiIiIiIqIzojTDaWbBfpjc9dlJSUp7FLi4u6lIeMUhERERERERERERERET0n3r16iG39957DzNmzEB+Bw4cwIkTJ1CccePGoXr16rBWDBIRERERUbkh9aPtjLZbn5qIiIiIiIiKZtAbYLAzWPTxRUhICLy9vWFWVBZRaGjodYNEw4cPhzVjkIiIiIiIiIiIiIiIiOg/EiDKHSQqyt13360utoxBIiIiIiIqN4xGy85JJI9PREREREREZcNo4TmJjP/NSVSRMEhERERERERERERERERkYVeuXMH69euh0+nU7c2bNyMmJgZt27ZF3759URYYJCIiIiKickOyiCyaSWTBxyYiIiIiIqLynUmUmJiYM4fRU089pf7K7UqVKqGsMEhERERERGRBhw4dwquvvoozZ86gZs2amDp1Kh5//PEi75+ZmYnvvvsOS5YsQVBQEOrVq4dHHnkEL7zwAv9PRERERERENqxdu3ZYsGABrAmDRERERERUbhgMgJ0Fs33k8W9ESEgIhg4dqoJCv/zyC3bv3o1HH30Unp6euPfeewvdRu4XGRmpAkV169bFrl278PDDD0Ov12PKlCml80KIiIiIiIhskEFvgMHOYNHHr2gYJCIiIiIispD58+fD19cXn332Gezs7PDggw9i69at+Pjjj4sMEk2ePDnP7bvvvhs///wzdu7cySARERERERERlSr70n04IiIiIqKyYzQaLX65EXv27MHgwYNVgMjstttuw8mTJ5GSklKi13P48GHs27cPo0ePvqn3hIiIiIiIqLxQ/TKDBS/GijcPLYNEREREREQ3KCkpKc9F5hEqTFhYGKpXr55nmdyWjkdERESxzyEZSC4uLujRoweeeeaZYucxIiIiIiIiIroZLDdHREREROWG0WC6WPLxRb169fIsf++99zBjxoxCt7G3ty/09vVGqIWGhiI1NRU7duxQJegqVaqEadOm3doLICIiIiIismEGvREGWHAeWn3FyyRikIiIiIiI6AaFhITA29s757Zk/BRGsoaio6PzLDPfzp9hlJ+Hh4e6TJgwAadPn1bzGjFIRERERERERKWJQSIiIiIiKjcMBiPsDBYcVfbfY0uAKHeQqCi9evXCunXr8izbvn07WrZsCR8fnxI/r8xppNfrb6LFRERERERE5YdRb4TkElny8SsazklERERERGQhMpfQlStX8PHHHyMjIwNbt27FL7/8gilTpuTcZ+3atfD09FT3E9OnT1eBpPT0dGRnZ2PDhg2YN28eHnzwQf6fiIiIiIiIqGIFiaQcx7///ntL942Pj1ed71uxb98+XL58WfPnLeljleZz3SxDRjL0qbGFrtMlhiE7Phj69PicZXLf7LggddGnJ0BrhoxEGDKTC12nTwpTF0NWasF1KZEw6rKgtczYEOjSEgtdlxZ6QV30GSkl3saSjAY9MqKDoM9MK7hOr0Pq1bNIC78IQ1ZGibaxNF12FkIuXUBWxrX25JeanIgrfuduaBtLyUpPQ8RlP+h1ugLrUuJjEXjqCMIDLsBguDaqIj0lGWGXzkGXVfjE6paUkpSAqwF+xc69ERV6BTHhoTe0jaUkxEQh4kpgoevCgwLgd+IwosNC8iwPC/QvsEwrIfFJCE3M+93PLTkjC8dCInE51rSf9YuKw6Hg8GK3saTL8UmITS/6exOXnolj4TEITzbtC45HxOB4eAwSMrT/7BqMRoTqM5FuLJi9kWE0wE+XlnOJ0WdfdxtL08OIWGRBV8iIqnToEY6MnEsarrUvEwbEQPvfNUsxGowWv9yI5s2bY+XKlVi8eHFO6bg333wTTz31VM59dDqdmnvIvN+WYJCUlqtZsya8vLzw0ksv4a233sLs2bNL/f0isgZnz57FsWPHbum+Fy5cwOHDh2+pHStWrFDfRa2ft6SPVZrPdaOMBgOSgk4j+co5ZKcW379IDjkPQ7bpdyXl6oUSbVPa5BjSkBIJQ2o0jPqsEq2/3jaWJL8DJw4fUJfsrILPnZGeptadP3Ucmf/1PwpbppW0uEhE+51AfLAfDIaCxzzZ6alqfcylUwXWZSTFIzE0AFpKio/FxeOHEHThTJH3Mej1qn936eQRdTsuKkJtc9X/IrSWHhmIlJCzyIwLK3R9VmIUUq6cURfzfeSvbJNRxDaWFnI1FEH/DXbJz++SP/YeOKguoWHhOcvPnjuPq6Hat9f//FmcPHwAsdGRha6X79Spo4cQHxuTs+zSuTNqG7kkJlw7r6WFE0eP4MjB/UhOSizx+uttYynyPbp88oi6FHbuIS05KWe9XBKiTJ+HjNRkXL10DlmZ2p9fkUFZZ86dR1Yh+97k5BTsPXBIXY4cO5GzXH4v/PwD1PLMTG37qUnJKThzwa/YCgNXwyIQEGT6PoZFRGLvoSNqm7I4v2LRTCILXyoaqy83FxwcjPnz52PEiBE3fV/5wgcE3NpByLJly9C3b180btxY0+fNPXHxF198gdGjR9/UekvLijyP1DOrYGfvAJd63eDWdGCe9RlB+2BIj1OBIddGfeHWuB8yAvdAlxgKOwdHONfpBId63TRrb3boMWSHHpE9O5wbD4BjtZZ5X0/wXvl1U212aTIEjtVaqOX65HBknFsF11Z3wMG7tmbtDduyACmBx2DISke9O1+HR902edZH7v1NrctOjESDu9+FW81m193GkoL+/gDZSVHquRvdNwsulWrlrNNnpSNi588wGnTQpcSjyUNz4eRVtdhtLEkOQt6fdIf6sZSgy3s/r4K7V8HyQSsXfoFLJ49ixi+rS7yNJSTGROKb5++Ds6sbXD298MxXf+SZEP2q3xls+/VbpKckwrtqDTwxdzFiQoOx4KUH4VWpijpoe2HBP3AqYu6M0hZ04SxmPT8Rbu4eaNquE56f+VWB+2RlpOPz6c+gQ+8BuO/5V0u0jaUc37UFP374GpycXdDvjnsw9smpedbv37gKZw/uQUxYCIY/8ARGTnoKiz6YjovHDsLO3h6Dxj2IEQ9N1qy9iw+cwbzdx1X5q7eG98T4js3ztjcwDJP/3IQGlbwwvFUjvDigM5Yf98OWi8G4s31TTBnYBVr6cNcxbAsMQ7pOjy9u74UedfLOx7LqYhA+2HkMTSp54/62TTC2VSPMO3gWyVnZCElKxbej+qJjzSqatffr1FBE6bNVQGi6Vz3UdHDOWRdv0OGvNFPHMcGowyjXyhjg4FvsNpYOEK1AuJq6U66PRU24wCFnvQSPjsLUSUyCDoNRFe7/rT+GBIQhE3dDm/1uRXT77bfj4sWLKivIycmpwPoxY8YgOTlZBZFEx44dsWbNGvU7I5fc+3mi8kgy5xISEtC5c+ebvm9sbKy63AoJxHbr1i3nu6jV85rt2bMHQUFBqg03s96SDPpsXNm4SO2TMqJD0Oy+t+HbrOBxRPzFQ/BfPhttn/kKblXqIHTHn8hKiUNmbBia3f82fBp31KbBRgMM4ab+njErBY7NRsDOxafY9XDyLH4bC8pIS8P8jz9AcMAl/PrvDtSoVSfP+rjoaLU+OysTyUlJWLx6M1ISEwss8/bx1aS9UReOw3/738hOS4a9kzNue/tH2Ds65gkinVw+H0nhQRg3b3OebU8s/RIZiXEY+Ip2x/jBF85i5fdfICMtFTP/3FhgfXpqCmY9dS8y09JQqXoNvL7gT/gdP4TNS39SAaZGrdrh2Y/madbe6OP/Ii3CH1kJkajcdhDqDn0i7/pj65HodxD2Lm7wbdEbNXuNR/zFfUi4uA9ZidGo0n4o6gx6WLP2/rZ0Od6a8T842Dtg+tQX8PQTj+VZ/7+P5yL4SggcHB3xzBOPYcLdd2H8gw/jkv9lZGVn4Y1XpmLSA/dp1t6/fv0Bx/bvwX1PPINxDz5aYP2vC75CdGQ4QoOD8P6X36FbnwGY/dY01X7xzPR30LF7L83aO/+zObh4/ixmfT4PfQYMKtH6621jKdmZGVizYA6iQgIx7fsVqFKrbp71UVcuq/UiPjIU4156Vw1oXTDtEXhWqqIGJDz31RJ4+FTSpL2JSUnoO2y0Oj6X0s57N62Fq6trzvpzFy/i4aeeQ726dVCtalUs+9n0O3jPpMdx4tRptfyPHxeiZo3i5xktLRL4uW38RHh6eqBB3dpYs2RRgftI8OjxKa+haaMGmD/7A+w7fAzzfvgF0bFxaNeqBf5cqN2+l2yL1QeJSkPt2rXVCMyK8rxlISNoL7x7PwV7Z08k7v0Grk0GqB2smWeH8eqvITMFKSf+VEEi4dZsCBy8asDBTZuDWbPsyNNw6/IojPpMZJ5fXSBI5NbuHvVXsol0kWdUkEh+CHSRZ+FYNe9JWEvTZ6YiLewCmj3+LdIj/BFzZEWBgE+jCf9Tf5P8DyHx4l44V6p93W0sJT3qsgr8NX/8W8Sf3Y7405tQs/+1A1RHNy80edB0UBBzeCWSLx+FW61mxW5jSSf3bEObHv3wwLR3sGLh5zi87V8MuPPePPe5dOoo6jZpAf9Tx0q8jaUc27QSA+9/Ar3ufBB/znwFQaePoHGH7jnrW/YYoC5i0fRHkRgdgYsHd2LYoy+ix+h7sWz264gIvIh6Ldtr0t6NS3/Cc//7Au169sO7j4xFXGQ4KtfIeyJ6/e8/YPDY+xAbGV7ibSzW3t8X4Z0f/4FvtRp454GRuPOJl/KcnB331DR1yUxPUx1JCRIFnDmOe198UwWJln39saZBot+OnMPOF+9FerYODy/5t0CQ6H8b9+PTuwaoAJGZBJO6N6yFs+HXRsZpITUrG4dCo7Fp4khciEnAN4fP5QkSSQbOR7tP4Pdxg9Gi6rXfhB/uNH2edwaFY4N/iGZBIskGkoDLTJ9GOJSVhF2ZCZjgfq29tRyc8aZ3fXX9m5QwdHP2uu42lhSCdNSCK/qgMo4hEQFIQ2t45ayvCzd1EesRidowBYqjkAkfOCEC2mdqVUSFBYiEg4ODKjeXnxxL5T6eIqKi9enTp0I9r9YcnFzQ9qkv1fW48/uRGHCsQJDIoMtG/IX98G3eNWdZi4dmqL9JgacRc3KrZkEiGbzo2GyUqV2xl2BIvAKH6u2uu764bSzJ09sb3/+9Hq8//Uih62vXb6DWi2/nfIjjB/ZiwPBRhS7TQsPet6uL2PHJC8hIioV75Ro5633qNMbQt77H5g/yBgtCT+xG9ZZdcOVg3sCRpbXr1V9d3n/4zkLX7/jnd9Ss3yhPIKjn8DvURbw3cbSqIOGc6+SxJdW//Vn115CdgYs/Ty/0PjV63wPPuq3hUqmmui2BIrlIQPfCD1M0DRJ9s3ARDu/eBhdnFwwaMaZAkEjMfP8dNG3cGLVqmj4n5y5cxNyZH6hSvJ98OU/TINHrH32GdX/9XmQGyEffLFZ/Tx87jDVLl6ggkau7O15660NUr10HPr7aBDDMvv9tGT6b9b8bWn+9bSzFxd0DL327DD+/92Kh6xu26ajWi+9fnYzWvQfh2JY1aNalF7rdPhYnd27E0c2r0X+8Np/ff1avw/3jx+LNV6bilbfew4Yt23DX6JF57jNi2FA8+chEtG5pGjz+7+atCI+IxPkj+4o8treUH/9Yjo/eno777hqNux5+GqfOXUD71nnPZy5ashT33jUaR0+eVrfHjxmhLqLv6AlISU2FZyGDYYhsZliilHr7/fffS5SZk/++N1uKLSIiAn/88QdOnjx5U228mec1l667dOmSeqzAwMJLHl2PPPdff/1VbMmE0mTUZ8PBrRLsHJxUwMeYnV7o/bIiz8G1QW913d69CjICdiJp3wKkB+61eBtz2mrQwc7RVbVVgloyUqwo+rjLcKzVQV3XRZ2DY402csYGWspOioFrtUaqIyPBlOyk6AL3kQPB1JAzSL58GO61WpRoG0vJSoxUbRDutVuq0U+FBb5SrpxSgSy3Gk1KtI2lRIeGoElbU2e1SdvO6nZuUvpn/4ZV6J8rCHS9bSwpLvwq6rUyPXf91h3V7fziI0Jxfv92GPQGNRpHAkKXju7DxUO7ERcWgip1GmjW3qiwEJUNJCc5m7btiKiwvO2NDruq1tWo17DE21iSjC6sXrcBnF1cUaNeAyTHFxwZHHzxrPpMNGjZVt1+6JUZ2LV6Kc4e2gN3Ty+VQq+FbL0eHi5O8HRxRjVP90LvE5mUii71auJcRGyZp5ZHpKSjRRUfONrbo231yohMyfs7IbfrennA28VZlaQzy9YbcCg0CtuDwtChhnZZRNH6bDR2MJ0IaOzghhiDqZxcfnGGbDjaAe52DiXexhIkO6j6f4Gf6nBGMgqWoxRSaq4aXGAHO8hUn35IRSsUDE7YMmsrN0dEJXfkyBHVB4qLi7vh+95sKTbZTvp85jnBtHhec+m6gwcPYunSparvdjMkq+i3337D5s2bNf2dT7x0BNW7Fqz0EXlwDWr1HlewTF3gacSd2wPP+q01a6N6bn02DCkRMKSEw86taonWX2+bspQYH4cje3fB/8I5NGresshlWkqJDoNH1dp5AkRF0euyEXn2MOp2zlt1xBpcuXQevUeMRdD50yqryCwyJAhHt2+Ep08lzQJEuSVeOoRqXfKesBbOPjUQfXQtLv78MkK3/ZSnTF382Z1wq2YayKQV6bdJloW3txe8PD1V+cTcmjdrinfen4keA4bgx19+U8s+/uA9LPzxJ5w4fUYFiqyJ7E9PHNqPHRvWom1nU+Zm7boNMOftl3H/0J44uGt7WTfR5oVfvojKterA0ckZnQaPhru3L7b9sQhelaoiVsMy7lIisVsXU4Zwj66dERic91jA28tLlaKb+ORzGDDiThVYlNv3jR+L8xf9EBN7/eOV0m3vVXTvZDpH2aNzBwSF5D1XEhsXj7DIKHRubzpXkTsDafXGLfDx9oKHe+HnDmyNQW+0+KWisYlMonPnzuHJJ59EjRo18MILL+DUqVOoU6dOie8rB903Wort6tWr6Nq1qxoZlpSUpMq5Sbm5G2njzTyvlK577rnnUK9ePfUaJRNp48aNcHYuebmasLAwVYJALrIDe+2111Rnxtu7YDksWZ979IS81puiyqEYYGdnD6MuUwVg8ssI2g87B2c41zR1DtwaS9m5vqrec9KB7+HWSKNReHYOMswtT9sLOyjIDt4Le5+6cPCsoeYg0scHwd7VF8asVBjSYmDvVUuT0b32Ti4wZKXljCSycyz4WTBkpiFy1y/qhJ9v64El2sZi7XXM9dxZaaot+WUlRKj2SrvkUpJtLEXKtpk7AlIH1yXfwf/uNctRq0FjNQ9NRlqKmqvmettYur2Z5udOS1FBoPwuHNyJw+v/Qo2GTWHQ6eDi5oHk2Cj8u3AuqtSuX2xgtLS5qPcqWQVP5K+LmymTwezv779A7+F3IsT/ImIjwpAUH3fdbSxJ1aI3GFT2kPyP5f3Ob9MfP6p65ubssTbd+6pLZno6pozsDld3bUblODk4IFN3rQ6xvpAT175urnj09w1ISs9E/cre+HViwU6mVtycHJCSZdr3SuaTk0PecSrOjvbI0OvxzLo9SMzMwqCGtfHugM5Izc7Gp/tPQV7eiKb1NGuvi50d0o2m+WEyYICzXeHjavZkJqKXs88NbWMJjrBD9n9zEWXDqG4X5gJS0Bmm9l5CKnzgqLKIZJsEZMMX2o6GIyIyW758uQqayPH1hx9+iDNnzhRZarGw+95MKbZNmzapOcCkJOS3336LkJDiT0yV1vNK6boFCxagQYMGKpPw9ddfx4kT1+Y6KOljyBxnw4cPV4GqefPmYdWqVZbr76kTMjoEr/8OVTsOgVvVvCWEspLjkBLmB486zZCdmoC0MH9Vbi6nTJ1BD9/mGpfJ06XDEHZE0oYK7Z8Wuv5625QhCQR9O3cmvHx84ejoVOQyrSSE+MNvyzJ0efDlEt3fb9MfqNSwBWIDziA7LUUFmDyraVfCvTjOLi7Y+McPSE6IV4PE3v7hb1SrXQ+Ht67HwU1r0axDF+iys+GoYdZA7Mkt0KUnoUbPvIFXUa3zCHWR+X7PL3oedQY/klOmLiX4NCq17g8t6XL1STIyM+GYq/SgeOf16eoipb3G3H0fHpv0IO4YNUJdomNisOSPpbAm8r/+Zs4H6m/X3qYqOG/ONmVSXva7gO8/n4Ue/bUr4VYe7Vv9J7qNMH22pRT+uJfeUdc3/vQ13Dy1KeUv3N3cVMln81w/cju3Vi2aY/u6ler686+8jv2HjsDFxQWr1v6LX/9cpjKKpNxc3149NGqvq2qnam9KwfbO+vJbDOrbC6fOnkdEdAzCI6NQq0Z1rNqwGUtXrkO3ju1U+esbOcdMFYdNZBLJhL0yOkpGST3wwAPYu3dvqdy3OEuWLMHjjz+Ov//+Wz1e/frFj8QorecVssORg34ZoSZ1sRcuXHhD2//666+qDTJJsmQlde/eXWUUFWbWrFnw8fHJuUhw6mY4etdCuv92ZIYcNQWLHJzU/EPGbNOIkLSLm6BPiYK9RxXoEk0TE8rf7LggZF49DjtH7YICKrDj4Izs8FPICjkIOzdTqrAhNUZ1XkTmhbUS7YCdvRMMabGqJrUxKxlZwbthSAqFLuxE3kCTBTn5VEdmbAgSL+xG5M6f4VHHFGRLjwzIKeuQGXcVNfpPgk+Lvog7vr7IbbTgXqs5Ev32IenSQUTt+zOnzF165OWcLCIJXNXo/zDcajZFwvmdRW6jhZade2Dr8l9VCbkNv/+A5p26q1ICYYH+an1ibDQObFytyohJObRDW9YVuo1WGrXvhh1/LMS5fdtwZMM/KpsoLTlRZQ8JySyq2ag5Rkx+GYnRkQjzP4/DG/5W2418+lV1H/9j+zVrb6vOPbDsm09xcMs6nD96EHUbN1OTwCbEmrLbkuPj8NeCz7B95R84e2Qfgv3OFbqNVhq2aod/FnyK3auXq5Jybh6eiLoanJMdJMEhySq7+5mXselPUxmC6NArOH9kv5rLqHW3PrB3uDYPjKX5urngp4NnMH/XcdSvZDqYvhgZh6z/OmqNq/riyd7tVYm5yzEJallwXBIuRcUhLDFFZRhppaanO64kpWCNXzDm7juFLrVMo3PPRptGT1dxc0VVN1e82KMtpvduj5ORscjS6+Efl4RXenXAmOb1seT0Jc3a28DRFSeyU3A8Kxmr02PQ3NF0AB6sy8hTIu90diraOLoXu40WpNTcOSQjGGk4hSR1Wwcj4nEtmykTBqRCr8rLiTTocRlpOIQEpECHAFg+81gL8n+x9IWISl+PHj3UXFyrV69WcwLJoL3SuG9xvvnmG/z444+q/yQBo+ImgS7N5xVdunTBunXr1GPddtttRfbXivLVV1+pQJP0WyVwJYEimfvMUv09CRBd/PUduFauqfpM6TGm154aZjpmzkyIRGZcuAoIpYYFIGzv3+p+qaF+qD/8CVTvNhLhe/+GViQjSKpb2NfuCjuvWjDEmdppTIstcn1R22jl9NHDSIyPx/mTxxEbFalOUAdcPK/WJcTFqvlQnn3tHdSoXQe7Nv9b6DKtxFw6hUOLZ6JBj9sQF3Qe2RlpyE5PQXKU6XMhA9Wi/U4gOyNV/c1KS0FmSiICdqzEmVXfIykiGOEntasmInOeXDx+SA2yk78yj1NSXIzql4gWnXuq8uL3vfgGGrVqr/qCMjiwWYeumPDC67h89qQ65tdKxP6/kBhwGB61myM19IJalpUYBV16ck7GUMqVM4g9vRWwd8hZVrl1fxUwij25SQVotdKoQX189tV8LFj0IypXMp1f8bvkj7Q000DQI8eOY8/+A1j8y2+oWtU0yPGSfwC279qDp1+chtEjhkNLMvdXSOBlNedQ4CXT+3vp3JmcAYPnTh7Ds6++i3EPPYo/Fn2jlp88fADHDuzF2uW/o3q+OcMs7fzZ04gIC4XfhfMICQ5Sy86ePlns+sKWaSXozHGkJsQj5MJpJMZEqXmcwwKu/T5lZ2Yi1O8c6v9XAl/mTb588ogqO7dz2U9oP0C7z0PfXj3x9XeLVAm5BT/+pII9yckp8L9squoUGh6OvQcOYd3GzTh45Kgql9i/dy9Uq1YFH733NsaOGYWjJ0pWfao09OvRTQWC1mzaipXrN6FL+7ZqrqGrYaZ9WWx8AubOX4ivf/gFh4+fwoGjJ+AfGKyyjma++TJOnr2AS4HBKA8kU9nSl4rGJjKJ5EDWnLEhB7bmH5pbvW9xJN3V1/fanAiVK1cutTZej2T8mEfNVapUqcg6qcW1XbYzK+4x3njjDUybNi3PyLKb6Ti4tRiOtAsboE+Ogmf7u9WyrMjzcKrWFI5ONWFIj4chIxHpfptUNo5nxwnICNqnltu7V4JH27HQkkvT25AVvAd2dg5waWxKd8+OPAPnej0AezcYs1OhT5BLEBy8asO5UX+4tTfVyM0OOwYH3waajSyT7Kw6I6Yg+sAyOHlVRbUepvmSYo+sQt1R06DPSEHkrp/VwaGzT03U6DexyG204ODqiVqDnkDciX9VKTnfNoNUBzH26GrUHTkFmXFhiNj5s5qDyKVKPdToMb7QbbRSt2kLDJ0wCVuW/aJqTrfo1B0x4aHY9+9KjH/2Fdzx2PPqIhbPfD3nev5ttNKm71BEXQnAgdW/Y/hjU+BbvRaCzhxDeMAF9LrzAVw4sAMntq6Fs5s7WnTvhwZtO8O3Rm1s/OFz7Ph9IWo1bYXWvQdr1t7b738Uf86bg51r/sJT782Fk7MLLp44oso1dOk/FNO/+FHdL+DsSRWAadejL1p26lZgG61MeP41/PnFR6pj+NQHn6tlx3ZuRssuPdGwZVs1Z5GUF/SqVAUTX/1ArT++a4ual6pBiza4++mSjaQsLXPuHIBZmw7C0cEeH442ZWMuPX4Rz/XriCqObvjfqD74aNNBJKRn4MPRpmzYjRcCsemC6cBQAkzyGFqwt7PDx0O6q7mIqnu64tluphr/Px2/iLnDeqrrb/XviK8OnlXz+swY0AXJWdn4ZN8pONjboa63B17ppc1cWkLKx93vXh3bMhNQ38EFff7LFtqUGY/JjqY5sq7qM9VcRPLaittGC5XghHbwxlkkoxHcURuuSIUOfkhBD5iOCSKRmae0XEf4qIvYhVh0gbbzAxIR5ZZ7UN71+lM3ct+S9vlkomoJ/BSntJ5X5O5rmvtrklV0M30+2U76kIX1+UqrvyeDvPSZaYg9s0tdfFv0QN2BDyB83wo0GTcNXvVa5cxZdHXbElTrcjsM2VkqaCSVHFwqVUeDEU9DM9lpMIT/lxHk4gP7mqZyzfros3Bs0L/w9UVso5WFn38MnS4bv33/DR566nl07N4Lq5cuwdR3Z6oT2As/n60GlDZr1RZ33Psgzh4/WmCZVpLCg2Dv4IjTK0wDWrs/9rYavBhx5gBajnhIBYdOLp8PZw9v9bfzA1PRccIL6r4SUDqz8ns0G6pdH1X6dzLoz93LR/19Yc4CVUJagkfmuYdiwkKw/teFaNq+M9r1GoC1P32DU3u3w9XDE/3G3INaDZto1t7MuFBkJUXj6rYf4ejqiab3fYDEgKNwrVoPXvXbqoyh9IgAOHlXQ6M7XlHbRB9dh/SoQDi4eaHOkMdhr2Em3GezP8KbM/5nysD58hO1bOnfK/DQ/RPQqEEDzP3ia8TGxak5ib6c+7Fav3zFKuzYvQfdu3TGW69q24favOYfHD+0T123t7fDc6/PwPKfv8f0/82F3qBXWUT29g6oWbsOXnpnZk5mkWTrNW3VBpOnvK5pe39d9B2uBAWqi5wrfHjy0/jpu28wd953Ra4vbJlWNiz+SgVidy5brObtbd6lN/av/gN3TzXNUXfl/El0H2mat1xkZ2ZgzYI5quTcg29/gtpNTFMRaKFPz+4Yf9cdWPDDT3hu8uNo06qlKiMnQaFXXnwOe/cfwrc/LFZl56a/+DxaNGuqths2eBC++OY7NGvSGI9P1G7f++D4O3H5Sgh+WLIMn8x4E5V8fbB55x4VKHpg3B1Y/JVp/m8JDEl5ubEjh6mg0fotO+Dl6YEH7r4DbVpoNwiXbIudsawnKbgOKZM2Y8aMnLl93n77bTRt2hSPPPJIie8rZeOmTJmCLVu2lPh5ZR6iYcOGqdJxiYmJaoTW119/jfHjx1v0eeWxpGSAPI88hoxuk+dt2LBhsY8lpQ7M66XU3ZAhQ1Tb09PTsWjRIlXzuiSdAek0SIen0tC3Yeekfc3dG5WZrG39z1vVqJdtpSTb2mTZ0+/V7gRyaYhItq2J4hv4apcVURoy9bYz8mPI+o9gS9Kibm7uhLKye8mNlfApawfjrasue3GyYMBihKhjpcLK2mrJfAzT9Knf4eBiuVrbcpLU/7sHrOI1E5UXUiotISFB9aPE0KFD1bKWLVuW+L7msm9SBq6kfvjhB3zyySd44oknVLk36c+dPn0adevWtejzynZz5szBpEmTVIBHMoJkXiPJZirusaRfZ16/ePFilSUkJc9ljiPZXl5D/jJPRe0ru727Co6utjFx9ZFlpjlMbMWeVaaT5Lbi8+3Xn/fZmoxsVxO25MuV52BL9k7VNjh6K44nWldJyOup6m4T4/RzrLsUA1vyTEvrP4eZm13mtfnPrJ2UuKvasotV9H/MxzF/N24PDwtWcUnV63H35VNW8Zq1YvV7qOrVq2PkyGtzKEjptGrVqt3QfSULaMyYMTf0vB06dFA1nSXtv3379iqdv0mTJhZ/XtGrVy81j9Hu3btVbWypkS3zDBX3WLmfS9orpe+kZrZ0EqT03c2WFSAiIiIiIrKktm3bqsFtZtKvyV0ZoST3bdWqleqX3QgpLy4ZPVKu7dFHH0W7du3g6elp8ecVEpiS+WwDAwPVQD/JUrreY+VeL+2V7bdt26aWf/rpp9cNEBERERER2WQmUWGio6PVXEG5SfqkjKK60e1SU1MLlBUo6rG0eN7OnTvnyUoqrTaUFDOJLIuZRJbFTCLLYiaR5TCTyLKYSVSxMomaPPkbHJwtmEmUlYaAhQ9axWsmKs8kO0YGzeVWq1Yt3HnnnTe8nUzS7JRvwvmiHkuL5w0JCcmTlVRabSgJZhJZHjOJLIuZRJbFTCLLYSaRZTGTqGJlEv3VsB08/puTzRJSDXqMDzptFa9ZKzY51EhGcEkqfW4lqd9c2HZSR9rd3b1Ej6XF80rWUO6spNJqAxERERERka2QwXH5+z0pKSk3tZ2UUM4/NrKox9LieaUsee6spNJqAxERERFRhQkSSSq+1G3Wajutn/fZZ58t9TYQERERVQRGgxEGg9Gij09Eljdw4EB10Wq7sn5eSz0WERERUbmjN8BotOA86gbbmeO6tNiXdQOIiIiIiIiIiIiIiIhIezaZSUREREREVBgp7WTJKTdtcDpPIiIiIiKicsOgN8JgwX6ZoQJWj2AmERERERERERERERERUQXETCIiIiIiKjdkziBLzhvEOYmIiIiIiIjKjlFv4eoRBmYSERERERERERERERERUQXATCIiIiIiKjdU/WgLjvyqiPWpiYiIiIiIrIXMR2TROYmMFa/PxzmJiIiIiIiIiIiIiIiIKiBmEhERERFRuWE06NXFko9PREREREREZUNvNKqLJR+/omEmERERERERERERERERUQXETCIiIiIiKjeYSURERERERFR+6Y2miyUfv6JhJhEREREREREREREREVEFxEwiIiIiIio3jAaDheckMljssYmIiIiIiKh4nJOo9DGTiIiIiIiIiIiIiIiIqAJiJhERERERlRtGvV5dLPn4REREREREVDY4J1HpYyYRERERERERERERERFRBcRMIipURlIM7BxdrP7dyU5Lgi3JTNfBlhgNRtiS5EzbGt2dnGFbn4dsG/s8eLvYzk+ck4crbImztztsiZsDx8RUJEaj3rJzEhlt67eGiKgwd/dpAFcPL5t4c9JT7oEt6eiWAlvy6R2tYEvSsm1rbsDQQY1hSxzjgmArujrZVp9E71IbtuT2plVhS4wOmbAl9tG2812zS0mFtTEYjWpeIks+fkXDsyZEREREREREREREREQVkO0MsyYiIiIiug7JIrJoJpEFH5uIiIiIiIiKp/9vXiJL0VfAfwAziYiIiIiIiIiIiIiIiCogZhIRERERUbnBTCIiIiIiIqLyS+Yj0sNyqUR6zklEREREREREREREREREFQEziYiIiIio3GAmERERERERUfkl8xFZct4gvQXnO7JWnJOIiIiIiIiIiIiIiIioAmImERERERGVG0aDQWUTWfLxiYiIiIiIqGwwk6j0MZOIiIiIiIiIiIiIiIioAmImERERERGVGwbJIrJgJpF6fCIiIiIiIioTeqMRehgt+vgVDTOJiIiIiIiIiIiIiIiIKiAGiYiIiIiIiIiIiIiIiCoglpsjIiIionLDaNCriyUfn4iIiIiIiMqGwQjoLfz4FQ0ziYiIiIiIiIiIiIiIiCogZhIRERERUbnBTCIiIiIiIqLyS280Qg+jRR+/omEmERERERERERERERERUQVk9ZlEV69exaZNm/DYY4/d9H0jIiKwevVqPPnkkzfdDtm+UaNGaNeunabPW9LH0qIN16NPDgd0mbD3bQA7O7s863RxAWqdnYsX7L3r5lmvjw+CnVsl2Lv6QCtGoxHGlDDAzgH2njULrDfE+ZuuuPjA3qOaaVniFUCflWeZVgz6bKRdOQFHzypwrdY477rsTKT47wXs7eFStTFcqtRXy3Wp8UgPPw/nSnVzlmlFn5mK9NAzcK5cD86+tfOs06XFI+3Kcdg5OMG1Rgs4eVdXy7MTI5AZfRku1ZvmLNNKUlw0Lp8+hoatO8C3Wt7PQ3JcDM7s3wFnF1c079wTXpWrquWhARcQGXwZTTt2g3dlbT8P8RFXEXn5Ahq06wo3L9886+LCghFy5hic3T3QpEsfOLt5ICLgPCIDLuTcp0m3fvCsZHodWgjxv4Co0Cto270fXNzcCr3PpZNH4eTigoYt25Z4G0u5ePIoUpIS0bH3ADg4OBR6n+N7d6BmvQaoVb8R/E4fR0iAH2rVb4jWnXto2lbZl+26FAIne3v0blq30PtciozDyatRqFvJCz0b1ynRNpaSpddj9+UwVPd0Q7tahX8GT4fHwC86Aa1rVEajyt5Ycy4QDvb2aFOjMlpUr6Rpe9MMepzLTkNtB2fUdnTJsy7ZoMOxrJSc23UdXNDEya3YbSwtHXpEIBPV4AzPfId4SchGODJzbteAC3zhpK6nQIcYZKEh3FEu6PUw2luwQrWecxIRlba9e/ciIyMDQ4YMuen7Hjp0CHFxcbj99ttvuh1ff/01Jk2aBB8fH02ft6SPpUUbinN27zakJMahcbsuqFavUZ51cnxx6N+/c277VK2Blt37ITUxHpeOHUDVOvVRt3kbaEWXnozESwdg7+AI91rN4VK5ToH7xJ3ZDqNBp677Nu+t/ib47ctZ71qlHjzqtNSszbFx8diz/yA6tGuDhvXr5VmXnp6B5avW5NyeeO94ZGVlYemK1er2+DtGw91d22PmsNBQnDx+FN169ELVagX7Q6dOHMfVkCvo1bcfKlWqXOQyLVwJCsThA/vg4+OLoSNGlWj99baxpDD/cwj3vwDf6rXQpHOvQu8TfOYYUuJj0Lx7fzi5uCLs0jmE/9fnq1qvERq06aRpm3cePIZsnQ5DencrcD7or/VbkZ6ZiYZ1a6Nft445ywOCr+LQybOoUa0KBvfqqllb9Xo9Nu/aB18fb/Ts3KHA+r/XbUJaRgbatmiGTm1b5Sw/dd4Pp85fRMumjdG1vXb7s7S0NGzbsRONGjZEm9bX2pPbgUOHcck/AD26dUGzpk3x6+9/wtHRAd26dFa3tSL7pbX/LFfXh4+5Ex4eniVabzAYcHDvLkSEhaHPwMGoXqPgeTpLCQ0Lx5Fjx9CzezfUqF7wXFR6ejp27t6LqJgYjLtjNK6EXMWR4yfg6eGBgf37onIlbfuofsFXcT7gCgZ17wBvT4886wJDI7DryCn4eHngrsF98qw7fsEfGZlZ6NWhNcoDvYXnJNJXvEQi688kkiDHP//8c0v3dXJyQqVb/NJu27YNFy9e1Px5zWJiYrBs2bIi12vRhuLoIk4jy38Lsq8eQnbgjgLrDSmR0CdeRVbgLmQH7bq2PC0WWYE7YZAAk4YM4UdgCD8O/dUD0EedLbS9huQw6IN3Qh99Ti0zpsWYll3ZDX3MtRPuWojY+CkSTq1DxIa5SLl8MM86owSQQs+oIFLoqhlIDT4KfXoSrv7zJlIu7UHk1q+ReuWEZm2VDmLYqneRdGErwlbPUIGf3AwqgHQWqUFHEPLXq8iMu4LM2GCErnkfyQF7EbFxDrLirmrW3vSUJHzy5N049O8/6m9ibHTe9anJuHT8AI7v+BdznrgLaUmJOLN3Gxa99SyObV2Lb195Qi3TSuzVQPz66iSc3LwSi6c9CH12Vp71KXExCDp1CKe2rFLrDXo9EiJD1TK5bP/lS9gXEfiwhLOH9uKzKY9h6/Jf8emURwu9T0piPH6Z+y4Ob11f4m0sZcs/f+Drd6bhr++/wncfvlHofa5evoRFH7+Dc8cOqdvhVwJxaNtGbP77D2jto/X7MGv9Xry1cgfmbTtSYP3qE364c/5f2HIuEP5R8SXaxpKmrNqFRYfO4rkVO7D+QlCB9Qv2n8aTf23DnqAwhCamIEtvwIErEdh1ORQT/9iELX5XNN2XzUy6gu0ZCZiVdAUB2el51mcYDSoYJJcN6XGIMGRddxtLyoIBKxEBP6RgFSKQCtNJL7N0GBCGDHU5hkR1f7PjSFTLiIjKyvHjx7F///5buq+bmxu8vLxuqR2//fYbkpOTNX9es1OnTmHPnj1FrteiDcUJuXgGB9Yshf8J0zFQbvIb6H/8oLoc3bgKfkf2ITM9DV88fQ+ObV6N5Z++h5M7NkArhqx0pASfRpL/EVz65RWkhhbsv0Xs/EXdRy4GXaa6mG/HHluHjKiCxyqWkpmZiQGjxmHJsn8w9M4JCAgMzrM+MSkJH336FXbtPaAu8n5nZ+vU9Q/nfoG4hARoKTwsDGNHDsNfS//AmGGDkZpybeCMWPbHb5jy7FNY+fdfePi+e1Rws7BlWomNjsbBfXvw1SezSrz+ettYUnxEKC4e2ol9K34pdP2htX9ixWfv4NC6ZfjpDdOA4DO7NuL0zn9x+eRBxFwN1LS9cxf+ipc//AJvf7oA73y2oMD6/cdPY/v+I3jundl4/8vv1bINO/ej7z2TsWbrbpz1C9C0vc+99SFmf/MDJr/yLhYvXVFg/cHjp7B1zwHc/+wrWLbGtN9a+NtyjJ70DDbt3IfLwSGatnfcfQ9i4Y8/Yey9D2DHrt0F1r/57gw89tSzal1YeIRpUOCevdi0dRtuv3McDh05qllb9TodDu3bg/mffoz42NgSr3/piYmYPeNtHNy7G0ka7s8iIiMx6PbR+O3PZeh/20gkJeU9BsjOzsaQkXdg7hdfqfc0IzMTV0PD1PU/li1Hj/5DkJqaqll7JdAz4uk38NOqjRjyxKvqf51bVFwCdh45hfe//TXP8vSMTEz/5Dv8umazZm0l22P1mUSloUqVKrjnnnsqzPOWRRt00efh2vYewNEFGSeWqB1V7tEjzvVNI7OMukxkXlx3bbuIU3CsWTA7y9IMSVfh2OJOSdGBzv9fOFTPOwrEsX6fa4GhWD913aFWZ9OyzGToww4BVbUZVSYBH316IurdPQtZ8aGI2fszPBtfy1ZwcPVEzaEvquvJl/YgKzYETj61YefoCrc6beGUEoukc1vgUf/aiB1Lyoy6BOcqDVBjyIsqEJTktxPVcmU/SWZTjSEvqOtxR/5CdkI47Byd4eRZDe512iHT2QPJfjtRpeeDmrT39N5t6DVqPEY8+iK2L12sOq/9756Ys756vUaY9PYn6vo/X89EiN9ZxEZcRZ0mLdG8S29cOLQbJ3dtQq/R2nzXz+xYj8GPvoy2g0ZhwzcfIujUYZUxZFa/bRd1EUvffx7JsZFo2XuouqQlxmPDtzPh7q3dSJddq5fh+dnfoEmbjpjz/EREXQ1G9boN8txn/a8LcfsDTyAs8FKJt7GUrSuX4qOf/4GXb2W8NG4IdNnZcHRyytvePxbj9gkP59yWDn2TVu3w9w/zobWNZy9j28sPIlOnx5h5y/D84Lwj8OZsPICfHh2N7o1ql3gbS0lMz0RYUipWPjIawfFJeHfjAYxs2TBnfbbegPn7TmHT5LtQy/vaiKhPx/RTfyWo5BeTgKHNtcmMDNRnoJaDM573qoOTWSnYnZmoMoXMqjk44xkv0/s6NykE3Zy9rruNJV1BOprCA93gizNIxmWkoR2882QOycUII9YjCtXhkrNdHbipTKLywmjUq993iz4+EVmdwio+lOfnLYs23P7Yi/CpVqPQdfb29njwrbnq+j9ffIDuI+9GVnoa7Ozt0axLLyTHx+LguuXoMNCy2U5mzj7V0WDMNHU9+sgaZMSGFMgKcvTwgU/TbnCr2RROnqasFvM2gX/PhG/r/tDK1l17MPK2wZjzwTv49c+/8NeqNXhtyvN57iPZRSOHDUbv7t3U++3p6YFFX3+K+x57Glpbt3olnn1pKh5+7AnMmfk/bN28CXeMHZez/uqVK2jfqRM6du6KjevWYteObYUuG3b7SE3a26lbd3UZP2JIiddfbxtLatP3NtRv3RErP3+30PUBxw/iwffnqYy+eU+PyzlR3Kh9N9Rp3gZNOpvOv2hl2drN2L18IRwdHNFlzER8+PIzedZ/+tYU9Tc4NBzTP/pKXf9w3o/4Yc47GDlQ27bKSX8JAh3b8BeiY+Mw4elpePTesXnuM+ftV9TfE2cvYN7i3zFhzO14/7P52Lp0MVo2zZtFaWkBgYFwdXXBymV/qmDPgu9/wMD+pv6RiIuPx0+//oazxw+jku+1KiOLvp2n/v6zajW279yF7l1N5wgszc3dHR9/9S2mPfVoidefPXUC/n4XsWrbPjg7O0NLK1evxZTnn8GzTz6BD2fPxfqNm3DfPXfnrF+1dj0qVfLFuv+yn8SwoYPVRUx46GHEJyTCwyNvRo+lLFmzBfPfehG39+2Gh9+ag8NnLqJ7u2u/bT3atVSXPhNfyrPd17+vxLP33YEtB46hvDBlEllyTiJUOFafSWS2a9cufPDBB9i5c+cN31eybBYuXHjDz3ny5En873//w4oVBUcWWOp5pWzcjz/+qDKX5LF27y44SqA02yAjlpKSkvJcbopBDzsnVxUYkpJy0BU+eloXfhxO9Xqarsf6w6FyE1XyTUtSUsDOwQV2dvaq5Jkd7Aq/n9EAQ9wl2Fdvl6cMnT7yFOx9r53YtDQpG+fsayqP4FypjirXVhgpO5cadBTebW6Ds28tVOk+AVlxIXCp2hC61FgN2xsHp5z21oU+pfDn1qUlIDsxHB4Nu8Kjfid4Nu+nMopcqjVRj6GVxJhI1GjQRF2v2bAJEqIjCtwnKiQQu1f8psrL1WveBn3uuA8N23RCRNAl1G/ZrtBtLCUlNgpV/ivvUbVeYxUEyi/80lkcXPmL+j56VbnWkT+5ZSXaDNCmI2YWHx2J2g1N6e21GzVFXFTe9+qK33lUqVkHPlWqlngbS8rOzIR3pSrqvatRpx6SEvJ+Fg9t34huA4fBSeOD18Jk6/XwcnWGo4M9PFyc4Ghf8Cc9NTMbni7OWHvKHymZWSXaxlKiUtPRuLIpaNGgkjfi0q6VPhORKWloUa0SQpNSsPVSiAoamWVk69SyBzu10Ky98XqdKhkn6ji4IP6/cjT5hekyUcneEc529iXexhLSoEel/8rHVYIjUotIvjcFhVzVdTmoDkE6GpeXMnNEZPOWLl2q+jJFVXAo7r5Scm3DhhvPVNm4caPq8x04cOCm2ngzzytl47Zs2aIyl2bOnIlLl0wDZUrqRtpQav29GziWirpyGTUbNlVlmu96/k1EBPqjTrPWiI8Mg9Z0GSlIC7uISq0HFFjnVqMJEi7ug/+SV5F69XzO8qykaNg7OcPBRbvfx/CIKLRoZjr+bdm8KULD8x7/urm5om7tWli+ci16DxuD4BDtKi8UJjIiHE2bNVfXm7VogYjwvP/bp59/EY2bNMVlf38VGIoIDy90Gd2c1n2GYOfv32Hnn9+r0nLSd6ndrDWirgRg0+IvsWzWdE3fWinR7ebqCicnR3h7eaost8J8s+QvvPW8aYqE8KhoNKxTC3//uw0J+bI3LCkmLgGN6tVR71n1qlVU+a3CytH9tGwlvlj0C+66fTASk5JRrUplpKSlYfWm7cjIyNuPsaTw8Ai0aNZMXW/ZvDnC8n1vgoKD0bd3L5w4eQqbt21XbRcJCYlY/MsSLPvrH9w2ZBCsWfDlAAwYMhx7dmzF0YMlyyouLWERkWies+8t+P76B1zGyOHDsGL1Gpw4dTrPumMnTqJ506aoWyfvFAuWFB4dixaNTOVIWzWqh7Do65/rCw6LVOXbG9fVrp1km2wiSHTkyBH88MMPamcnWTF+fn43dN/rlWorzLlz5zB06FCkpKSoA+6//vrrhtt4M88rQZ033ngDixYtgk6nw3333YfNm0uWDngzbZg1a5aquW2+1KuXt/bxjTCaR+1KeR2HvPMwyMiWrKA9sPeqBQfv2ioAIyXqjFkpMKREwJAUCmO2NunmdvaOqkSbqV0GGHOV3Ln2WnTQh+yHfZUWpqBXrjJ00GfAqNMuNd7exUNlEgldeiLsnQuOSpflUTsWoGqvh+DgYhrB4NmkF6oPfAqws4ODm2/Ztfe/9uSWlRCG2P2/olq/J2BnbwoS+rS+DdUHPAV9RjIc3LWbn8rd01uNbBRJcTFw87o28t4sMSYKAScPqzJtyfExcHRyxtAHJmPCtPcRfTUY3v/NU6QFF08vpP0XuEhJiIWrR8HSIvHhIQg9f1LVYM9IvXYiwO/ANjTrrt2ISOHh5a3eVyF/5XZuKxZ+DkdHR5zevwtXLp1X8xBdbxtLktGu2Vmmg/6E2LzPLfvWdb8vRlx0pJq36PyxQwWCSFpycnBAWpZpX2YwGFUAKD8JCM3esB9/HjqL8d/+XaJtLMXbxRkxqRk5WUXuTnmTmb2cnRCbmo7v9p/BD4fO4qVVpsEO8WkZeOPffXh5QCf4uGk3x4+HvQMS//tdSzTo4G5X+CHTtswE9HfxuaFtLMEZ9mpOInNpOZciDvEuIhXNYNovn0Wy2u4iUpABA8Kh3W+bJRkNBnVMYrlLweMGIrp1ixcvVoGahIQE1Q+T0d43ct/rlWorjARpnn76aTUXwty5c+Hv/9+8pBZ+Xikb98QTT6jgTmxsLPr27YuwsJIFUG60DaXZ3yuJkzv+Rdu+Q3Nut+41EPe88gGcXd3gWakKtCTBntBNC1Bn6JOwdyw4wKfeiBfQ8K7XUG/kS2r+IrPYExtRqa222SMyN0p0jOn4NzI6Rt3OzcfbW2UN/fHDt3jlxWewbdeNfeZKm7ePD2KiTWW6o6OiCszj5e7hgRenTccHs2bjwrmzqFatWqHL6OZkpqUgLSkBIedOqLlz9bpstO0/HPe8NhvPzluGyKAbCzzfKp1On5PNlJqWpoJFuUnQ6KX3P8EdQ/ujQytTwKNa5Up49/MF+G3VBgy496kCZbMsxcfbE1Expj6cBHvyz58k5Ddh54HDSEpOUff1cHdDZmYWZn29EN8tWYq7n8ybpWFJvj4+iIo27RuioqPU7dy8vbxwOTAI3y5chFlzP8XUV1/P+T/s3rtPzQEUFZW3pL618fTyxrFD+7Fy6e/48K3pWLzAlAWl2b435/2NLrAv8/b2wso167D8n5V46LHJWLveNCBj646dWPrX3/jg3begJR8vT1VSTkTGJsDX6/oZTDO++QVenu5Yu3M//IKu4kKgdiXcLZ5JZOFLRWMT5eYaN26Mn3/+WV2XWo8SDGnevHmJ79u+ffsbfk4JCk2bNk0FbMSECRNuuI0387yibt26+P3333NKB/z555+YOnXqdbe7mTbI65PXaSYjy26m42DvUxdZAVth5+yhSs7JiX99cjjsXXxg5+yOLP9NcvoVxqxUUwaRbwN1X5mnyJieAKO9PYyGLNj9N7rZ0uycPaEPPQyjPhP27qaDUwlU2XnWUEEkvf8Gdd2YFqsCQvZetVUWkb1HdRizU2FMuAJU02aiQievqiqzJu7IcmREXoJHA1OasMxNJGXn9JmpCP3nbXi1HIi0kJNwrlwXrjWaIyVgP3QpsYg7tgI1Bj0LrbhWb46o7d/A0bMqUi/vh2+nseqALy3oMDwadUdWYjjC1v4Pvh3uQMrl/aqtknGUdGEb9GkJSDi5CnXGztSsvTKh7tcvTVRz90i20OMfzlPzFIX4nUPzzj0RGRyAmLAQNOvUA8d3bFDLZSJeuR4dEogz+7Zh7Atvatbepl37Y8uiOYgJCcDpLavQY95EJEVHIDUhFrWatUGY3xlkZ6SjQftuOL7hbyREhKrycsGnD6NW09YqcKSl9n0G4adZb6F19z4IOn8GdRo3R+hlPzg4OqFm/UaoVKMWLhw/iIToKDUfVFJcbKHbaKVd996Y987LqF6nHlzd3OHi6ga/U8dQrXZdeHr7oHL1mjhzeD/Cgi+rMh/pqalIjI3BwW0bEXElELvXr0S/kXdp1t5GVX3xzqqdSMvMRqd6psk9d1wMVuXl3J2d0KVBTXUfZ0cHLDlwpshttFDDyx3x6Zn4cvcJnI2IxcAmpozDTReDMaxFAxUAalLFBz0b1ERGtl7NQ5SSmY17fv0Xd7VtjL2B4Wha1Red6mhzQqGxoysWJIdhbVosjmQlY4SbqQzNkcxkdP1v8IDOaMSl7HQ85FGj2G20UBeuWANTZuF5pGAwqqp5h6KRqcrJCZmnSIZGePx3+CcBolhkIQU6dd94ZKOWRr/DRET53Xnnnfjss8/UdQmehISEqP5NSe97MyRIJFUcBg0apE4KFvV8pf28YtSoUfjyyy/VdTlJuW7dukJPVt5qG0qrvycuHTuAwNOmUjVVatdTpZfP7NmCNn2G5LT98IYVmPS+6XWJs3u3IS7iKrb/+QNGPfkytJKdHAv/395A1S6jkeh/EO41m8KteiMk+u2HT/Ne0KUnq8CQUa9D/JltqNp5VM4gwuTAY6jZ7wFoqX/vnnjzg1mqZNEvfyzHF7M+QHp6BvYcOITbBvVXmUM79+5HWlo6Fv3yOxZ8Nltt98dfK3Dlaij+Xr0Otw8dhBZNTRUSLG3QkNvw4tOTER0ViZ8WLcTy1etV0Cgo8DK6du+B2JhobNm0EefOnMHZ06cxcMhthS7TiswvtGPrJsTFxuDvP3/D7WPuRExUJLIys9CsZatC12ekpRdY5uHhqUl7Y0KDcfHADiREhuPYphXoPGwswi6dg7u3L3xr1MbJbevQuGMPuHn5YP+KJUhNjEd4wAU1P63MR+Tsok25Y7PObVvguXfnwNXFBc0bmcqE7zx4TAWEfL29MP7Z11CvVg34B4UgISkFowb1Qe8u7VHZ1wc+Xh447x+oAk35g0uW4O7mBjdXF7z3yTxcvnIVg/uYSvmv27oTtw/sq85d/L5yPQb07IagkFCs3rQNj903Dp3btUbX9m3Ub8XydXJ+SxutWrbA8ZMn8ckXX6mycePHmvqaa9atx5hRI9GkcWPUqFFdlaCLiY3F6TNnVQm6rdt3oH/fPjhy7Bi2bNuB24dp931b889yXL0SjI1rV2HQbbejbv0Gaq6hfoOHFrq+c4+eqt/ds29/VK5SFVER2mUZ3jZ4ECZNflq9Z/MWLMS/q/5WAXvJIOrVozuGDRmMdRs2YmC/vur+4RER2Lh5K15/dwamPP8slvy5DCOGDUW1qtoMHL69Tze8+ulC3DWkD9btOoCZLz6Ky1fDkZqegXbNGiE6LgHrdx9CTHwifl61CeOG9kW9mtVw4OR5xCclIzw6DmFRsWjZSJsS7mRbbCKTqHLlaydZXF1dVSS8NO5bHBnZLqn5ZtebVLG0nlfk3laeV1J3S+Jm2uDi4gJvb+88l5vhVL8X7FW2ihEuzUx1po3p8SoII+ycTNFtCQpJNo6UeXNpNkxdZFunBn1h76JdtoBD/b5SOFsFi+zrdDe1NyNOhh+b7uDiozKbJMtJXodQGU+pkeo+Dg2u1YDVQq0Rr0GfkQq3Ou3g23GMWpYRZRrpKFlRLjWaqfmK0kLPICvBVJpAAkqyrMbg5+DRUJv6s8LeyQW1hk+HLiUa3q2HwbORvL9GZESbJqM06rLgVqs1MqP8kR56Nqe0nOl6PGqPfgculbX7wapSqy7ue/VDxIRewbjn31RzDWVlZCA80C8ni+jS8QO44ncWnYeMQpeho5GdmaGW6fU6TFuwHJ4+2s3x07BDd/S+5wkV/Bn72ly4efkiLTlBZQ+J+PArCDp1CFFBl9D3/qdQu3lbtTwpJhKdR94LrQ286z50HXy7CgBN+WyRysaKiwxHQrTpZPak6e/jqfc/x0OvvIexT05F03adCt1GK/c/Px31m7WATpeNaXNMcwyFBQciNSkRTs4ueGnmF+ry0EuvY+LUN1VJupiIMIQGBaig1/kTh6Glz+8dCncnJ9T08cT/7jKVUTkXHoMsnSmj5MO7BqrrV+OS8N3EEUVuo5Vv7x6E5MwsdKtfA0/2NH02T4abRm6Jj0b2RnRKOpIyM/HpHf2QpdejQ+2qCIxLwoErEeqvVqR83FTvuog1ZGOQqy96/PcbFZCrnGqEPgvD3SpddxsteMIR/VEFSdChO3xRFc6qnFwcro3ET4AuzzxFLeGJQaiqLm3hhdaw/KTnWrBsFpHpQkTlr88nGcPFZS+V5vOK/H1NS/X5Squ/JyICL5kqMRgNCL9sOla+cuF0TgZARloKWnTrAw/va1UMQi6eQXjgJdz1wlvoctsd0IpBlwWPuq2RHnkZKcGnkZUYpZanhZnabchKV8vTI/xRpdMIVGozUC3PSohE1U4jVWlyLVWrWgWL53+OwOAreHPaC+jTsxuysrNw6uw5tT4mNg679h7AuYt+mPnO6+jepZNavufAYbRu0Rynz55HeETBMtSW0rZ9B3WSNDgoEHO/nIc6desiKSkRlwNMfdTk5GTs37sbrm6uWLVhC9zc3ApdppXk5CQc3LcHXXv2Vn8zMzJVBlR4WGiR6wtbppWUuGiE+Z9DzSYtEHjS1L+ICw9RwSBx9ysz1ZxfUcEBGPnM6/CuUh3h/ucRePKQGhT4wAzTvD9a+fydaahZrQo83F3x3UemgdYS+ElNMx03N6xbC2kZGSpwdPysqUzmrOnPwdHBAQHBV/HzpzM0CRCZ/frVbGRlZ6Ndy2Z4b6ppQO2x0+dUAEivN2UR7T18HA4O9ljw8Qy1/usP30J8YhIiomPxy5ezNGur/DYs/fVnREVF4a4xo/DIRNPczUeOn1B/JUA//4vPcCUkBHqDAV9//glSUlKxa89eHDx8GI0aNsB7b5myi7Ry7OB+NG7WApcunEdkRBiyddk4f/ZUkes9Pb3wxgcfITDAHzXr1MFzr2jX3nZt2+CjGe+qbKxvvvwMDevXR1JyMvz8TeevmjZpjFemvIhzFy6iT6+eeHTSQ6rqU5dOHVWmlrzPcn+t3Dm4N564ewRCwqOw9NN34O7misjYeHVbJKWmYeeRU+jTqa36m56ZhQ+efwQ//u8VfPn6c3jt8fswuIfp98PW6Y1Gi18qGjujVjmdN0myYWbMmIG1a9eq22+//TaaNm2KRx55pMT37dq1K6ZMmaLqPpdUYGAgevTogZEjR6rRVqdPn1ap+uPHj7fo88pjjRs3Dm3atEH9+vWxcuVKrFq1Cp6ensU+Vmm1QV6rpFe6dX8ado7alfW5Wdlp2p0wLA0N+2rXMSoNRoNV7x4KmPqQbf3YRSTZVmmnFtW1GTlXWjydtZ3v7Fb03jQHtiQ91lRS0lbs+65k80xYix0xabAVkoG0GCFITEy8pROPpcF8DOM94BWLHsMYdZlI2vmJVbxmovJi3rx5qnya9GGElFGTZS1btizxfaXcWlBQED788MMSP++aNWvw5JNP4q677kJAQIDq8x0+fFhVdrDk88p2X3/9NTp16qROAMq8sjIf7j///FPsY5VGG8z7yo83nii0bLE1+mn99eeosiYHp2hTcaK0xDhql/lcGtKybavs619ntQvelYbp9cuunPaNMjrZ1ryaeh/bmhMmKMM036mtaOiiXSC3NDgEH4etSEpJRZW+46yi/2M+jnnDpRFcLTigI8NowKzMQKt4zVqx+nJzkgr/+OOP59weMWIEfH19b+i+NWvWxFNPPXVDz9uoUSPVQdi0aZMK2MhIrzp16lj8eYWUiPvoo49UremXXnoJrVu3VrWqi3us0m4DERERkS1ScwZZcN4gzklEVPpkTp7clRuk/1K9evUbuq8M8GvVqtUNPe+YMWNUmTcZcCflveVvUX3N0nxeMXHiRPTq1UsNTpTycZIhdL3HKu02EBEREdkiGdJuySEDRlQ8Vp9JVJirV6/iiy++yLNMRmDNnj37hrdLS0uDu7t7iR5Li+e955578mQElVYbSoqZRJbFTCLLYiaRZTGTyHKYSWRZzCSqWJlEXv2mWTyTKHn3Z1bxmonKs71792LFihUFBvI999xzN7ydlBCSuQRL8lhaPK90wXNnBJVWG0qCmUSWx0wiy2ImkWUxk8hymElkWcwkqliZRK+7NIKLBTOJMo0GfMxMIuvnJHMp1Mw72Xb+g++SbpeamqomhyzJY2nxvPkzgkqrDUREREQVgZozyILzBnFOIiJtSF8pf7+nSpUqN7WdzB8g8w+V5LG0eF4po1fcnLc32wYiIiKiikDNG2TBfB+97eXUlP9yc4WpUaMGXnnlFc220/p5x44dW+ptICIiIiIishUdO3ZUF622K+vntdRjERERERGVyyAREREREVFhmElERERERERUfumNgN7Cj1/RsE4ZERERERERERERERFRBcRMIiIiIiIqNwwGPew4JxEREREREVG5xDmJSh8ziYiIiIiIiIiIiIiIiCogZhIRERERUblh1BsAO71lH5+IiIiIiIjKBOckKn3MJCIiIiIiIiIiIiIiIqqAmElEREREROWG0agHLDknkTw+ERERERERlQnOSVT6mElERERERERERERERERUATGTiIiIiIjKDaNkEVlyTiILZikRERERERFR8QxGQG/hx69omElERERERERERERERERUATGTiIiIiIjKDWYSERERERERlV+ck6j0MZOIiIiIiIiIiIiIiIioAmImERERERGVG8wkIiIiIiIiKr9kPiK9BecN0qPiYZCI8jAaTd8woz7LJt4ZW2mnmT4rDTbFxmZqS09Nhi3JTM2ELUlLsa3Pg72zA2xFcoZtfRbSM21r35tmsK1DvCwYYGttNR8/WAV9NizaGn22JR+diMiizPvrjNQUm3mn9Zm21YdKSratPkmyg22dFkrT2c5xkq1910RScipshdHJtj4Lejvb2jekZNrWviHJxvqoDim2811LSk2zuj6fpfvMWTbUJy8tdkZr+g9Tmbt69Srq1atX1s0gIiIiGxISEoK6deuWaRsyMjLQqFEjREREWPy5atasicDAQLi6ulr8uYiIShP7e0RERHQz2Ocr3xgkojwMBgPCwsLg5eUFOzu7Unt3kpKSVPBJdije3t5W/a7bUlsF28v3l58Hft+4b+C+t6x+K2SsUXJyMmrXrg17+7Kf6lICRVlZlh9F6OzszAAREdkk9vdM2IeyLL6/fH/5WdAGv2t8f7X4PLDPVzHYVu4gWZyc4LHkSGDZSdlC4MXW2irYXr6//Dzw+8Z9A/e9ZfFb4ePjA2shmT3M7iEiKhr7e3mxD2VZfH/5/vKzoA1+1/j+WvrzwD5f+Vf2Qz6JiIiIiIiIiIiIiIhIcwwSERERERERERERERERVUAMEpEmXFxc8N5776m/1s6W2irYXr6//Dzw+8Z9A/e95e23goiIbIut/c6wvXx/+Xngd437Bu57+VtBdI2dUWafIiIiIiIiIiIiIiIiogqFmUREREREREREREREREQVEINEREREREREREREREREFRCDRERERERERERERERERBWQY1k3gMq/oKAgbNy4Ee7u7hg7diw8PT3Luknlyv79+3HkyBE0adIEI0eOLOvmlCt6vR6rVq1CWFgY+vXrhw4dOpR1k8qV2NhYrF69Wr3PY8aMQY0aNcq6SeXKuXPnsH37dlSrVk3te52cnMq6SeXK1q1b1Xvcrl07DBw4sKybQ0REVKaOHj2q+iX16tVTx3X29hyPWlpkGum1a9ciODgYPXr0QLdu3UrtsQlISkrCihUrkJmZqfrTdevW5dtSii5duoQtW7bAx8cH48aNg6urK9/fUrRz506cOnUKrVq1wtChQ/nelqKsrCysXLkS0dHRGDRoEFq3bs33l8o1HrmRRZ05c0YdxMqJtPXr16N79+7IyMiwqnd96tSpqp3m6/LDmvtiXmeNvvnmGzz00EMqiPH+++/jmWeegTWR/7V0Es3X87+35nXWSg5i582bh6tXr2LUqFGq82BNdu/ejQ8++CDnev7317zOWgNEEnSTdkuQs2PHjupzbE0+++wz/PvvvznX87+/5nXWaNOmTRg8eLA6mbB48WLcfvvt6gSDNZF9V2RkZM71/O+veZ01euutt9TvhXxmn3vuOcycORPWJCQkBI899ljO9fzvrXkdERFRafjtt99w1113qWPmTz/9FA8++KDVvbHDhw+HwWDIuZ7/t9G8zho98sgj+PjjjxEaGop77rkHP/30E6zJ8ePH8eqrr+Zcz//emtdZo5SUFHTu3BmbN2/G6dOn1XV/f39Yk4ULF2L58uU51/O/v+Z11ujAgQPo3bu3ek+lnf3794dOp4M1mTx5shrYbL6e//01r7NGs2bNwtNPP636JK+88greeOMNWJOYmBjcf//9Odfzv7fmddZI+s7Dhg1TfWnpU0vfWgYJWhM5HyHnKczX87+/5nVEJWVntLazRlSuTJ8+XWUOvffee+q2RN/lIHHEiBGwFoGBgSqDQTKd5Hpqamqe9Y0bN1brrJGMYF+0aJEaUZacnIyaNWsiISHBajIGpLN18eJFNapFrkuwMDcHBwe1zhpJJ0w6CdLZlfdTMookKCdZcdY06i0xMVGN2JTrV65cybNeRmvJOmv0ww8/qBFlf/zxh7otAc7mzZurE+/WQg623dzcUKlSJXU9Li4uz/o6deqoddZITiCMHj0aDz/8sPruNWvWTI1Atabvm5+fHxo1aqS+X3JdRmrl1qJFC6vZl+VXuXJldSJBPgMBAQGqwyv7DGsh76V0ZuT/Ltfl/c3NxcVFrSMiIioNknH/9ttvq+CLZGNIn+Ty5ctWdZx09uxZtGnTJud6/tMgbdu2hTWSvqm8nzJ4RvqkkjUg/emDBw/CmtooI+0bNmyorkufOjc5HyDrrNHff/+t+tPmwV9ykt3Z2VkNwLQWERER6pi4SpUq6rqcbM+tVq1aap01kqCLfLdeeukldbtTp06qT92rVy9YCwlgSZ9Zjo/lev5BzXLMLOuskbRb+tTSb5LPRsuWLdX5IGshAUHpK0n75PqFCxfyrJfPtayzRvI7cccdd6hMOMmM/fnnn9XA96VLl8JaxMfHIz09HbVr11bX8/dHpc8q64hKiuXmyKLk5KScZDWTH1cpLWVN5CRl7utywC0n1MztrF69utUGiXK/v/IDa2dnp9ptLSdW5cfUfFJarksQQA5ipBMhHTNppzWdtM7/3jo6OqqLtX52vb291cV8XT6ru3btUiPihHTGrDVIZAv7htwHVOYDrxMnTiA7OzunzdZ08qOo91e+e9LZtbb3V/YHuTtf8tmVwIZ5JK98fq1lX5afvJfmUhnW+NmV/7c5CCTX5bdN9r3yGTYHkBkkIiIiSxx3yLGzDASztt9Gc4BIyIlUGREeHh6eEyyS9dKXsjbyPsqxnPmYyBqPOzw8PNTFfF2Om3fs2KEGsQm5ba1BosL6JNaW6SJBwtzX5X2VSgzmAVZSucVag0S20Odr2rRpnut79uxRQW5zn0TKD1prkCj3+yttlNuyT7OWfZn8HpiDQHJd3l/J2pPBl9JO2V9Ya5BI3kt5T82lU63xsyvnIsznI+SvBIzk85uWlpbT32aQiG4Eg0RkURMmTFDRd9nByugnySQZMGCA1b7rc+fOVemkd955p+rcCGsrj5fbAw88oMoPyGXDhg2qJJo11/iVDDL5cZUMHWGtB1tCgiuSRSbvsWRqzZ8/H2+++SaslXy/2rdvr+aeqVq1qlpmzZ8F+axKyS4ZdSqdXilTIuUIrJXMnSTZTlI+RU66mzP5rJWk7kvJAcmEkxrVEui21oCsmDJliprH4Lbbbsvp1FhbBz032S+MHz9efd9kNJk1ltUxk86MlHqVUZzmwJDMU0VERFSav4tS8uipp55Sgz66dOmSczxqrX1UGW2fO5vBmk6s5iYDwSRjWbLEpSqHZL1Yc9lYCWDIMbKUOpYMF2s3ZMgQleUi2Vm+vr6qzycDa6yVZJLJ53fixIk5fT1rHvhz3333qXMV8rmQjBIZsGTNc2pJ/1SyReS8hTk4kL/agbXte6Xfd++996rS+HLdGvdjZn379kX9+vVz5vaRgWvWStoo/X7Z30o/6ssvv8QXX3wBayVlEbt27aq+c7IvE+a/RCXFcnNkcVKXWEp1ySiBSZMmWfXk9HIwIAeJclBrK+RgQEYSNWnSRB0sWuvIewkUyolqyXIxZ+dYOxmBITW/pdSYBDflBLa1kjJ4n3/+uQoW2gopRSHl5uQkthzY5s4ssTZSBq9BgwYqmGErpBMpnVwJCDz66KPw8vKCtZKTCfJZsNZSL4Xtz37//XecP39etV0+v9baIZOyAxLoloAhERGRpciJ1b1796oTgFLu1poHK8mIazkOtZUTaFLC75dfflEnAXv27GnV87pKVQ7pT1vz4K/CjpWWLFmiBofKnLTWPBBMyuBJQHPGjBmwFfJZkHJ+EhCQgJGUwLJWEjiWPrV8z2yBfBaWLVumBgVKhqQEjcyDna2NlMmXjELJIrLWflN+EtyU80FSCUfOBUnA3lrJwEU5N/jnn3+WdVPIhjFIRJSLTLQq2S3PP/883xcLkB/Wb7/9Nk9KN5UOOXCRrD1JL7bWA0NbJvP5SCDu66+/LuumlEsShJNArEx6TaVPRpVt2rTJqjvlREREWpGsnNdee039PlLpkjmJ5JhOAobWXDXCVkmmnsztKvOjUOl75513VPmzhx56iG+vBfTu3VvNA2YLWYa2Rsq2y8BQKaVqK0E4sj4MEpFFSrbJydSiSCr3sGHDrOadlw6CTEqXe/S9jBwxH9TOmTMnJx3WGkh6uYy+KMry5cutZp4UGY0lJZlyj8SQyQolxVx+uKR+rrTXWkgKvJTJKIoEtxYsWABrIZ2vWbNm5dyWiTblfZWsMnM69+uvvw5rIaNapERGUSQ9/vHHH4e1kIBQ7n2ZlEOTevXmeaBefPFFq9qXvfzyyzh58mSxQfAOHTrAWsj/WsokCskmk8+zdBzMmYY//vijmmfLWjKHivtfS2C2uN89rUnWkJT9MZP3WSazNf//pZymBOyJiIhulpTguXLlSpHrpYS3Nc2NKeXEzfNJSLUAmWdS+nzmklJSWth8vazJ73ZxpWxlXhrJfLEWcvwpZbpyn6yUuTEks0F07NgRH374Iaxp8FdxZaMkU0uyoayFHBP/888/ObelioiURTcP/pFjaimBbC3ee+89dVxflA8++EAd81uL5557Tn1mzcf8u3fvVhlF5hLj33zzjcqQtBYjR44stgSeZG1ZS3WZ2NhYlVma+7a81506dVLnLaQsqWTqWNO+TPrURZG+lPSprYUMAvzqq69ybkuVC6ngZP68Dh8+HC+88EIZtpBsDYNEZJF0bfPk2IWRif+sKbX/4sWLSE5OLnK9HNx6enrCmtqbnZ1dbHutpZybHGQdO3as2BOrcoBgLSSoJYGWokhQyxyAsQbyPZPAVlGk4yAdCGsRExOjTlQXRcqiWVM5SjmAlQytoki6vDXV25eyKTJ6syiNGjXKmVTYGpw+fVqVTymKzLFl7pxZgzNnzhS5Tjo5uSfEtoZ9WXHtlX2ZNbWXiIhsz6VLl4r9HZcywtb0O3706FFVmqko1pRVJH096fMVRd5XayrTLH3p4torA6ysqb3Sh5JzFsX1oaxpsncZ/FNcH0qCsdbUh5LgsQwOLYqU8LamMtgyYFiCmkWRctjWVD5T5vmW8yxFkWN8a8kkkX1ZcYMYZV8mfT5rIX1p6VMXRfrS0qe2pvMrUoa0uPMr8n0jKikGiahMyGgXSUOXOYpswcCBAzFz5kz06dMHtkACcdIRsqaDxeJ+iOvUqaMmj7UFcpAjn18ZwWULZNSZ1KeViy2QzD2pVyzfN1sgmWcy34s1TyCcf3Led999V+1/bYFk70nKvC0c3MroZOk4SIDGFsjJnLvvvrvYYBIREdGtkJOrUlpIyjfZAjkRLH0TWyjdLIOZ5LiuuAFu1mTbtm2YPXu2VWVeF2fhwoWq3zd//nzYAqmOIn3/adOmwRbIMahkmEi5dFsg2XC//fabzQyykkHOkrVjCyUfJWAr59mKC3ZYE6k8JNlw0ke1BZKJKHPFfffdd2XdFLIB1pFPTRWOnEwrbvSDtdHpdMWOPLM2ttReaae011bYWnvle2YurWELbK29trYvk/bayr5BcF9mOba2LyMiIttjS7/jttZeW/sdt7VjfFtrry32SWypvba0b7C19travkzaa0v7Blvbl1HZYpCIiIiIiIiIiIiIiIioAmKQiIiIiIiIiIiIiIiIqAJikIiIiIiIiIiIiIiIiKgCsjPaSqFKKldkUjovLy9UqVIFtjLBd506ddQEgLZAJtmUSRUdHR1hCzVSpb2dOnWCLUhPT0dgYCBat24NWxAfH68ujRs3hi2IiIhQNXPl+2YLZNJgDw8PVK1aFbayL6tdu7ba/9qC06dPq8munZ2dYQuOHj2KLl26wBZkZGQgICDAZibgJSIi23P27Fk0adIErq6usAW29DuelZWljuvatWsHW5CcnIzw8HA0b94ctiA6Olr1++rXrw9bcPXqVTg5OaFGjRqwBXIMKueCfH19YQtsbV927NgxdX7Fzs4O1i47Oxvnz59H+/btYSv7srCwMNVHtQUxMTFITU1FgwYNyropZAMYJKJSN3fuXGzcuLHI9a+++iqGDRtmNe/81KlT1YnIonzxxRdo27YtrMWECRMQFxdX5Prly5ejUqVKsJaTkKNHjy5yvZubG9asWQNrOlh96qmnilzftGlTLFiwANZi9+7deP/994tc379/f7z77ruwFn/++ScWLVpU5Pr7778fjz/+OKzFZ599hvXr1xe5/uWXX8aIESNgLaQ9EnAtyqeffooOHTrAWjz00EMqKFiU3377zWo6uhLMLu53y8HBodjfPa2FhITg0UcfLXK9nPD48ccfNW0TERGVL4899hiuXLlS5PrFixejXr16sBbDhw8vdvLuTZs2wd7eOgqtREZG4sEHHyxyfc2aNbFkyRJYi+PHj2P69OlFru/cuTPmzJkDa7F27VrVxy/KmDFj8NJLL8FaLFy4EMuWLStyvfRf77nnHliL9957D3v37i1y/QcffIDevXvDWkyePFkNAi2K9F8bNmwIazFy5EgVJC7Kv//+qwKG1hKguO+++4pcX61aNfzxxx+wFtKXlj51UaQvLX1qayH/6+LaI5+VadOmadomsm0MElGpCw0NVZkLRalbt65VjRiRAwKJrBdFMjDc3d1hLWTEmIy2KErLli2tJoNITqyeO3eu2BOrrVq1grWQoJa/v3+xQS0ZQWQtkpKSiu2c+/j4WFXnXA4SiwsKyEGitQQFhIwQKi4gK9lO1hKQLcm+rFGjRirryVr4+fkV28GR0VnW0sERZ86cKXKdjNKzpowceV/l/S2Ki4sLmjVrpmmbiIiofLl06RIyMzOLXC8ZI9aUCSyZAMUVUbGmQYHS15M+X1HkfbWmjBw5/izuJLtU47Cmk+xyrkLOWRSlcuXKKvPeWkj/SfpRRalVq5ZVVWiR/qn0U4siGQ3WVNVA+v9yHqAocswsx87WQs6vyHmWokifxFoyiHQ6HS5cuFDkeunrWVNGzvX2ZdKXlj61tbC1fRlZPwaJyOJWrlypLhJoefrpp606jVR+bH/66Sds375dnbCWEUTWnJYpHbNvvvkGR44cUcGsKVOmWNUBYn6JiYlq1JacvJRRGC+88IIKvFgr+cH98ssvVbCgX79+eOKJJ1Rgy1pJmvb8+fNVCvSoUaNU1pk127dvn/q+yajOBx54AEOGDIE1W716NVasWKHKDMiIvY4dO8JayUmQn3/+GVu3blWl8GRfZk2d88KCGpKld/DgQbXPlQxP2QdbK/mOyb5BvnNyUunFF1+0qgBcYScXZN8rGUYyclM+v9YymICIiGyfnAj87rvv1LGdDFCSPolku1grORH41VdfqQEgMmBNjpOs6aR1YaXPPv/8c1XmuEePHnjmmWesaiBNfpcvX1bvrwy2Gjp0KCZOnGg1J62LKs0l2TrSt7777ruLrYRhDTZv3qwqNEi/VDL6evbsCWsmbZWMB29vbzz//PNWFRQobF/2/fffq0woGRAo+zIJwlmrtLQ09V0zl+iW9sr7bK0k2Cl9EgnEdOvWDc8++6xVDSbIT9op76+0W85VPPzww1a9Lztx4oT6LZag59ixY3HHHXeUdZPIxlhHPjWVW6tWrVKp57KDkpqosmOVernWSg6+5QBRTljLCUop1yW1iK2VHATs2LFD/VhJAMbaD2jHjx+valFLeyWw9eSTT8KaDxAHDhyoTvxKWSwpffXhhx/CmjuP0l6ZK0neZ0nzlzZbKzmQHTdunPqOSRkv6Tzu378f1mrdunUqcHHnnXeqWvW33XZbsVlcZU0OZiWALPsy6dhIkFM6Edbqtddew4YNGzBp0iQVMLKmkqSFkfdVToDIvkxOMD3yyCOw5oDh4MGDVVBI3t9//vkH77zzTlk3i4iIyhH5XZFBgfI7IyXbpM9nzVMfy++2jMaX33HJIpDfdWsmx0XSN5H3V0ohS/l2ax5II8edUtpWykzJCWE5JrXmk8C33367Cr5JmbnnnnvOqkoIF1ZuXEoKS5v79u2rTgLLoCVrJYPWZs6cqcrhSfbbgAEDiq3UUNaklLuU75e+qQRiBw0aVGzWTlmTUu2nTp1S+zIJIt97772wZlKqXc6vyb5Mgp3FlXazhsEEcq5CMnGkLP68efNUH9tayf9ffiu6du2q9gsy+EHOYRDdECORBb3wwgvGr776Kuf22LFjjX///bfVvucjRowwrl+/Pud2u3btjCdPnjRaq2bNmhn9/PzUdYPBYPTw8DCmpqYarZFerze6uLgYs7Oz1e3Q0FBjrVq1jNYqICDA2Lhx45zbO3fuNPbr189orTZu3GgcNmxYzu0ffvjB+Oijjxqt1fz5843PPvtszu033njDOHPmTKO1mjp1qvGzzz7LuX3PPfcYly5darRWY8aMMa5evTrndqdOnYxHjx41WivZ1546dSrnduXKlY2xsbFGa+Xu7m5MS0tT16WdlSpVMlor2dfWrl075/bBgweN3bp1K9M2ERFR+SK/K4cOHcq5Lb878vtjreR3Oy4uTl2XvpP8rlsrOc6oUqVKzm3pm8pxk7WS44zu3bvn3Ja+v5wDsFa//fab8YEHHsi5PXv2bOP06dON1ur99983vvvuuzm3J0+ebPzuu++M1mrixInGX375Jef2oEGDjFu3bjVaq169ehn37duXc7tevXrG4OBgo7WqVq2aMSoqSl3PyMgwurq6qvNC1igxMdHo4+OTc/vs2bPGVq1aGa2V9J07d+6cc3vVqlWqj22t5NzEhAkTcm5/+umn6hwG0Y1gJhFZlIwiknrVQkpKySScMgrfWslIAXN7ZRSUsKZ6z8W1V7IapP6sNc2flJuMKpRUeBl9L6TknLTfWknJK8kiSkhIsIn2Svk+qUkrowxtob3yWZDRLmYBAQHo06cPrHlfZp6vSkaTSUacjNKxhX2DjIKSz4XMV2YL7ZXyjvL9kxrKtvDbZu3fNckkk5KD5lr21t5eIiKyPbl/x6OiotTvjjWXaMrdXvldlN91ayXHQ5KVY57X09p/x6V8n5RtM1fjsPb2SskrOfY0Z77J58KaPw+SPST9JiFtlnMA1lxuLvd3TaoFSPUTa55+IHd75dhZ5tKWObVtob3yV/rT1loOTcrgyfxO5jl0rH3fIOX7pA9trsZh7e2VcxPyO2HOfJPPgzW3l6wTC9KTRckPwK//b+9O4LSe/v//n2wVWohKSYtCoVUSRZs2EkWRElla7BFRSLZP8VGfKEQLStKCyhaVVFJatSmVFEqLokQL1//2ON/f+/pfM+aaBjPmdZrn/XabW9fMVTNn3l3XeZ/XeZ3zOq++6j7//HO3bds232GxhRuUbrJ2Bglbn3v37u3GjRvnkxkMyindhb59+/pSXpaQyGLrNtvj2WbMTTcqOTdixAg/qLGCuqgkChnEcpbL3Llz/WPay7lEbOu2hAE35Zn4P2eA8Nlnn/nt5rSX85+sbTUmgcGgm7ZSA55ybgwKxo4d64OJ7t27O0t4fy1evNgHCVxngjOSGU8++aR/TbNl3hLKZI4aNcrNnz8/nozjHBrwp7XyaPS3vEYp+UkZjQIFCsTPqHriiSfcmWee6SzhcFvKOtInUL6NA0Gjvmzo0KGucOHCzgruY0x88P6qWrWqL53JoJz2Uhuea27tPky5DM5OYuKGvozSlLSXcyOee+657G6iiIgEjjEoZ3dylgelr7i/ULoLnDlobZJ1165d/qwc7t+M7YhNonEHZ1CyuM0Kkm7cxxkzcy/nPs4YhPYWKVLEDRkyxFkb40eLk1j0w/Ulhvroo4/8dbZWvpsYat26dT7m41wqEjAsBuNcj6ZNm/ozUyyhrZTpYuEtyTjG0Pfff3+89Bhl/i1hkTAl/adOnerjvdy5c8fLNPfs2dNcgou+jHO0hw0b5r788ktfaiw614WyiSRsrfVljRo18slOztain4j6Ms7StXR2GfEpsRL9ALEofRkJWdrLwgLOKrbWlxH30ZcRmzKnSQlvjnug/cTU1voy5izoy4j9SRJxHu1LL73kXyOcBy5yILnYTnTAvyXyDwYx0erl1OhoCxUqZOrarly5Mr6DKDVuDkcffbSzhEn2ffv2pfkcN19LB5Nzg2XgkhYGC5xZZQmTwMuWLUvzOXZrWUsYkriIVpWlRrKTxJYlrHJJdj4ZK085KNQSdj1x7lNaSpUq5Qe2ofRlJBKtHc5MUpNANy0EO9YONCUxlBZW7lnbLUuCnsRbWkjQswNVRETkn2DMnOwcVxIbefLkMXWBSVwkmwaxtlOc3RcsBkwLE+7WFv4w/mQcmmwngbUqHYzvE6sbJGKRkrWkAPFTtKssNZKzJA4tIT4lTk1L2bJlTS1qDbEvY34l2ZlJxCSWdhUxb8X8VVqI9aztMEuvLyOWJqa2hHlX5l/TwnnrJO1FDkRJIskWrHLh0EIOrAsBq645cNFyOaxErNYj+LE2SEwLu0dICERl3axjYMPrN9kksTUcUj969Gj/EQJ27LGCi/dbCDp27Oh38nXo0MGFgN2bDz74oO9/Q0DwOGXKlCAGteyUpEQlSZkQEPS0bNkyaQJJRETkn2JSlQoN1ibTkmECmNiEBWzWkdhgXBeVQ7aOnSR9+vRxH3zwgQvB4MGDfdw3cOBAF4J7773Xx/5du3Z1IWAMSuWIaJeOdSzAHTlyZDCLq1jczM4dEsnWUfWAebZkCQ5rpk+f7qsPEaOGgKocc+bM8bsjRQ7Ezj5qyVGYTEu24sEitsyHtOkupPbSzugcnRCE1l7eZ7zfQhFae0Pry2hvKH0D1JdlndD6MhERCU9I9/HQ2hvafTy0MX5o7Q0xJgmpvSH1DaG1N7S+jPaG1DeE1pdJ9lKSSEREREREREREREREJAdSkkhERERERERERERERCQHUpJIREREREREREREREQkB1KSSLLF7t27/Uco9u7d63799VcXipDqpNLWXLlyuVDs2bPH7du3z4WC91lIr10OCw6pb6CttDkUv/32W1CvB/qxUN5v9GOHHHJIUPe1kOp/i4hIeLjPcL8JBffxUOISxkehxHvRGJSPUCgmyVrEIyHFfKH1ZaH0Y9G1Del8Kl676svkYHVYdjdAcqYjjzzSf4TiiCOOcHnz5nUhDQoOPfRQF0pbQzlUEblz53aHH364CwXvs5Beu0cddVQwSYHo+tLmUOTJkyeo1wP9WCjvN/qxkAIc7muHHaZhoIiIZB3uM9xvQsF9PJS4hPFRKPFeNAblIxSM70OarwgtJiEeCen6htaXhdKPRdc2pIV2vHbVl8nBKpx3ooiIiIiIiIiIiIiIiGQaJYlERERERERERERERERyICWJREREREREREREREREcqBcsZCKVUqQ1q1b5z744ANfc/ayyy5zRx99tP9avnz5XKFChZw1S5cuddOnT3eFCxf27aVG6sqVK13x4sV920OwePFid/rppwdx3gS1v2lvlSpVnDUcBvv222+777//3tWuXdtVqlTJH1T49ddfuwoVKjhraNv48ePdzp073YUXXuhOPvlkt337dv9RpkwZF4JNmzb56877LQTffPONr/993HHHOWuWL1/upk2b5o4//njfl1G7nr6sWLFivv+1Zu7cuf6jVKlS7qKLLvLnlS1ZssSdeuqpwdQAnz9/vqtWrZoLAQeurlmzxt8rREREMuMeOHv2bFeiRAnXrFkzf8bEsmXL/HjU4vkNH3/8sR9nMKavX79+cPfxvXv3+nHdmWee6UJAfLJx40Z3yimnOGt+/vln9+abb7o9e/a4pk2buhNPPNFt2bLFx1YnnXSSs4br+M477/gzqS699FJ3zDHHuG+//daP9YsUKeJCwBiUuaCCBQu6EFjuy5i7+uKLL1z58uVdgwYN/NcWLFjg51eIpyxh+vndd9/18ylnn322/+A84hUrVriKFSu6UPoy5oeIUUOwdetW98svv7iSJUtmd1MkAEoSSZYnXOrWrevatGnjNm/e7JMB3LAs3lzx3nvvueuuu85dc801vq0gwWVJq1at3AsvvOAHgzz+8ccfUzw/ZswY/5yVScgrrrjCTZw40T+++OKL/3ToH89Z1bx5cz8IYPAyYsQI98wzz/jJdosYXFWvXt0PXhkA0F5ez5YC3ddff90PUK6//nr/+KWXXkrx/FVXXeWfs+Lpp5/2g+0mTZr4xwxoE911113+OYsmT57s+zE+mABhIuGjjz4yFyhEhg4d6nr37u1fA5988omfQBg2bJizlMzm/5r7AY8bNmyY4nmCdEv3ig0bNriHHnrIX1cec19LxIQHz4mIiGSGkSNHuu7du7urr77affrpp36xz6hRo8xe3EceecSPRS+55BI/2c6Yn69Z8cMPP/hxJuN5HnNdExUtWtQ/Z8XChQv9/3ffvn39427duqV4vmrVqv45i3bt2uUqV67szjnnHB9Djx492r+Gy5Yt6yxav369j/latGjhk1gkOxctWmQq2cIYtFGjRu7cc8/1j2fNmpXiecb8PGfFjTfe6Hr06OEXqvGYBEYiYlaes+iJJ55wr7zyik8WEvsTr/A1q4j1WchYp04d/17juluK/5kD5Hr+97//9Y/phxOxaJjnrOD/nARb165d/ePUbSPpzXMiGaUkkWQpBojsvmFwABJG99xzj9mJVRIAl19+uR+IMxHI7osPP/zQlStXzlnBijESAdEOJ5IDiU477TQzO4i4hrSRiXYeMyBIPbHKcxZ99913PqCJVmWxo2jQoEGmJoJTryC6++673eeff+4/Z4Cwdu1aN3DgQGdpFQu7hFjhxmN2DSVix4ul1W+sECKRScDI49QJWSZArCRkUyM5S1K2ffv2/r1HHzZp0iSz7zeC3aeeespdcMEFbvfu3f51wKQIO1AtrSCMdt2wACIRyTdLO3JICrLLjf93Hq9atSrF87lz5zZ1XxMRkbCx475nz55+YpjdGCQxGIdaHSedcMIJbubMmT6mYszPxB9jUyuI75ioZtEMj4mnErHD2tKOHBaBsfOGiXQep55kZz7A6iT7uHHjfBKACVbcd999/vo+/PDDzqI+ffr4+JTFi9GYn2Rnu3btnKVEFu99KhfwmJ1aiVjQaKmqwerVq/0OSMbHPGZxayLGzDxnEe1mISC7WoitmQvasWOHs4jryu4xFo9TiYPk4W233eZ3cFpB/0X7SpcunWZfRrt5zgoqxpAsplIIj7mfJTr22GP9cyIZZWMmWQ5aTE4yyRrh5sokcQjtpUQCyQFr7U3c1kpwQDkpVopHlSOjBJIFXMNoUprHtJdBDEEE7eX6Wp205rXAdYyuZUivXavtTSzJxmNW7n322WfxgTjlMiwliRIHVNHAi5V6UWKWa2x18iN1X0awa+31kKy99Au02Vp7E5NABGD0ZSSy6Mtor6UkEf/fURKIxwQztJfXMAoUKKAkkYiIZMl9nLEzC8Gs3ceTtZfx3P79+50ljIWiJBCPie+4j7NgiXEHSRdLSSImTvmIHjNuZodLlBzgc6tJorRiKGuvh9BivsQSfTxmdxmLraLrmj9/flNJosRdYzwmgUySm2sNyg9aTRKl7sv4nD7CYvUIXqfRHJvV125iEojHLAplvi3qy1gAYSlJxFxENB/BnySMeP2y6BLcJ5Qkkr/CxkyyHLQoh8bKFm5WTKaxk4SV4lZR6ojdGKzAZjBDR2tpAJ5ap06dfCkpdmhFAwFrN9pE7CBjMMAOHVgdbEWrcthJRqnEGjVq+B05999/v7OKEgmsHLn11lt92ymPxhZuq9i+Tf13Vr1FQaXlAcyECRNc586d/S7D6HwcyzXgo76MlYbUqGZHjtWELHif3XDDDb7cwJQpU3wJAkvBY2qUpSFZSClKMBlmFfcE2nnGGWfEE0Ps2hMREcnM+zhxSceOHX3ZWModWzyvMbG9rVu39rsw2EmSupybJUz41qxZ009es0gFVhcpgclUxsiNGzf2O7asIx65/fbbfbUTSrYR85GQs4oyc7Vq1fKJQhba0VZL5a9So3w0pRyZF4oWXxJbW0X5M0qMM29BQgPsyrfclxH30Z9xrhaPLSaIQMzPmUlU7uF9R+nrtm3bOqtY0Mo5SZz1HM1TWJ5r49z3s846y1155ZXx8pOWylBKGFRuTrIcyRZKdXFT4HwMSzsF0sJKgalTp7rChQv7cxwYgFnFhB9b463WTE5EopCJam62VnY6HQgrMIYPH+5LjZHcZIBgGWUyXn75ZR+cUWqMEl5WEYBRP/fZZ591Ibjzzjt9aYQ77rjDhYIShASOJAToyywnXcD5ZOwsY6UpZfKiZJxF9GXbtm1LsZLTKpLHBOMkDEVERLIKE6uUD2LnAPdxq2fQRnEJi6koH1uhQgU/sRpNCFvDeIPkEJUYQjBnzhyfdGFMFwrGSpzxRNKFJIzlhWAghho7dqyPqUlwJu7csYYFgZShtJwMSERCtl+/fn4BZghIIr/xxht+USD9BEkjy4vXSLhx5g9l3IhPWFBuFWXwWPgwb948FwLuaSQKOW9P5O9SkkgkYOxsYEU7K7VCQJLlueeeCyKpJVmLgewDDzzgE8gh4DwfzqOK6n9LztasWTN/KCy7c0LAqrLJkyf7utQiIiISFiZT2dVufbElOMeDxXUkDC1XjZB/BwsYSQj06tUriEtOfEp5/1CSWpJ1KN3Grj36MsuLHiJUQ2JhKFU5rO4mE/uUJJJM9+STT/rJ1GTYyt2wYUNTOwQo2ZZM//79TU0EshonOleCckezZ8925557bnx3DqugrGwrZTUW24kj7HD58ssv/Q4Xblyswh8zZoyzYs2aNX61SDIkt55//nlnxYwZM9I9VPX88893Dz74oLOCFW/szEosOUewS21fUPKDladWkBBK7Mt4r3HuDHW0wUGblvqyu+66y1/TZChFwcHMVhB8ccBqMiNHjjQzGcKK48SVbuyIZPUxZdxYecyKPUsJT3YNUfYnQrlXrnX0/09JShL2IiIif1eHDh38ofTplZnifmMFuxnSKxXEYgoru4k4OJ3rm1gtgDNoK1eu7GMoxkdDhgxxVjD+pExX4mQlE6xReTza/eijjzpLi7+I8dNbDMRuKCsGDx7sd4skQ/xKHGXFY4895uOmxJiVEpTsxEfPnj1N7dS5+eab/Ws2GvPTXnYURVUNBg0aZGq3VtOmTdMtgUelmejcn+xG30X5s2SoeDFq1ChnBfOC9913X/xz7nEkvqOy7ewyZKGgFdy3BgwYkGKXIRWcotcr9z2OIxDJKCWJJEu2a0dJjLRw8J+VJAZY2ULHnwzn0kQDGiuDcJJDyTAIt1LOjUHWggULkj7PxGqVKlWcFSS1Vq9enfR5klocHGsFSbf0gnMOp7cUnDNJnV7JK+qWczikFQQL6ZX2oCyapXr7B+rLOGQzOv/JglWrVqUb4LCKz0qAg/RKDTBhQ/BrqS8jiZVeX0bCU0RE5O/66quv3J49e5I+z7mulkrHLlu2zJdmSsbSokDGR+y6T4YdOpZKou3cudOtXLky6fMssLJ0zi9zFcxZJMPOa0tnpRJDMdmeXgxVqFAhZ2nhZXrzQSy8tDQfRN9AUjO9vsHSThLO+WaeJRnG+FZ2kuzfv98vEk6GWI+YL5S+jPLtltpLv8BZROkl4SiZL5JRShJJtuBwcrahc0ZRCDhEnRUx5513ngsBiThqqFpZhZ8eJrVJDOzYscOFgCQdr99QatOOHz/e16flIwR9+/Z1P/30k3+/hYCVe5QASVztaRmHhLK7jP43BASRbJkPYXDL6mSScCRoQkAA1LJly3STSSIiIv8Ek6vjxo0zNamWHiaCiU0sn+mRuJiJcV16C9ws4czfPn36pFtxxBJ27xD3cY5qCO69914f+3ft2tWFgDEoFSQsn0mTeiEuVQ5CWWTFudqcZxZCyUcStsyzpZfssHbub+/evX2MGoKhQ4f6s+JeeOGF7G6KBMDGfmrJcZhMS2/1gzWsgEhv5Zk1IbWXdtLeUITWXt5n6ZXWsCa09obWl9HeUPoGqC/LOqH1ZSIiEp6Q7uOhtTe0+3hoY/zQ2htiTBJSe0PqG0Jrb2h9Ge0NqW8IrS+T7KUkkYiIiIiIiIiIiIiISA6kJJGIiIiIiIiIiIiIiEgOpCSRiIiIiIiIiIiIiIhIDpQrFkqhSgnGvHnz0j10rnr16r6OZ758+VyhQoVcdvv444/d1q1bkz5ft25d/3zx4sX9AYDZbdKkSekejN6sWTP35Zdf+kMVDzvsMJedqH365ptvJn2e9nFYJYeCVqlSxWW3n376yX344YdJny9YsKA/VPHrr792FSpUcBYOeZw9e3bS50888UR/UPD27dtdmTJlXHZbuXKlW7JkSdLnTzvtNHfcccf51w3vt+w2f/58/3+dzFlnneVy5crljjrqKN9uC4dobtmyJenzF1xwgfvxxx9dsWLFfP+b3d577z1/OHQyTZs2dWvWrPGv4SOOOMJlJ+5ZHLydzCGHHOJatGjhXzPVqlVz2W3nzp3pHgydP39+d/755/vrG8oBvCIiYsvkyZPdzz//nPT5Ro0aufXr17uTTz7Z5cmTx2W38ePHp3sGSsuWLd2CBQtM3Md3797t3n333aTPM/asX7++H1ufeeaZLrtt3rzZffLJJ0mfL1KkiKtcubLbuHGjO+WUU1x2W7t2rf+/TobXLHHUr7/+6k466SSX3Yif+L9OpmLFiu7II490hx9+uL/W2Y34lDg1mXPPPddfW+aCiK+zG/E/8wDJNGzY0G3YsMFMX8b8SnpnzBCTLFq0yM+vEKtmJ/6f33nnnaTP87q98MIL3YoVK/zrOLsRSxNTJ3P88ce7qlWruu+//97HqNmNeVfmX5MpXbq0K1mypI+5+VPkQLJ3BlkOSqtWrXJz586Nfz5q1Cjf8f/www9u06ZNbsiQIe7ss892VnzxxRd+oIi9e/e6MWPG+MTF0qVL/cQkgYKFG0CEaxsFZAxmmAi86KKL3MyZM13ZsmV9QFapUiVnAYMX2hUhUGSQW6dOHT9BTFsvvfRSEwki7Nq1K0V7eW1wjWnf22+/7W677TbXoEEDEwkikLxMbC/BWYECBVyJEiX89e3Tp48755xz3DHHHOMs+Pbbb1O0d+LEie6MM87wAc2cOXPcc8895z+3YvXq1SmScK+//roPyLnu/C70ZVxfK3hv0Wbs27fPjR492vdly5cvd4ceeqgPzi31ZZ9//rlPWkVJDQIIktxcc17D9erVMzHxESWJEl+7BL5MLtBGJslq1arlAzILE0tRQJbY3mXLlvmgh0Ua9GU33XSTD3iVIBIRkb+LhRHEd1FSg/sL4w5iFSaqWYxg6T4za9as+MQq7Wa8QYw6ZcoUf/+2dB9nQWDifZwEAWNPxp0TJkxw11xzjWvSpImZcdKOHTtStJdxPQtoSAgx3n/wwQdd7dq1TSxSAnMSie1lLMcEKgkLFpAOGDDATHwKFq0ltpeFSzVr1vTjPRaHDh482C9stYLYI1oYSGKW+SAWf/F70FcMGzbMzGsXjOlJYIJrShKGvozJdxYCMs631Jd9+umnPtaLErT0bYzrp06d6l+3zK+QyLCA+bXE1+5XX33lExskCukb2rRp418bFhJEYO4nsb3Eq7yGy5cv7/veHj16+IWXVmLq1H0ZCU/iaBKw3Nv69+/v72sWFrRKINhJJJJVBg8eHOvZs2f889q1a8dWrFhh9oLfc889seHDh/vHe/fujZ166qmxn3/+OWZVixYtYjNnzvSPt23b5ttrWaVKlWIbN270jxcsWBBr2LBhzKo9e/bEypYtG/vtt9/852+88UasS5cuMavWrVsXO/vss+OfP/7447Gnn346ZtX06dNjl19+efzzDh06xCZMmBCzaujQobHu3bvHP69bt25syZIlMavuv//+2EsvveQf79u3L1a+fPnYjh07Yla1adMm9tFHH/nHtLNcuXKxP/74I2YV77Wvv/7aP16+fHns/PPPj1m1f/9+35f98ssv/nPeZ7zfREREMgv3lUmTJvnHu3bt8vcd7j9Wcd/+8ssv/eO1a9emGENbw3iIcVEUk06ePNmPm6zaunVr7Iwzzoj9/vvv/vNBgwbFHnzwwZhVixYtShGTdu3aNfbKK6/ErBo/fnysU6dO8c+bN28emzVrVsyqvn37xp566qn4a7lq1aqxb7/9NmbVTTfdFHvrrbf8Y8bO9GXEUlbVq1cvtnTpUv94/fr1sWrVqsUsO+WUU+Ix6dSpU2OtWrWKWbV9+/ZYhQoV4vcy5jaJsa1ibqJ+/fop5jaZwxD5K3QmkWQpduOwsyHC6iG2kobQXnY3IL3SeZbaSyk8tkGzOsciVmCw0ilaQcZKLdpvFTsFWMUXlbmy3l528OXNmzf+ufX2srMhsW+w3t6Q+zLKOrKaM9oxab29lFBhVRyrUi2/fqPyGNZfu1xLVsVF5TGst1dERMKTeB9nPMp9h/tPCO3lvsh93SrGQ1xLyjKFcB9nvMkOdsaeIbSXXS+JO5xCi0l4bPn1m9heSp8xHiVuDaUvo9JItMvIenv5k9ez1RNFqBxBmTZivRDea+x84zVLfxZCe5mbSDwew3p7xSaVm5Msxbb91q1b+86Vwe3ChQt96QGrrrjiCtetWzefbOGcHBJFnJNiub3XX3+9a9++vS87xzbzKICwhkCBrc+0+eKLL3YjR470j62i5MAJJ5zgr22NGjXcs88+626++WZnFW1kINO1a1e/xfjJJ5/05dCsaty4sXvooYfcww8/7N9nL7/8si/vYNVll13mLr/8cp9w2bZtmy+l8tJLLzmreG/de++9Pqih3ANJWktlEtJqLyXQbrjhBl8qgfOerJRJTAuvhVatWvk+7Y033jDdl9GPUe6lbdu2/ky1559/3vdrIiIimYX7YJcuXVzHjh3djBkz/H2H+4/l9l511VX+T84q4r5uFeMhxkVXXnmlj/VefPFF0+MOyoix0I7XA3H0008/7f7zn/84q7imxE+UkWKS/ZlnnvFlpayiDBplEmkr15mzq3r37u2s4r3Fa4FkC7EqSQLKH1vFe+uWW25xnTt39mXdSpUq5c+nstzeq6++2s+5vfXWW37+LbvPIUqGZCyxCG2lhDtzFZb7MmJnYuhOnTr5x/369TP9XqMMHscj3Hfffe7YY4/17U3vfHCRtORiO1Gaz4hkYr1q6o2yEoP6yZYDBjBByUfhwoXdtdde6w/4tozAhnq5ZcqU8dc3uw94T8/+/fvdiBEj/Oohzk1iktXqIAYc8Dd8+HA/0U4tYhIblnHmCMkWzqyitq+l83KSrTSkRjWDLwaIlhOyIMlNvX36snbt2rlixYo5y6ZNm+ZrEXPAJn1Z4qpDi7i2JN9I0NKXWTgYNhnONHjttdf8ii3O0WLiJloxaxH11enLOM+AGuCcYyciIpKZOFuQSVUmVBl3JO5wt4axJ2dNssqasyY4FyNaLW4RyYBXXnnFffPNN/5s3+bNmzvLtm/f7mMSFlZxniuTl5YxPiJG5TqzMMzK+b7JsFtk7NixfvEar10SGZZx/hfn5RKLsFDJ+vkoJN4456d48eK+L7O6CDfqy1iwxlnKxNK8HnhdWLVnzx7fl5EwJFnI+80ydsUSQzHPQl/G2dqWkYR99dVXfezHfcLS2WoSBiWJJFuw+4XBIhOBIeBm8Nhjj/mVDyEgOCM5x6Gx1pGIYQBmubRUInaY8folMRcCkoijR4/2HyHo27evH4zxfgsBK2bZxdWhQwcXAlZtcXiw9WA9UrZsWZ/oInFkHYkjyicwwRACDsFu2bKlyhCIiEiWYSHFuHHjzBzyfSAsUCE2sZwwipAwYly3evVqFwIWYfbp08dXvwjB4MGDfdw3cOBAFwIqCBD7sysqBIxBSRixMyoElStX9pVQLFdmSETZMZK0uXPndtZR5p95NsvHPCSaPn2631FEjBqCoUOHujlz5rgXXnghu5siAbC77FUOakymseohFOyACWnTXUjtpZ20NxShtZf3Ge+3UITW3tD6MtobSt8A9WVZJ7S+TEREwhPSfTy09oZ2Hw9tjB9ae0OMSUJqb0h9Q2jtDa0vo70h9Q2h9WWSvZQkEhERERERERERERERyYGUJBIREREREREREREREcmBlCQSERERERERERERERHJgZQkkmyxe/du/xGKvXv3ul9//dWFIqQ6qbQ1V65cLhR79uxx+/btc6HgfRbSa5fDgkPqG2grbQ7Fb7/9FtTrgX4slPcb/dghhxwS1H0tpPrfIiISHu4z3G9CwX08lLiE8VEo8V40BuUjFIpJshbxSEgxX2h9WSj9WHRtQzqfiteu+jI5WB2W3Q2QnOnII4/0H6E44ogjXN68eV1Ig4JDDz3UhdLWUA5VRO7cud3hhx/uQsH7LKTX7lFHHRVMUiC6vrQ5FHny5Anq9UA/Fsr7jX4spACH+9phh2kYKCIiWYf7DPebUHAfDyUuYXwUSrwXjUH5CAXj+5DmK0KLSYhHQrq+ofVlofRj0bUNaaEdr131ZXKwCuedKCIiIiIiIiIiIiIiIplGSSIREREREREREREREZEcSEkiERERERERERERERGRHChXLKRilRKkdevWuQ8++MDXnL3sssvc0Ucf7b+WL18+V6hQIWfN0qVL3fTp013hwoV9e6mRunLlSle8eHHfdmtmz57t5s2b504++WTXtGlT/7XFixe7008/3eR5E5MnT/bXs3Llyq527dq+9jftrVKlSnY3zf3000/uww8/TPp8wYIF3Xnnnee+/vprV6FCBZfdvvvuO///n8yJJ57oTj31VLd9+3ZXpkwZl934f1+yZEnS50877TR33HHH+UN4eb9lt/nz5/v/62TOOussf6YW9b9pdwj4PyhWrJjvf0PA64XXcCg1wHnNVKtWzYWAA1fXrFnj7xUiIiKZcQ9kXFqiRAnXrFkzf8bEsmXLfIxi8fyGjz/+2I8zGNPXr1/f9H2cKZtJkya5b775xtWoUcNVr17d7d2714/rzjzzTGcNbXvzzTfdtm3bXL169fwYf+fOnW7jxo3ulFNOye7mubVr17oFCxYkfZ7XLHEUB9SfdNJJLrvxOuX/OpmKFSv6uRbOqSpSpIjLbvQDxKnJnHvuuf7aMhdEfJ3diP+ZB0imYcOGbsOGDWb7srTw+mZ+hVjVOs4jXrFihX8dh4C+7Pvvv/cxagi2bt3qfvnlF1eyZMnsbooEQEkiyfKES926dV2bNm3c5s2bfTKAG5bVm+t7773nrrvuOnfNNdf4toIEl1WDBg1y//3vf12rVq3c1KlTXdWqVd1zzz3nrLrrrrvctGnTXKNGjdz48ePdTTfd5L9mBYPZJ598Mv75F1984QeMDLDefvttd9ttt7kHHnjAWcFrdNiwYfHPP/nkE1egQAEfnPNa7tOnj+vQoYOzYsqUKW7ixInxz3l8xhln+IBmzpw5/rV78cUXOytGjx6dIgn3+uuv+0kEBlrffvutGzJkiDvnnHOcFbyX6LsqVarkH0d9WIS+guesaNu2rW8TwSyPN23alOL5kSNHmgh0QTK7SZMm/n7AY4LFRBwcbeleQSD70EMPuaFDh/rH3NcSMeHBcyIiIpmBe3b37t3d1Vdf7T799FO/2GfUqFFmL+4jjzzix3WXXHKJe+edd1zz5s3916xq3769W716tTv//PP9de3Vq5e79tprnUWMky644AKXP39+n4B75ZVX/Ji6Tp06zgpeo2+88UaKRYxMoJKwIHk4YMAAd8UVVzgrJkyY4GP9yLhx41zNmjV9ouXLL790gwcP9nMuVhAjRQsDeT3wmmUxK4vvdu/e7eNXS8lNYmYSmOCakuCkb2AhLgsBR4wY4YoWLeqs4FoyN0EMzWOSsomYB+A5C4ibb731Vv8a4PGVV16Z4vnjjz/e1L2C+Jk+ixiVx6nnqoilec4K/q9JsHXt2tU/Tt02Xh88J5JRShJJlurWrZvffcNkFRi83HPPPX6yzSJ2Dl1++eU+wGFAw+4LVpaUK1fOWcTg6qWXXvIryljRwOBlx44dZgYFqVfAESwwOGSwxcpCBl+sZLeIwRYr7El05s6d240ZM8YHDQMHDnQWsbKQZCHJFjzxxBM+GXvnnXc6i0hoPfPMM/664vrrr3eXXnqpX3lqEcHMqlWr/HUFqyIJIElyWcF7ix2Q7GziMSuGEpUuXdo/ZwXXkzbRX/E4dYDD6ixLfRl9VrTrhn4hEav0LO3I4VrSJ3Dv4jHXNxF9mtX7moiIhIfqAD179vQLwfbs2eNjEnZrHHPMMc6iE044wc2cOdPvDGCRGBN/TGBaxHiO6/nDDz/43SJUvCCejsb81jCxSkxC8oLxEQkDEhwkEq22l+sZLfZhUpiKF+3atXMWkcAgqRUtDCV+ov3szrGIBZjsKuS6Mh9AJQaSXhaqRqSlY8eOfmKdxDEJLfoGJuEtVWhZvny5K1++vH9/8Zh5q0TEJFZ2EO3fv9/P9xDX8Zh+IRGxnqUdOfS3LG4nRuVx6qoixNI8ZwUVY0hsUimEx6l38B177LH+OZGMstPTyUGJG1bevHlTTExRSiqE9jKY4aYVSntpK4MB2mtpYjXCoJD28hoI4bXADZZyTFGZK1aWpZ4YtoRJ4MT3Gu1l9ZNVTLiz6ykSXV+rSSLaxkqnCOXaCBgsJYkSB6w8njt3ri9NEb3PSM5aShIllhshYUHikMRGFOiUKlXKVF+WmASibMpHH33kJ2zo27hfWEoS0W9FSSAe83qgvQQP4L2nJJGIiGRFTMJkKjtsLY/zE9tLTMLkpVVcxyguDSGGiuK9aJLaenuZZE8sw2w95qNtiTEUj4mrrCaJaC9JZPCaYBEjcavVJBHtveqqq/xj+ohdu3b5XUZU6rAisew9CRaqddDG6CQREkj0wRZwP4iSQDwuW7asX4T9448/+vYSm1pKEiUmgXjM65RKOD///LP/Ggl7S0kiFmJEizH4k4QRCyBIcEbxtpJE8lcoSSRZilVE7BZhsMhkGoMwtp9bxYDg7rvv9hOVCxcu9B2thbrJyVDGj1IDfLz//vvuoosuMlvKj+CG10OLFi18IoDVZOzYsoqSA6wypLwDO7WeffZZd/PNNzuraCMrXdhOzCCWVVus3LOqcePGfofhww8/7IPel19+2e/UsiraZcjglvrqJGDYxWdVv3793AsvvODbHQUJDBqtuuOOO3xpvwsvvDA+qWB5wobVhdTPPvvss/3nVgKxtDAxQztJaEaJocSEp4iISGbEJJ06dfKr8Fn0wbk+ls9rpL2tW7f2JcUo3WU5JqESA2XmaCtVORh/WionnValCyZ/b7jhBj9Z3b9/f7OVGMA1JX7q0aOHT7hQ6YCdLlYxt8J4mbayoPHdd991vXv3dlYRP3Xp0sUnW4hVOcuFM7Ws4n12yy23uM6dO/uyhCxa43wqq6JjHTg3OWL52PlatWr5stdRoisx4WkNr1nOSeL9FiVaLCe8OfednXqU9IvO+rJw5peEReXmJMuRbKFmKpl4zsuwcsZEMqwUYEs8ZZs4x4FyeZax5ZwdI5RLYFu8pZX3qXFTpaYvq4fYxs+g0cpW6LSwxXj48OF+ZQ4DGhIblm3ZssUnW1jpwjZ5S+flpIUyJNQgJonMgJzdGSH0Zawq471meVUOiVgmaiyd8XSgCQVeC5Z2ZqWHci8kCxN371neFUkSmXO0REREsgqT1bNmzfITgCyysrpwDYw9OSeHXQNMVrJQkAVtVlHCj3MymARkfG91532EM10p1cxYiRKExFGWMUYiRiXpwgIrS2d4poWFt2PHjvWL10gSkMiwjIVgnJdCQoC+wXICObEvYxcJi3EZ91vFwi+qW1i/plG/wGuVXUSW54Ai8+fP9/G05eosibinMTfIeXsif5eSRCIiIgeZ//3vf34VGTt0QsDZWewypa56CJic4XyqUJJarCqjfj11qUVEREREJHwkuG+//Xbzi0MjlEVk9yYVW6yjCgcJbhKGlhc9RKiGxCJ3yg+GkIQTm5QkkkxHmavo4Me0cLBiw4YNTU1OLlmyJOnzbJG3NBFIyTZWXyQzZswYM4fEshorvZ0MrMKfOHGis4JDFXn9Pv/88/4xK0cSUUOX56yYMWOG3/n24IMP+seUbktEaQqes4JVLezOuv766/3j1OXaGOTynBVPP/20L5PRpEkT/5hVZYk4gJXnrLjvvvtS9GUcbMwOkmhQS1KDHTtW8H9NGdJolyEDcAKH6GDYoUOH+h2dVlYcU94jsfwAq48p48bKY8rNscvM0opYyv5EuM6bNm2Kr4ylJGV04LGIiMjfQcmz9evXJ32enSSWzvFgR0t6pYJYTGFlNxH37fRK4HEuBjtfLO2279atW9Lnq1at6vr27eusmDRpko/1mFznMfF+6sVAPGfF4MGDfXxP5QUev/HGGymeJ2blOSsoKc77jXE9jxnjJ6I8nqUzlG688UZfbpBdLjymLF4iYlZLu7Uo37937954ImPBggU+SRSVv2Y3iZXqMuwmZPdY4uckM6pUqeITGeyAonKLFcTSxNQR7nHMXzAnAGJpYmoruG8NGDAg/jm7yqjgxI5e8D689dZbs7GFEholiSRLSttEh2OnhZqulmpjMgig40+mTJkyprYYcxA9Z2EkQ8muaJLVwsQq2+GTYSAT3XCtJLV4/VK6j8erV6/+U1KL56ygrBzbtgnAeZw6UGdLv6XgfOvWrT44p+Qkj5m0Tr1d3lI5Smpm839OUMbj1MlZShBYSchGfcPOnTuTPs+hoIkH81oYhFM+JRlqQB9xxBHOivRKDRDkcP6CFfRf6R26zOv69NNP/1fbJCIiB5evvvoq3fs457pauo8vW7Ys3bM6LC0KJNZjXJcM19XSubnE0qkn1hNRvt3SJDtzFUyuUzqax8R/idh5bamsNDETk/6FChXyj4mjErErg+esICYlRiLu4DFxauqzfy3FJMT8xMy5c+f2jxlHJ+JMT56zgqQQ8yzJEJNY2UlCX7Z48eJ0+zJiPiuIpdPre3ndElNbQV9AGdJkmF/h/SaSUUoSSbatIKe0EGcUhaBOnTruscceS3EgoGUk4qihamnCPb2ggsn2HTt2ZHdTRA569evX97vL6H9DwO49tsyHMLglAcrKrdSBpVUEQC1btkw3mSQiIvJPkHihtJClSbX0sPua2CRakW8Zq/EZ16Ve1GYVZ/726dMn3YojIpI5SMyya8dScisZkrTMs6WX7LCEah3shiNGDQHVOebMmeNeeOGF7G6KBMDGdgPJcZhMS2/1gzX79+9Pd+WZNSG1l3bSXquH/3FTZbcOWLHHobGWz6GZMGFCfGdcvXr13OOPP+4s4v+8V69ebubMmfFJdZLGXbp0cRaxAq5nz56+nEa0k+/+++9PUYIslL43lL4B6styZt8rIiIHh5Du46G1N7T7OLF/euX+svM6Ei8x4bt79+54OS9K5FvEDihiks8//zxecuy2225zbdq0cRZt2bLFl3Kjukj0eqX0IGXRre6QJEZlZ1w0X/Xaa6/56jIhUV+WtX2Gxb4stL5XbFKSSEQkSS3wm266yZ+ZFNV0tVQyI7XPPvvM9evXz40fPz5+OD27GqziTKKPPvrIJ+Gico6Ux7OKM4k454X2RjWeqaEsIiIiIiJheuedd3xcQiIgKoFmqRRaauwGoGTziy++GN8lEsV+FnFmLgkLzp2JSuJbOXM0LZzfwvk+jzzySPyMMqqeiIjkBEoSiYikgcEsA1irq5xSoyY851FxMGwo7T3rrLNchQoVXCjt5bVATWoREREREQkfY/xKlSr5A+lDae+5554bTAlH2kt1C0pIh9Lexo0bB7dzSEQkMyhJJCKSBlYMtWrVypcfYCAe1fYlsWFR7dq1/S4XVpVFiQwOKrR6OP2VV17pJk2a5FfuFS1a1H+NHVtWB+SUcWBnGe3Lnz+//xrBGQfFioiIiIhIeJo1a+ZLjL/66quuRIkS/mvFihXzZcYtuvHGG921116bonrEySefHG+7Nffee6+76667fFuj6hGcVWa1IgPnUFOd44YbbohXj6hevbrpCh0iIplFSSLJchyKvXjx4vg5HrVq1fJ1Xq1u4+ZQt1WrVsXrdjZt2tQNGTLE5DZj6qFOmzbNbdiwIV4/u3Xr1v5AUIsDL+omU2KM2sS0l4HXVVdd5WbMmOGs4f+fa5srVy73xRdf+K+VKlXKbJKIc5M+/fRTt3HjxnjAwFZ5q0kiDqjkvcZ1jgKG5s2bm00S0SfMnTvXlxyMAoaOHTuaThLR3pUrV8b7MlbFUaKCwNca+oNPPvnEH8Ic1f++/PLLfblHi+1lpyF9GWUpaTvlKDhTi9e0RZxTRnu3b98eL+3YpEkTf5i4iIhIZmA8n3iOx2WXXebvM4yfrWFsxBk0jJujGKp9+/b+Ph6VxLJ2Dg338R9//NG3l4VrnIvJOMki2vnxxx/7Mz3BWK5mzZpu0KBBzprNmzf7st2MlaL5iQsvvNBskoj3GGP8Qw89NF5urm3btmaTRMwF0V7ip+i9dffdd5ucqwBnPfF6iMb3YBeU1SRRWn1Zu3bt/LxA9PqwhLOIP/zww3hfxnW99NJL3bvvvussInZiTijqy1jcet555/mY2qLvv//en/kcna9GP8b1pU8TyQh7IyA5qLDChRX4V199dXxile3cVapUcRY9+eSTbtiwYX6ymoFXdCOzup27U6dOviZx3bp1fTIjGihwjS1iUpLBSlQSjccMviy+HhiAc5Ndt26dCwFBbcmSJd3kyZNdCGgnwe3zzz/vQvDmm2+6bt26+aAmBKyAY/DKBE3UlzHBYLUvu+OOO9zs2bP9ADbqy0jEWC39wT2ChQ9nn322/zy6xtWqVXPWcE+gnazaTNxlmCdPHrNJZBERCQu7BUhiNGrUKD6xyn3S6n2GagE7duzwiYsIE5YW7+O0i3YyUU1paRxzzDF+4ZLFcRKTqbSLxUmJi6lIwFhcJEoyi6oRVDcIAYlBdrk8+uijLgRjx451TzzxhN/9FIJRo0a5kSNH+oXNIWjTpo1PdJK4SOwzrJag57pSPSQqOc/CNeYJK1as6KzZtWuXbxfxabRokbiKfsxiTM28FQuaqdhSsGBB/zX+pH+zmpQVe5QkkixF1r179+7ulltuCeJKT5061R9Qz6A2lPa+9957QdT4ZVXhrFmz/M3W4gq91EqXLu1XtuzcudNkQJNa5cqV/W4iBi7RhLVlBLvseAsFA29W5YSCvuGpp55yF198sQulvQRlJDJCubdt27bN5c2b11m3adMm3zdQSkVERCSr7uMsTrG64z6t9rIgLJpIs4wV9+zAX7RokQvBihUr/K4WKnGEgIU0obQ1iklGjBjhQmovVU9CEVp76ct4z4WQBCAeWbNmjd+tFS0KtIyKHEWKFPEl/UNZNNygQQP37LPPZndTJGD/t8xHJIvUr1/frV69OpjrS6eq9mYNVhVybk4oO3PWr1/vk1mshGOinQ92xVnFgIsVkYnt/c9//uOsIjCfP3++n0yI2vvyyy87yxPttO/888+Pt9fyri31ZVl/b+M9FwJKpVIagUkmERGRrKBxR9YpVKiQL8dMidsQsOCH3eB79uxxIWA8t3btWr/gLhrjWyyLFyGB8c477/jdAVF7qXhgFYuq+vbt6+rVqxdvL+XcrGJBK+fQsisyai/zAlaF1Peya6h8+fI+rg4Bu53YlUVloRBw1AC7yqKygyJ/R66YXkGSyfr37+/LDUQoIcRgMdqNceedd/oJNkvlEZYtWxb/fPr06b6DjWq4MqiJtsNaQI3Z6FwJyjhwfRkkRrtzWFlkZVUcN1TOFUksP/Dll1/6wx9ZPcIq/DFjxjiLKM2V+LoAZ+dYei0k4jWRetKas4msnvHD4PDbb79N8TVKUlg8+wuclcNZWomosW9p1dZ9993ny08m9mU1atTwZcVAqQdLZUmuv/76+IQHO+DYaZjYl7Fqq3Dhws7KTkjKIyYGkNRYZ/UpCXB277399tvOCt5blCONcJ15z0WlSFnh+9xzz2VjC0VEJHSdO3eOr7jnPsmZRNzHoxLjlBQ+8cQTnaVSsdE5jZzXwM4cYr6oPN6ECRPij7MbE30dOnSIf75161Z/rUlkEEOxut3S7hfOH+7Ro0eKcTOxVFQej3ZbLY/G+J72JmL8SUksixjjpZ5kZ1zHa8Ii4tNo7iJCFRQr8xWpEf/z2k3EXFYUT1nQokULf9YzaOuCBQt8XxZVEyFpGPXDFpKEnPeW+DnvN44boC8jlh4+fLizgliamDpCgpDzykhugViamNoKFq0OGDAg/jm7yqiGE/VfJDtvvfXWbGyhhEZJIsl07BRhIJteGS9WRFnaRkpJsWQY3HI4qKVBOMmhZBiEWynnRsDIoCUZBjIWzyM60PVncnvevHkulHPBKDMVSqkpkrJsRX/sscdcCDp27OgTMYmBvNW+jNrJlkonMghPb5UpNaCpt29Feu95ghxL5xiQoCeJlQwJeqvnRIiISBjSmky1PLHKDvb01sdaKpXHBPAXX3yR9HkWM1pa+MP4k3FoMvnz5/cHqIdk8ODBPu4bOHCgCwELX0kUde3a1YWgZcuWPnmQuAjLMuZYOKvIwviZ+RXmWZIhJrFSzo15K95HyRDrWTqP6EB9mbXziJh3Ta9SD+fQcm61SEbZmEmWgwqr6/kIhaVOPiOileAhYDWepYArMxBcUkIhFAwgo1WTIQitvbQ1vUH6vym0vszS5EZGhNSXMSkXUntFRCQ8FiZL/wpLizkOhInTkO7jTJyG1N6MUEySc2KojCD+t1KEqWrVqi4U7GgKqW8IrS9jJ5alyiYSPhv7qUVERERERERERERERORfpSSRiIiIiIiIiIiIiIhIDqQkkYiIiIiIiIiIiIiISA6UK2alsKaIiDEcXDhlyhS3Y8cO//mxxx7rLrzwQvf111+7ChUqOGs2b97sPvnkE7dr1y7/OWeDcYbV9u3bXZkyZZw1HLL42Wefud9++y1+Rk3x4sV9jWr+tHhA86JFi/wBnDjvvPN8zfijjjpKtYBFRERERAL07bffuk8//dTt3r07fs5n2bJl3a+//upOOukkZ82qVavcvHnz3N69e/3n1atXdwUKFPDnvxQpUsRZs3DhQh9HRef61qtXz8dThQoVcgULFnTWzJw5061duzZ+ZtKll17qvvvuO3fyySf7cz9FRA5Wh2V3A0RELGLQzQG3DLpLly7tv3bCCSe4vHnzmkwQ/fDDD65ixYrusssuiycsGMTWqVPHHXPMMc6axYsXu/r167t27dr5JAuKFStm9lDhCRMmuM6dO7urr77aJ4aipJbV9oqIiIiISPrWrFnjatSo4a666iqfaIkWBtasWdPkpZs+fbpr1aqVj6GihEW5cuXc6aef7iwaNmyYe+SRR3ybDzvs/6Yfud7ly5d3FvXo0cO9++67rkmTJu6QQw6JzwtYvb4iIplJO4lERJIEDI0aNXKrV68O4vp88MEHrl+/fu799993IRg4cKBbsWKFe/bZZ10I7rzzTleyZEl3xx13ZHdTREREREQkE4wcOdK99957bsSIEUFcz4cffthRDKhXr14uBCSziKnbtm3rQkBykJj6nHPOye6miIj863QmkYhIGtg9lD9/fvfTTz8FcX2qVq3q20qpthDUrl3bbdiwwYWCXU8kDkVERERE5OBAUmDTpk0+8RKCunXr+tLnoWjQoEEwiy5DbK+ISGbSTiIRkf+HAfett94avx4bN250W7du9WXFwLk+AwYMMHO9Zs2a5Z544on45wxoc+XK5eslo1atWq579+7OirFjx7rhw4enKDlH3eyiRYv6z6+44grXvn17Z8Uzzzzjd2hFZs+e7UsNkDzEbbfd5ho2bJiNLRQRERERkb+CnUNUNYgsX77c5cuXz5UoUcJ/3rRpU9elSxczF3Xo0KFu/Pjx8c85j4i4lLJ4uP76633JcSsee+wxHzdFZsyY4Ut0H3nkkf7znj17mtqpc/PNN7tvvvnGP+YcItpL8jAqMT5o0CCTZ1OJiGQ2JYlERP4fDiflUM1kGNhaOo9o+/bt6e5uIXAggLCCVXocDJsMZz4VL17cWUGwsGXLlqTPlypVKn7+k4iIiIiI2Mf4PkoKpKVw4cKmkgLET8RRyZDcYuGdFcSnxKnJlC1b1hUsWNBZQfzPPEAyZ5xxRvz8JxGRg5mSRCIiIiIiIiIiIiIiIjmQziQSERERERERERERERHJgZQkEhERERERERERERERyYGUJBIREREREREREREREcmBlCQSERERERERERERERHJgZQkEhExZurUqW7JkiUH3c/KKvv27XOvv/662759e3Y3RURERERE5IDeeOMNt2nTpoPuZ2WV77//3o0ZMya7myEictBSkkhExJjevXu70aNHZ/r3nTJlilu6dOm/8rP+Tb/88ou76qqr3Jo1a7K7KSIiIiIiIgfUpk0bt2jRoky/Uiye++GHH/6Vn/VvWrBggWvXrl12N0NE5KClJJGISA7x0EMPubFjx6b4Wr169VzFihVdyI444gjXunVrd+yxx2Z3U0RERERERLINi+dSV4ogVjrhhBNcyIoXL+5atWqV3c0QETloHZbdDRARCckff/zh5s6d61dnlS1b1p1++unx5+bPn+9+/vlnV7du3RT/Zt68ef7rJGRmz57tvvnmG5crVy533HHHucqVK7tChQql+zMnTpzof06ZMmXiX5s+fborUKCA//c40PedOXOm27p1q99JxOoyXHrppa5WrVru+OOPT/HzfvvtNzdr1iy3c+dO/31KlSqV4vnJkye7EiVKuKJFi/oVXYceeqg755xzXJ48eZL+DtG/oW1z5sxxBQsW9D8b/Bzaz7Xl5/F9U/vqq6/c8uXL/TXgWlAyoUGDBv77HX744f53OeaYY/zf/fXXX93bb7/tLrroIv//tGLFCv+zo2tF0LR27VpXoUIFV65cuT/9rIy0R0REREREDk67d+/28cCePXv8groTTzwx/twHH3zgihUr5s4888wU/+b999/3iQy+PmHCBP89DjnkEP9vq1Sp4vLmzZv05yXGL/ny5Yt/fdy4cT7O4vviQN/3zTff9H9+/PHHPvbje/E9mzdv7ooUKZLiZ27bts3/jnyvmjVrxmMpEAdF8RZtI34iZqxevXrS3yHx31AGnLiTeOvUU0+Nl4sjLiaGrVq1aorfM0KcSLvPOOMM/3tRGv3KK6/0z9H+Zs2axf/u+vXrffx9ySWX+PYRCxO7lSxZ0v3+++/us88+8+2gzal/94y2R0QkR4mJiEiGrFu3LlaxYsXYaaedFmvWrFmsePHisUsuuSS2Z88e//zrr78ey5cvX2z37t0p/l2FChViDzzwgH/8v//9L9a6dWv/UatWrVj+/PljY8eOTfH3L7jggliPHj3in5csWTL24osvpvg7jRo1it11113xzw/0fZ966qnYcccdFzv99NPjf2/Hjh1/+lmLFy/2v1f58uVjDRo0iB155JEpnke1atViderUiZUpUyZ20UUXxUqXLu2vCd8vGf4NP4vfhWtHe/Dmm2/Gjj322Fjt2rVjjRs3jhUoUMD/Lokef/zx2BFHHBGrX7++v/5NmjSJcfuaMWOGf3779u3+888//9x/vmHDBv85P4/ft2nTpv7f33bbbbG2bdv678HPyp07d2zIkCEpflZG2iMiIiIiIgenKVOmxAoXLhyrWbOmjyMKFiwY69WrV/z5W265xccZiTZv3hw77LDDYh988IH/vEuXLj7euvzyy308UqpUqdjSpUtT/JtDDz009t5776WIX1asWJHi7xx11FE+Pokc6PveeOON/vsQq/H3br/99j/9LLz22ms+zjvvvPP873n00UfHxo0bF3/+119/9d+HWK9cuXKxiy++2MdIV1xxRdLrFv0brhlxYsuWLWMTJ070zz300EP+OhJfEatyfbnOkf379/u4Ovo7J510kv8zccqS70X8Fhk1apRv91lnneVjNz74Pxg4cGA89jz33HN9PDdv3rwUbT1Qe0REciIliUREMohB5h133BH7448//Oe//PKLTzg88cQT/nOSQySJSBZFFi5c6Ae3K1euTPN7jhw50idvfvvtt3+UJMrI9yUIYECcKPFn8XtVr1491qpVq9jvv//uvzZt2rRYrly5YjNnzoz/GwbdtGnbtm3x37tEiRKxp59+Oml7+DdFihSJbdq0Kf61b775xgc+iQPyBQsWxPLkyRMPdr766isf1CQGR9dee22GkkT8vej/iuvH1zp16hT/Pv369YudcMIJf6k9IiIiIiJycCK+IXkwZsyY+NdWr17tkxFR7DF79mwfH61fvz7+d5555plY0aJFfbIjLZ07d/YJiX+aJMrI9+X7fPjhh0l/FgktFhT2798//vyjjz7qk0A//vhjioQPi/v27dvnv7Z8+XL/e8+dOzfNtkT/pl69evFFlBg/fry/NvyOkeeffz5WrFix+N975ZVXfBy9du1a//nOnTtjlSpVOmCSiOcT4+T27dv7r7366qvxr5GsSkxuZaQ9IiI5kc4kEhHJgFWrVrlPP/3Ulztj2/+YMWPcpEmT3Mknn+ymTZvm/w5b4lu0aOFGjhwZ/3c8Zov7KaecEv8aJdCmTJniRo8e7fbu3eu31FP+7J/6p9+Xv/v555+77t27+7IDqFOnjjvvvPP890xEPejoDCB+7xo1ariVK1em+/35N4lb/Sl7R9m5H3/80V9PyhNQVo4Scp988km8ZELp0qV9ObnI3XffnaHf56abbvLl90AJhehrEb62ceNGt2vXrgy3R0REREREDk5vvfWWL5sG4gE+KGl20kknxWM+yr8RA44aNSpFzMdZQJThjqxevdq9++67PsagpBklyzPDP/2+/FvcfPPN8a/dddddvozdhx9+mOLv3nDDDe6ww/7vlIry5cv7WO5AMV/Hjh39mbGRYcOG+fJxlH+LYiyep9xb9L04N7dly5Y+7sPRRx/tOnfufMDfhbLj1113XYr47qijjnJt27ZN8TVi+b/SHhGRnEhnEomIZMC6dev8nyQLEgf/DCgTE0BXX321r/tMjWfqOhM83HPPPfHn+/Tp43r37u3rRzPIjgbdmzdv9gPvvyszvi91nJF49hEIgqLnIlGCKJI7d25/llF6Uh+WyjXdv3+/DwoSkZQqXLiwf7xhw4Y/nYmU+vNkEutq075kX6PdBCIZaY+IiIiIiByciAdYLJc6HuCcocRzidq0aeMTQ8R5LLQj4TBw4ED/HOfhkDDi7KKzzz7bxx8s3mMhGs8lxpJ/RWZ9X+I6kl5RvAjOluXco6yK+djglPqatm7dOr6gj5iPc4H+asyXP3/+FL837WPRX3ptzkh7RERyIiWJREQygAEoHn74YX8AZzL169f3O09YlcQhnezuiQ7bJHF03333+d0+devW9V/btGmTX730f5UB0kagEq1oiyQOdP/u902NdoNAg1VpET7noNJ/KvWgm2vKB6vgkiEw2bFjR4qvcQBpVshIe0RERERE5OBELEDMcqB4gJ0qLNBbunSpr3zAorwoyfHOO++4yZMn+x0/UQxFQoKdSMlis6iKQ2LMx2OqQ0T+zvdNFvMR36XG16J4MLNjPhZVDh061EzMd6D2iIjkRCo3JyKSAQz62U3y/PPP/+k5tqbHO9VDDvFJIVaW8XHhhRfGd6GQMGIAT/IoknoFU1pY1UUwkDhgXrx4cfzzjH5fdsukt/KLwTJtHT9+fIoEFIFHrVq1XGZr3LixL+dGcisR5d9+/vnn+C6eRYsWpVjV9vbbb2d6WzLaHhEREREROTg1atTIx1ostktEDEVcFClXrpwvKR7FfInlzVisR9IjcZHdgWI+YjB29iTGfLNmzXL79u37y9+XcmvpxXzEV5TcZvdThPjnp59+ipfozuwYi/iSCheJvvvuuxRtIgnGjqh/I+Y7UHtERHIi7SQSEckAysq99NJL7oorrvBJmQYNGvjt/RMmTHDXXHNNiprJBAn9+/d3CxcudIMHD06RhGGVGWUC2rdv71asWOFeeeWVA/5svv+dd97pSwqww4c6yokrtDL6fc866yy/Ko5kEucIJZ7zE5UZePLJJ92NN97of7cSJUr4pBg7p2hDZuO8oy5durjmzZu7W2+91ZUtW9bXi2bQTq1sVnmRZLvgggtcw4YN/d/ZsmVLfNVXZpcDyEh7RERERETk4ERZuQceeMDHPrNnz/Zn11BOjjNpR4wY4QoVKpQi5uvRo4f75ZdffMnxCPELsVuHDh188oNzfigRlx4SRJSwu+OOO3yyggVqw4cPT1FKLaPfl5hvwIABfmcO8SOl0BNVqlTJnzVELNitWze/Y4nS5bfccos77bTTXGbr2rWrTwCRVCPWohwc5zxxfZcsWeL/zu233+5efPFFn8Ah3uacpY8++ijT25LR9oiI5ETaSSQikkHNmjXzJQUYPM+cOdMHBM8888yfDtVk11GnTp3cJZdckiIRw+B/+vTpPhnx8ccf+/rIDEapf5x45k29evVcxYoV45+TtCFIoH4yO2oGDRrk619z/tBf+b7333+/HwhzrhKHsu7Zs+dPP4uAiICD3TMMzgkg2EmUGKCwwi51yT1WndWoUSPptUvr34Da3SRhOCiV1XKsjONPVudFSMRxDRi887vyOfLlyxdP4PG7RjWzjzzySP95YlKHv8vXWFkXIWjia9HZRBltj4iIiIiIHJwoI0c8xII0Yj7iCBIWJF8SUT2CBAyJlpIlS8a/Xrp0ab9Lh8V9M2bM8MmI999/38cdUVk58Hni+T0kSUgSzZs3z8dptIFEVHQWUka/LzubOLOIBBLfI62fxULAfv36+cWFLIojvmSRY4TYj3+TuuQ48TDn1aYl2b8hNiP+fOSRR9yaNWt8lQjiRn7PCHEcsWe1atX8n8SntCeK96LqGq1atYp/zjVv0aJFip/FNWLBXyIWVCYmyjLSHhGRnChX7K8ULxUREfmXUR878dBUEmbRriJ2P4mIiIiIiMjBE/OxWJEkDgsWRUQk66ncnIiImNarVy+3f/9+v7KMlW7PPvuse/DBB5UgEhEREREROQi0bNnSn4PLDiF2+nDe0sSJE7O7WSIiOYZ2EomIiGkc2Pryyy/70gOUiGvSpIkvrSciIiIiIiLh40zcIUOG+EWBnI3LOU8q+S0i8u9RkkhERERERERERERERCQH+v9PtxMREREREREREREREZEcQ0kiERERERERERERERGRHEhJIhERERERERERERERkRxISSIREREREREREREREZEcSEkiERERERERERERERGRHEhJIhERERERERERERERkRxISSIREREREREREREREZEcSEkiERERERERERERERGRHEhJIhEREREREREREREREZfz/H8KcKha5POOBgAAAABJRU5ErkJggg==", "text/plain": [ "
" ] @@ -481,10 +482,10 @@ "id": "cell-08", "metadata": { "execution": { - "iopub.execute_input": "2026-07-26T18:10:53.112690Z", - "iopub.status.busy": "2026-07-26T18:10:53.112608Z", - "iopub.status.idle": "2026-07-26T18:10:53.121257Z", - "shell.execute_reply": "2026-07-26T18:10:53.120802Z" + "iopub.execute_input": "2026-08-01T20:07:37.803218Z", + "iopub.status.busy": "2026-08-01T20:07:37.803108Z", + "iopub.status.idle": "2026-08-01T20:07:37.814389Z", + "shell.execute_reply": "2026-08-01T20:07:37.813840Z" } }, "outputs": [ @@ -496,21 +497,21 @@ "community_detection — accuracy (rows = trained on, cols = evaluated on)\n", "eval_homophily lo mid hi\n", "train_homophily \n", - "lo 0.318 0.357 0.410\n", - "mid 0.283 0.398 0.559\n", - "hi 0.219 0.384 0.621\n", - " trained hi -> evaluated lo : 0.219 (native lo = 0.318)\n", - " trained lo -> evaluated hi : 0.410 (native hi = 0.621)\n", - " best generalist (row mean over all eval regimes): trained on h_mid\n", + "lo 0.347 0.351 0.496\n", + "mid 0.299 0.405 0.606\n", + "hi 0.275 0.385 0.666\n", + " trained hi -> evaluated lo : 0.275 (native lo = 0.347)\n", + " trained lo -> evaluated hi : 0.496 (native hi = 0.666)\n", + " best generalist (row mean over all eval regimes): trained on h_hi\n", "\n", "triangle_counting — log10(MSE / triangles) (rows = trained on, cols = evaluated on)\n", "eval_homophily lo mid hi\n", "train_homophily \n", - "lo -0.506 0.006 0.430\n", - "mid 0.344 0.276 0.452\n", - "hi 1.186 0.751 0.627\n", - " trained hi -> evaluated lo : 1.186 (native lo = -0.506)\n", - " trained lo -> evaluated hi : 0.430 (native hi = 0.627)\n", + "lo -0.557 -0.003 0.434\n", + "mid 0.355 0.269 0.450\n", + "hi 1.130 0.702 0.606\n", + " trained hi -> evaluated lo : 1.130 (native lo = -0.557)\n", + " trained lo -> evaluated hi : 0.434 (native hi = 0.606)\n", " best generalist (row mean over all eval regimes): trained on h_lo\n" ] } @@ -541,8 +542,10 @@ "metadata": {}, "source": [ "Transfer is clearly asymmetric, and the asymmetry runs in opposite directions\n", - "for the two tasks. For community detection the mid-homophily model is the best\n", - "generalist, and a high-homophily model degrades badly on heterophilic graphs.\n", + "for the two tasks. For community detection the high-homophily model is the\n", + "best generalist (row means 0.442 vs 0.437 for mid and 0.398 for low), even\n", + "though it degrades the most on heterophilic graphs (0.275 vs the native\n", + "0.347); its advantage on homophilous evaluations outweighs that loss.\n", "For triangle counting the low-homophily model is the safest, because it is\n", "calibrated to the sparse-triangle end of the scale.\n", "\n", @@ -557,10 +560,10 @@ "id": "cell-10", "metadata": { "execution": { - "iopub.execute_input": "2026-07-26T18:10:53.122575Z", - "iopub.status.busy": "2026-07-26T18:10:53.122505Z", - "iopub.status.idle": "2026-07-26T18:10:53.132719Z", - "shell.execute_reply": "2026-07-26T18:10:53.132222Z" + "iopub.execute_input": "2026-08-01T20:07:37.816355Z", + "iopub.status.busy": "2026-08-01T20:07:37.816257Z", + "iopub.status.idle": "2026-08-01T20:07:37.831321Z", + "shell.execute_reply": "2026-08-01T20:07:37.830698Z" } }, "outputs": [ @@ -570,18 +573,18 @@ "text": [ "axis degree homophily powerlaw\n", "task side \n", - "community_detection eval 0.026 0.722 0.001\n", - " train 0.001 0.035 0.000\n", - "triangle_counting eval 0.003 0.008 0.000\n", - " train 0.147 0.178 0.046\n", + "community_detection eval 0.025 0.810 0.000\n", + " train 0.002 0.022 0.000\n", + "triangle_counting eval 0.004 0.010 0.000\n", + " train 0.132 0.172 0.054\n", "\n", "Mean value by regime level (log10 scale for triangle counting):\n", - " community_detection eval_homophily {'hi': 0.53, 'lo': 0.273, 'mid': 0.38}\n", - " community_detection eval_degree {'hi': 0.414, 'lo': 0.374}\n", - " community_detection eval_powerlaw {'hi': 0.398, 'lo': 0.391}\n", - " triangle_counting train_homophily {'hi': 0.855, 'lo': -0.024, 'mid': 0.357}\n", - " triangle_counting train_degree {'hi': 0.723, 'lo': 0.069}\n", - " triangle_counting train_powerlaw {'hi': 0.212, 'lo': 0.58}\n" + " community_detection eval_homophily {'hi': 0.589, 'lo': 0.307, 'mid': 0.381}\n", + " community_detection eval_degree {'hi': 0.447, 'lo': 0.405}\n", + " community_detection eval_powerlaw {'hi': 0.427, 'lo': 0.424}\n", + " triangle_counting train_homophily {'hi': 0.812, 'lo': -0.042, 'mid': 0.358}\n", + " triangle_counting train_degree {'hi': 0.682, 'lo': 0.071}\n", + " triangle_counting train_powerlaw {'hi': 0.18, 'lo': 0.572}\n" ] } ], @@ -615,9 +618,10 @@ "id": "cell-11", "metadata": {}, "source": [ - "Homophily dominates in both cases — on the evaluation side for community\n", - "detection and on the training side for triangle counting. Average degree and the\n", - "power-law exponent are secondary throughout.\n", + "Homophily dominates the evaluation side of community detection (0.81 vs at\n", + "most 0.03 for any other axis). For triangle counting no single axis\n", + "dominates: on the training side homophily leads (0.17) with average degree\n", + "close behind (0.13), and the power-law exponent is secondary throughout.\n", "\n", "## 4. Seed stability\n", "\n", @@ -631,10 +635,10 @@ "id": "cell-12", "metadata": { "execution": { - "iopub.execute_input": "2026-07-26T18:10:53.133957Z", - "iopub.status.busy": "2026-07-26T18:10:53.133885Z", - "iopub.status.idle": "2026-07-26T18:10:53.139578Z", - "shell.execute_reply": "2026-07-26T18:10:53.139029Z" + "iopub.execute_input": "2026-08-01T20:07:37.833810Z", + "iopub.status.busy": "2026-08-01T20:07:37.833680Z", + "iopub.status.idle": "2026-08-01T20:07:37.841444Z", + "shell.execute_reply": "2026-08-01T20:07:37.841099Z" } }, "outputs": [ @@ -676,16 +680,16 @@ " \n", " community_detection\n", " accuracy\n", - " 0.0035\n", - " 0.0014\n", - " 0.3978\n", + " 0.0027\n", + " 0.0013\n", + " 0.4124\n", " \n", " \n", " triangle_counting\n", " MSE / triangles\n", - " 0.0549\n", - " 0.0065\n", - " 3.6515\n", + " 0.0538\n", + " 0.0146\n", + " 3.6532\n", " \n", " \n", "\n", @@ -694,13 +698,13 @@ "text/plain": [ " metric max_seed_std median_seed_std \\\n", "task \n", - "community_detection accuracy 0.0035 0.0014 \n", - "triangle_counting MSE / triangles 0.0549 0.0065 \n", + "community_detection accuracy 0.0027 0.0013 \n", + "triangle_counting MSE / triangles 0.0538 0.0146 \n", "\n", " spread_across_regimes \n", "task \n", - "community_detection 0.3978 \n", - "triangle_counting 3.6515 " + "community_detection 0.4124 \n", + "triangle_counting 3.6532 " ] }, "execution_count": 6, @@ -729,18 +733,18 @@ "## Takeaways\n", "\n", "1. **The two tasks fail differently.** For community detection, GREAD's score is\n", - " set by the regime it is *evaluated* on (variance explained 0.77)\n", - " and barely by the one it was *trained* on (0.04). The apparent\n", + " set by the regime it is *evaluated* on (variance explained 0.85)\n", + " and barely by the one it was *trained* on (0.03). The apparent\n", " OOD penalty is target difficulty, not failed transfer. For triangle counting\n", - " the ordering reverses on a log scale (train 0.43 vs eval\n", - " 0.11): that task does exhibit a real distribution-shift failure.\n", + " the ordering reverses on a log scale (train 0.41 vs eval\n", + " 0.12): that task does exhibit a real distribution-shift failure.\n", "2. **Counting fails by miscalibrated scale.** Triangle counts are extensive and\n", " their magnitude shifts by orders of magnitude across regimes, so a model\n", " calibrated on one regime mispredicts the scale on another. The worst cells of\n", " the transfer matrix are exactly the largest scale jumps.\n", "3. **Homophily transfer is asymmetric**, in opposite directions per task: the\n", " best community-detection generalist here is the model trained on\n", - " mid homophily, whereas the safest triangle counter is the\n", + " high homophily, whereas the safest triangle counter is the\n", " one trained on low homophily, which is calibrated to the sparse-triangle end\n", " of the scale.\n", "4. **Methodological note.** Any OOD summary of a regression task on this grid\n", diff --git a/2026_tdl_challenge/outputs/2026-07-25_13-46-02/results.json b/2026_tdl_challenge/outputs/2026-08-02_gread-vc/results.json similarity index 56% rename from 2026_tdl_challenge/outputs/2026-07-25_13-46-02/results.json rename to 2026_tdl_challenge/outputs/2026-08-02_gread-vc/results.json index 85e4e1c25..ac1289456 100644 --- a/2026_tdl_challenge/outputs/2026-07-25_13-46-02/results.json +++ b/2026_tdl_challenge/outputs/2026-08-02_gread-vc/results.json @@ -1,8 +1,8 @@ { "metadata": { - "study_id": "2026-07-25_13-46-02", + "study_id": "2026-08-02_gread-vc", "model_config": "graph/gread", - "generated_at_utc": "2026-07-25T17:01:06.057701+00:00", + "generated_at_utc": "2026-08-01T20:05:11.192867+00:00", "n_runs": 72, "train_seeds": [ 42, @@ -14,843 +14,843 @@ "results": [ { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_1.5-2__s42", "train_seed": 42, "homophily": "h_lo", "avg_degree": "d_lo", "power_law": "pl_lo", "run_slug": "h_lo__d_lo__pl_lo", - "test_loss": 2.2027204036712646, - "test_best_rerun_accuracy": 0.3146917223930359, + "test_loss": 2.12618350982666, + "test_best_rerun_accuracy": 0.33883413672447205, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3126671314239502, + "test_best_rerun_accuracy": 0.3303537368774414, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3287111222743988, + "test_best_rerun_accuracy": 0.37172433733940125, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.32661011815071106, + "test_best_rerun_accuracy": 0.36725494265556335, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.34807854890823364, + "test_best_rerun_accuracy": 0.35037052631378174, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.34383833408355713, + "test_best_rerun_accuracy": 0.34376195073127747, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.35598593950271606, + "test_best_rerun_accuracy": 0.3383757472038269, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.3577813506126404, + "test_best_rerun_accuracy": 0.3170219361782074, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.38031935691833496, + "test_best_rerun_accuracy": 0.43887996673583984, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3824203610420227, + "test_best_rerun_accuracy": 0.4400259852409363, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.39800596237182617, + "test_best_rerun_accuracy": 0.4566429853439331, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.40927496552467346, + "test_best_rerun_accuracy": 0.45859119296073914, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__00__h_lo__d_lo__pl_lo__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_community_detection__community_detection__00__h_lo__d_lo__pl_lo__s42" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_1.5-2__s43", "train_seed": 43, "homophily": "h_lo", "avg_degree": "d_lo", "power_law": "pl_lo", "run_slug": "h_lo__d_lo__pl_lo", - "test_loss": 2.1980178356170654, - "test_best_rerun_accuracy": 0.315952330827713, + "test_loss": 2.1327407360076904, + "test_best_rerun_accuracy": 0.3361601233482361, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3117503225803375, + "test_best_rerun_accuracy": 0.32874932885169983, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3279089331626892, + "test_best_rerun_accuracy": 0.37229734659194946, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.3265337347984314, + "test_best_rerun_accuracy": 0.3653831481933594, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3481549322605133, + "test_best_rerun_accuracy": 0.3535793423652649, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3421957492828369, + "test_best_rerun_accuracy": 0.3428451418876648, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3549163341522217, + "test_best_rerun_accuracy": 0.3394835293292999, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.35526013374328613, + "test_best_rerun_accuracy": 0.3188173174858093, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3750859498977661, + "test_best_rerun_accuracy": 0.4389563798904419, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3789823651313782, + "test_best_rerun_accuracy": 0.44036978483200073, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3924669623374939, + "test_best_rerun_accuracy": 0.4598899781703949, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.40297195315361023, + "test_best_rerun_accuracy": 0.4634425938129425, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__00__h_lo__d_lo__pl_lo__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_community_detection__community_detection__00__h_lo__d_lo__pl_lo__s43" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_1.5-2__s44", "train_seed": 44, "homophily": "h_lo", "avg_degree": "d_lo", "power_law": "pl_lo", "run_slug": "h_lo__d_lo__pl_lo", - "test_loss": 2.211678981781006, - "test_best_rerun_accuracy": 0.3149591386318207, + "test_loss": 2.135568141937256, + "test_best_rerun_accuracy": 0.33623653650283813, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.30919092893600464, + "test_best_rerun_accuracy": 0.3302391469478607, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.32416534423828125, + "test_best_rerun_accuracy": 0.36905035376548767, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.32252272963523865, + "test_best_rerun_accuracy": 0.36450454592704773, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.34425854682922363, + "test_best_rerun_accuracy": 0.3523569405078888, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.341508150100708, + "test_best_rerun_accuracy": 0.34257772564888, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3537321388721466, + "test_best_rerun_accuracy": 0.3391397297382355, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.35392314195632935, + "test_best_rerun_accuracy": 0.3185499310493469, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3761173486709595, + "test_best_rerun_accuracy": 0.43712276220321655, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3786003589630127, + "test_best_rerun_accuracy": 0.44201236963272095, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3949499726295471, + "test_best_rerun_accuracy": 0.45809459686279297, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.4067155718803406, + "test_best_rerun_accuracy": 0.46374818682670593, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__00__h_lo__d_lo__pl_lo__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_community_detection__community_detection__00__h_lo__d_lo__pl_lo__s44" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_4-5__s42", "train_seed": 42, "homophily": "h_lo", "avg_degree": "d_lo", "power_law": "pl_hi", "run_slug": "h_lo__d_lo__pl_hi", - "test_loss": 2.19852876663208, - "test_best_rerun_accuracy": 0.31450071930885315, + "test_loss": 2.1571204662323, + "test_best_rerun_accuracy": 0.3300863206386566, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.31545573472976685, + "test_best_rerun_accuracy": 0.3354725241661072, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3269539177417755, + "test_best_rerun_accuracy": 0.36316755414009094, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.3251967430114746, + "test_best_rerun_accuracy": 0.36183053255081177, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3446023464202881, + "test_best_rerun_accuracy": 0.3534265458583832, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3418901264667511, + "test_best_rerun_accuracy": 0.3449079394340515, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.35040873289108276, + "test_best_rerun_accuracy": 0.3383757472038269, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.3525097370147705, + "test_best_rerun_accuracy": 0.3239743411540985, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.37527695298194885, + "test_best_rerun_accuracy": 0.43456336855888367, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.37325236201286316, + "test_best_rerun_accuracy": 0.43712276220321655, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3888379633426666, + "test_best_rerun_accuracy": 0.45396897196769714, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.39930474758148193, + "test_best_rerun_accuracy": 0.45668119192123413, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__01__h_lo__d_lo__pl_hi__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_community_detection__community_detection__01__h_lo__d_lo__pl_hi__s42" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_4-5__s43", "train_seed": 43, "homophily": "h_lo", "avg_degree": "d_lo", "power_law": "pl_hi", "run_slug": "h_lo__d_lo__pl_hi", - "test_loss": 2.2020461559295654, - "test_best_rerun_accuracy": 0.3129345178604126, + "test_loss": 2.149336338043213, + "test_best_rerun_accuracy": 0.330048143863678, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.31438612937927246, + "test_best_rerun_accuracy": 0.3353961408138275, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.32512032985687256, + "test_best_rerun_accuracy": 0.36427533626556396, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.32443273067474365, + "test_best_rerun_accuracy": 0.3617541491985321, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3420047461986542, + "test_best_rerun_accuracy": 0.35205134749412537, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3402857482433319, + "test_best_rerun_accuracy": 0.3424249291419983, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.34536632895469666, + "test_best_rerun_accuracy": 0.3371533453464508, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.34700894355773926, + "test_best_rerun_accuracy": 0.32240813970565796, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3654213547706604, + "test_best_rerun_accuracy": 0.4278401732444763, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.36725494265556335, + "test_best_rerun_accuracy": 0.431774765253067, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3788677453994751, + "test_best_rerun_accuracy": 0.4464435875415802, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.3884941637516022, + "test_best_rerun_accuracy": 0.44747498631477356, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__01__h_lo__d_lo__pl_hi__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_community_detection__community_detection__01__h_lo__d_lo__pl_hi__s43" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_4-5__s44", "train_seed": 44, "homophily": "h_lo", "avg_degree": "d_lo", "power_law": "pl_hi", "run_slug": "h_lo__d_lo__pl_hi", - "test_loss": 2.215632200241089, - "test_best_rerun_accuracy": 0.31167393922805786, + "test_loss": 2.157198429107666, + "test_best_rerun_accuracy": 0.32867294549942017, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.31423333287239075, + "test_best_rerun_accuracy": 0.3344411253929138, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.321605920791626, + "test_best_rerun_accuracy": 0.36221253871917725, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.32145312428474426, + "test_best_rerun_accuracy": 0.36083734035491943, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.33856675028800964, + "test_best_rerun_accuracy": 0.3512491285800934, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.33528152108192444, + "test_best_rerun_accuracy": 0.34578654170036316, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.34658873081207275, + "test_best_rerun_accuracy": 0.3371151387691498, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.34498435258865356, + "test_best_rerun_accuracy": 0.32389792799949646, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3647337555885315, + "test_best_rerun_accuracy": 0.4320421814918518, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.36469554901123047, + "test_best_rerun_accuracy": 0.4342195689678192, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.37974634766578674, + "test_best_rerun_accuracy": 0.4488883912563324, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.3862403631210327, + "test_best_rerun_accuracy": 0.45522958040237427, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__01__h_lo__d_lo__pl_hi__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_community_detection__community_detection__01__h_lo__d_lo__pl_hi__s44" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_1.5-2__s42", "train_seed": 42, "homophily": "h_lo", "avg_degree": "d_hi", "power_law": "pl_lo", "run_slug": "h_lo__d_hi__pl_lo", - "test_loss": 2.15183687210083, - "test_best_rerun_accuracy": 0.32852011919021606, + "test_loss": 1.9829374551773071, + "test_best_rerun_accuracy": 0.3817327618598938, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3089617192745209, + "test_best_rerun_accuracy": 0.3265337347984314, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3028497099876404, + "test_best_rerun_accuracy": 0.3129727244377136, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.3258843421936035, + "test_best_rerun_accuracy": 0.3680189549922943, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.35934755206108093, + "test_best_rerun_accuracy": 0.36767515540122986, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.35392314195632935, + "test_best_rerun_accuracy": 0.35694095492362976, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.37584996223449707, + "test_best_rerun_accuracy": 0.3666437566280365, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.3787531554698944, + "test_best_rerun_accuracy": 0.3367713391780853, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.41007715463638306, + "test_best_rerun_accuracy": 0.5024065971374512, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.4181373715400696, + "test_best_rerun_accuracy": 0.5117655992507935, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.44556498527526855, + "test_best_rerun_accuracy": 0.5468332171440125, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.4679119884967804, + "test_best_rerun_accuracy": 0.5645580291748047, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__02__h_lo__d_hi__pl_lo__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_community_detection__community_detection__02__h_lo__d_hi__pl_lo__s42" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_1.5-2__s43", "train_seed": 43, "homophily": "h_lo", "avg_degree": "d_hi", "power_law": "pl_lo", "run_slug": "h_lo__d_hi__pl_lo", - "test_loss": 2.1434247493743896, - "test_best_rerun_accuracy": 0.3331805467605591, + "test_loss": 1.9749042987823486, + "test_best_rerun_accuracy": 0.381465345621109, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.31236153841018677, + "test_best_rerun_accuracy": 0.32714492082595825, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3061731159687042, + "test_best_rerun_accuracy": 0.31530293822288513, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.3290931284427643, + "test_best_rerun_accuracy": 0.3708457350730896, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3573993444442749, + "test_best_rerun_accuracy": 0.3721063435077667, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3535793423652649, + "test_best_rerun_accuracy": 0.35812515020370483, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3772633373737335, + "test_best_rerun_accuracy": 0.3664909601211548, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.37894415855407715, + "test_best_rerun_accuracy": 0.3397509455680847, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.40667736530303955, + "test_best_rerun_accuracy": 0.5073344111442566, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.4155779778957367, + "test_best_rerun_accuracy": 0.5170372128486633, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.4422033727169037, + "test_best_rerun_accuracy": 0.554969847202301, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.4629077911376953, + "test_best_rerun_accuracy": 0.5726182460784912, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__02__h_lo__d_hi__pl_lo__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_community_detection__community_detection__02__h_lo__d_hi__pl_lo__s43" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_1.5-2__s44", "train_seed": 44, "homophily": "h_lo", "avg_degree": "d_hi", "power_law": "pl_lo", "run_slug": "h_lo__d_hi__pl_lo", - "test_loss": 2.158991575241089, - "test_best_rerun_accuracy": 0.32771793007850647, + "test_loss": 1.9897323846817017, + "test_best_rerun_accuracy": 0.3790205419063568, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.30865612626075745, + "test_best_rerun_accuracy": 0.321644127368927, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.30120712518692017, + "test_best_rerun_accuracy": 0.31094813346862793, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.325464129447937, + "test_best_rerun_accuracy": 0.3659561574459076, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.35770493745803833, + "test_best_rerun_accuracy": 0.3692031502723694, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3555275499820709, + "test_best_rerun_accuracy": 0.3533119559288025, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3780273497104645, + "test_best_rerun_accuracy": 0.3675987422466278, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.378447562456131, + "test_best_rerun_accuracy": 0.34303614497184753, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.41294217109680176, + "test_best_rerun_accuracy": 0.5008785724639893, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.4188249707221985, + "test_best_rerun_accuracy": 0.5110397934913635, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.44690197706222534, + "test_best_rerun_accuracy": 0.5526396036148071, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.47127360105514526, + "test_best_rerun_accuracy": 0.5707082152366638, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__02__h_lo__d_hi__pl_lo__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_community_detection__community_detection__02__h_lo__d_hi__pl_lo__s44" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_4-5__s42", "train_seed": 42, "homophily": "h_lo", "avg_degree": "d_hi", "power_law": "pl_hi", "run_slug": "h_lo__d_hi__pl_hi", - "test_loss": 2.164921998977661, - "test_best_rerun_accuracy": 0.32554054260253906, + "test_loss": 2.026320219039917, + "test_best_rerun_accuracy": 0.36889755725860596, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.30854153633117676, + "test_best_rerun_accuracy": 0.3183589279651642, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3030407130718231, + "test_best_rerun_accuracy": 0.30823591351509094, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.32767972350120544, + "test_best_rerun_accuracy": 0.37172433733940125, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3594621419906616, + "test_best_rerun_accuracy": 0.37772175669670105, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.35365575551986694, + "test_best_rerun_accuracy": 0.3658033609390259, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3731759488582611, + "test_best_rerun_accuracy": 0.3750477433204651, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.37856215238571167, + "test_best_rerun_accuracy": 0.35759034752845764, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.41160517930984497, + "test_best_rerun_accuracy": 0.5195202231407166, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.4196653664112091, + "test_best_rerun_accuracy": 0.5330812335014343, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.44816258549690247, + "test_best_rerun_accuracy": 0.5726945996284485, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.47096797823905945, + "test_best_rerun_accuracy": 0.6026816368103027, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__03__h_lo__d_hi__pl_hi__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_community_detection__community_detection__03__h_lo__d_hi__pl_hi__s42" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_4-5__s43", "train_seed": 43, "homophily": "h_lo", "avg_degree": "d_hi", "power_law": "pl_hi", "run_slug": "h_lo__d_hi__pl_hi", - "test_loss": 2.1560213565826416, - "test_best_rerun_accuracy": 0.3271067440509796, + "test_loss": 2.0091536045074463, + "test_best_rerun_accuracy": 0.3725265562534332, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.31098631024360657, + "test_best_rerun_accuracy": 0.32332491874694824, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.30808311700820923, + "test_best_rerun_accuracy": 0.3122851252555847, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3285965323448181, + "test_best_rerun_accuracy": 0.3753533363342285, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.35812515020370483, + "test_best_rerun_accuracy": 0.38299334049224854, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.35254794359207153, + "test_best_rerun_accuracy": 0.36614716053009033, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.37363433837890625, + "test_best_rerun_accuracy": 0.37661394476890564, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.3764611482620239, + "test_best_rerun_accuracy": 0.3592711389064789, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.4082435667514801, + "test_best_rerun_accuracy": 0.5208954215049744, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.414622962474823, + "test_best_rerun_accuracy": 0.5351822376251221, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.43914738297462463, + "test_best_rerun_accuracy": 0.5756742358207703, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.45943158864974976, + "test_best_rerun_accuracy": 0.6003132462501526, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__03__h_lo__d_hi__pl_hi__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_community_detection__community_detection__03__h_lo__d_hi__pl_hi__s43" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_4-5__s44", "train_seed": 44, "homophily": "h_lo", "avg_degree": "d_hi", "power_law": "pl_hi", "run_slug": "h_lo__d_hi__pl_hi", - "test_loss": 2.1702499389648438, - "test_best_rerun_accuracy": 0.32649552822113037, + "test_loss": 2.0166783332824707, + "test_best_rerun_accuracy": 0.37256476283073425, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.30800673365592957, + "test_best_rerun_accuracy": 0.3214913308620453, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.302620530128479, + "test_best_rerun_accuracy": 0.3109099268913269, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3253113329410553, + "test_best_rerun_accuracy": 0.3748185634613037, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.35911834239959717, + "test_best_rerun_accuracy": 0.38123616576194763, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.35365575551986694, + "test_best_rerun_accuracy": 0.3663763403892517, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.37393996119499207, + "test_best_rerun_accuracy": 0.37474215030670166, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.3790205419063568, + "test_best_rerun_accuracy": 0.35720834136009216, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.41603636741638184, + "test_best_rerun_accuracy": 0.5196729898452759, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.41970357298851013, + "test_best_rerun_accuracy": 0.5318205952644348, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.45056918263435364, + "test_best_rerun_accuracy": 0.5702880024909973, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.47345098853111267, + "test_best_rerun_accuracy": 0.5955764651298523, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__03__h_lo__d_hi__pl_hi__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_community_detection__community_detection__03__h_lo__d_hi__pl_hi__s44" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_1.5-2__s42", "train_seed": 42, "homophily": "h_mid", "avg_degree": "d_lo", "power_law": "pl_lo", "run_slug": "h_mid__d_lo__pl_lo", - "test_loss": 1.9939359426498413, - "test_best_rerun_accuracy": 0.38845595717430115, + "test_loss": 1.9429086446762085, + "test_best_rerun_accuracy": 0.4033157527446747, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.28497210144996643, + "test_best_rerun_accuracy": 0.2992207109928131, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.2793949246406555, + "test_best_rerun_accuracy": 0.2894033193588257, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.30689892172813416, + "test_best_rerun_accuracy": 0.3318053185939789, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.30074870586395264, + "test_best_rerun_accuracy": 0.33367714285850525, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3811979591846466, + "test_best_rerun_accuracy": 0.38929635286331177, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.4097333550453186, + "test_best_rerun_accuracy": 0.4090839624404907, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.4174497723579407, + "test_best_rerun_accuracy": 0.40488195419311523, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.5027503967285156, + "test_best_rerun_accuracy": 0.5576056241989136, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.516808032989502, + "test_best_rerun_accuracy": 0.5723890066146851, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.5733058452606201, + "test_best_rerun_accuracy": 0.6230422258377075, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.6190312504768372, + "test_best_rerun_accuracy": 0.6621208786964417, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__04__h_mid__d_lo__pl_lo__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_community_detection__community_detection__04__h_mid__d_lo__pl_lo__s42" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_1.5-2__s43", "train_seed": 43, "homophily": "h_mid", "avg_degree": "d_lo", "power_law": "pl_lo", "run_slug": "h_mid__d_lo__pl_lo", - "test_loss": 1.9884096384048462, - "test_best_rerun_accuracy": 0.3901367485523224, + "test_loss": 1.9408735036849976, + "test_best_rerun_accuracy": 0.40312474966049194, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.2869967222213745, + "test_best_rerun_accuracy": 0.3021621108055115, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.2819543182849884, + "test_best_rerun_accuracy": 0.29138970375061035, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.30487433075904846, + "test_best_rerun_accuracy": 0.33379173278808594, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.29983192682266235, + "test_best_rerun_accuracy": 0.3366185426712036, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3789059519767761, + "test_best_rerun_accuracy": 0.39116814732551575, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { @@ -858,4415 +858,4415 @@ "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.41767895221710205, + "test_best_rerun_accuracy": 0.4002215564250946, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.5052334070205688, + "test_best_rerun_accuracy": 0.5564213991165161, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.5202460289001465, + "test_best_rerun_accuracy": 0.5712048411369324, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.5740698575973511, + "test_best_rerun_accuracy": 0.6235770583152771, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.6203300356864929, + "test_best_rerun_accuracy": 0.6574986577033997, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__04__h_mid__d_lo__pl_lo__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_community_detection__community_detection__04__h_mid__d_lo__pl_lo__s43" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_1.5-2__s44", "train_seed": 44, "homophily": "h_mid", "avg_degree": "d_lo", "power_law": "pl_lo", "run_slug": "h_mid__d_lo__pl_lo", - "test_loss": 1.9952137470245361, - "test_best_rerun_accuracy": 0.3864695429801941, + "test_loss": 1.9397659301757812, + "test_best_rerun_accuracy": 0.40461456775665283, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.2871495187282562, + "test_best_rerun_accuracy": 0.30021393299102783, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.27836352586746216, + "test_best_rerun_accuracy": 0.2915043234825134, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.30632591247558594, + "test_best_rerun_accuracy": 0.3329513370990753, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.2990679144859314, + "test_best_rerun_accuracy": 0.335778146982193, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3788677453994751, + "test_best_rerun_accuracy": 0.39143556356430054, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.40977156162261963, + "test_best_rerun_accuracy": 0.41290396451950073, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.4180227816104889, + "test_best_rerun_accuracy": 0.4056841731071472, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.5043166279792786, + "test_best_rerun_accuracy": 0.5565360188484192, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.5172663927078247, + "test_best_rerun_accuracy": 0.5707846283912659, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.5742226243019104, + "test_best_rerun_accuracy": 0.624264657497406, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.6211704611778259, + "test_best_rerun_accuracy": 0.6609748601913452, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__04__h_mid__d_lo__pl_lo__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_community_detection__community_detection__04__h_mid__d_lo__pl_lo__s44" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_4-5__s42", "train_seed": 42, "homophily": "h_mid", "avg_degree": "d_lo", "power_law": "pl_hi", "run_slug": "h_mid__d_lo__pl_hi", - "test_loss": 2.0197396278381348, - "test_best_rerun_accuracy": 0.37963175773620605, + "test_loss": 1.9772688150405884, + "test_best_rerun_accuracy": 0.3893727660179138, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.2844373285770416, + "test_best_rerun_accuracy": 0.3001757264137268, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.2800443172454834, + "test_best_rerun_accuracy": 0.2939109206199646, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3004431128501892, + "test_best_rerun_accuracy": 0.33165252208709717, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.3001757264137268, + "test_best_rerun_accuracy": 0.33337152004241943, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.38249674439430237, + "test_best_rerun_accuracy": 0.399419367313385, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3996867537498474, + "test_best_rerun_accuracy": 0.3965543508529663, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.4174497723579407, + "test_best_rerun_accuracy": 0.3988463580608368, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.4986630082130432, + "test_best_rerun_accuracy": 0.5517610311508179, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.5124531984329224, + "test_best_rerun_accuracy": 0.566315233707428, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.5634502172470093, + "test_best_rerun_accuracy": 0.6116204261779785, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.6109328269958496, + "test_best_rerun_accuracy": 0.6505462527275085, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__05__h_mid__d_lo__pl_hi__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_community_detection__community_detection__05__h_mid__d_lo__pl_hi__s42" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_4-5__s43", "train_seed": 43, "homophily": "h_mid", "avg_degree": "d_lo", "power_law": "pl_hi", "run_slug": "h_mid__d_lo__pl_hi", - "test_loss": 2.0160462856292725, - "test_best_rerun_accuracy": 0.37542974948883057, + "test_loss": 1.993801474571228, + "test_best_rerun_accuracy": 0.38792115449905396, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.2864237129688263, + "test_best_rerun_accuracy": 0.30025210976600647, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.282756507396698, + "test_best_rerun_accuracy": 0.29360532760620117, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3004431128501892, + "test_best_rerun_accuracy": 0.3337535262107849, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.29994651675224304, + "test_best_rerun_accuracy": 0.3354725241661072, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3821147382259369, + "test_best_rerun_accuracy": 0.39922836422920227, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3976621627807617, + "test_best_rerun_accuracy": 0.3977767527103424, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.41431736946105957, + "test_best_rerun_accuracy": 0.4036213755607605, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.4986630082130432, + "test_best_rerun_accuracy": 0.5467950105667114, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.5134463906288147, + "test_best_rerun_accuracy": 0.5651692152023315, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.5636412501335144, + "test_best_rerun_accuracy": 0.6097104549407959, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.6115058660507202, + "test_best_rerun_accuracy": 0.6521888375282288, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__05__h_mid__d_lo__pl_hi__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_community_detection__community_detection__05__h_mid__d_lo__pl_hi__s43" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_4-5__s44", "train_seed": 44, "homophily": "h_mid", "avg_degree": "d_lo", "power_law": "pl_hi", "run_slug": "h_mid__d_lo__pl_hi", - "test_loss": 2.011624574661255, - "test_best_rerun_accuracy": 0.37676674127578735, + "test_loss": 1.9878644943237305, + "test_best_rerun_accuracy": 0.39200854301452637, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.28894492983818054, + "test_best_rerun_accuracy": 0.302582323551178, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.2828711271286011, + "test_best_rerun_accuracy": 0.2940255105495453, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3039957284927368, + "test_best_rerun_accuracy": 0.33191993832588196, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.3015127182006836, + "test_best_rerun_accuracy": 0.33489954471588135, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.382267564535141, + "test_best_rerun_accuracy": 0.39972496032714844, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3988845646381378, + "test_best_rerun_accuracy": 0.3977385461330414, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.4143555760383606, + "test_best_rerun_accuracy": 0.4018641710281372, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.5001146197319031, + "test_best_rerun_accuracy": 0.5473679900169373, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.5137901902198792, + "test_best_rerun_accuracy": 0.5636793971061707, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.5611964464187622, + "test_best_rerun_accuracy": 0.6095576286315918, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.6087936162948608, + "test_best_rerun_accuracy": 0.6524562835693359, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__05__h_mid__d_lo__pl_hi__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_community_detection__community_detection__05__h_mid__d_lo__pl_hi__s44" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_1.5-2__s42", "train_seed": 42, "homophily": "h_mid", "avg_degree": "d_hi", "power_law": "pl_lo", "run_slug": "h_mid__d_hi__pl_lo", - "test_loss": 1.9068946838378906, - "test_best_rerun_accuracy": 0.4208495616912842, + "test_loss": 1.8586361408233643, + "test_best_rerun_accuracy": 0.4339521825313568, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.26747649908065796, + "test_best_rerun_accuracy": 0.2720223069190979, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.2562839090824127, + "test_best_rerun_accuracy": 0.2612498998641968, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.29287952184677124, + "test_best_rerun_accuracy": 0.31247612833976746, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.2860417068004608, + "test_best_rerun_accuracy": 0.3133929371833801, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3847505450248718, + "test_best_rerun_accuracy": 0.3969745635986328, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.37118953466415405, + "test_best_rerun_accuracy": 0.3794025480747223, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.4337993860244751, + "test_best_rerun_accuracy": 0.4368171691894531, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.5080602169036865, + "test_best_rerun_accuracy": 0.5506914258003235, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.5256704092025757, + "test_best_rerun_accuracy": 0.5674230456352234, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.5920238494873047, + "test_best_rerun_accuracy": 0.6389716267585754, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.6447016596794128, + "test_best_rerun_accuracy": 0.6830926537513733, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__06__h_mid__d_hi__pl_lo__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_community_detection__community_detection__06__h_mid__d_hi__pl_lo__s42" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_1.5-2__s43", "train_seed": 43, "homophily": "h_mid", "avg_degree": "d_hi", "power_law": "pl_lo", "run_slug": "h_mid__d_hi__pl_lo", - "test_loss": 1.9076277017593384, - "test_best_rerun_accuracy": 0.4181755781173706, + "test_loss": 1.858083724975586, + "test_best_rerun_accuracy": 0.4343723654747009, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.2659102976322174, + "test_best_rerun_accuracy": 0.27251890301704407, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.25605469942092896, + "test_best_rerun_accuracy": 0.2636183202266693, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.2917335033416748, + "test_best_rerun_accuracy": 0.31262892484664917, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.28539231419563293, + "test_best_rerun_accuracy": 0.3133929371833801, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3804721534252167, + "test_best_rerun_accuracy": 0.3981969654560089, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.37172433733940125, + "test_best_rerun_accuracy": 0.3818855583667755, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.43116357922554016, + "test_best_rerun_accuracy": 0.43578577041625977, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.5079838037490845, + "test_best_rerun_accuracy": 0.5514172315597534, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.520628035068512, + "test_best_rerun_accuracy": 0.5699060559272766, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.5925968289375305, + "test_best_rerun_accuracy": 0.6407670378684998, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.6452746391296387, + "test_best_rerun_accuracy": 0.6848116517066956, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__06__h_mid__d_hi__pl_lo__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_community_detection__community_detection__06__h_mid__d_hi__pl_lo__s43" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_1.5-2__s44", "train_seed": 44, "homophily": "h_mid", "avg_degree": "d_hi", "power_law": "pl_lo", "run_slug": "h_mid__d_hi__pl_lo", - "test_loss": 1.9092298746109009, - "test_best_rerun_accuracy": 0.4204675555229187, + "test_loss": 1.8485448360443115, + "test_best_rerun_accuracy": 0.4346015751361847, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.2661777138710022, + "test_best_rerun_accuracy": 0.27370309829711914, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.2574681043624878, + "test_best_rerun_accuracy": 0.26025670766830444, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.2918481230735779, + "test_best_rerun_accuracy": 0.3146153390407562, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.28611811995506287, + "test_best_rerun_accuracy": 0.3146153390407562, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.37963175773620605, + "test_best_rerun_accuracy": 0.39796775579452515, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.36885935068130493, + "test_best_rerun_accuracy": 0.3818473517894745, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.43185117840766907, + "test_best_rerun_accuracy": 0.4347161650657654, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.5057681798934937, + "test_best_rerun_accuracy": 0.5539002418518066, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.5177248120307922, + "test_best_rerun_accuracy": 0.5712048411369324, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.5874398350715637, + "test_best_rerun_accuracy": 0.6423714756965637, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.6438612341880798, + "test_best_rerun_accuracy": 0.6863396763801575, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__06__h_mid__d_hi__pl_lo__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_community_detection__community_detection__06__h_mid__d_hi__pl_lo__s44" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_4-5__s42", "train_seed": 42, "homophily": "h_mid", "avg_degree": "d_hi", "power_law": "pl_hi", "run_slug": "h_mid__d_hi__pl_hi", - "test_loss": 1.8688710927963257, - "test_best_rerun_accuracy": 0.4370845854282379, + "test_loss": 1.8221337795257568, + "test_best_rerun_accuracy": 0.4431583881378174, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.2569715082645416, + "test_best_rerun_accuracy": 0.2592635154724121, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.24967530369758606, + "test_best_rerun_accuracy": 0.2465810924768448, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.2807319164276123, + "test_best_rerun_accuracy": 0.2977309226989746, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.27687370777130127, + "test_best_rerun_accuracy": 0.2982275187969208, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.37825655937194824, + "test_best_rerun_accuracy": 0.38666054606437683, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3698907494544983, + "test_best_rerun_accuracy": 0.3724883496761322, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.4148903787136078, + "test_best_rerun_accuracy": 0.4268851578235626, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.5055007934570312, + "test_best_rerun_accuracy": 0.5498892068862915, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.5232256054878235, + "test_best_rerun_accuracy": 0.5669264197349548, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.5895026326179504, + "test_best_rerun_accuracy": 0.6368706822395325, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.6494002342224121, + "test_best_rerun_accuracy": 0.6908854842185974, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__07__h_mid__d_hi__pl_hi__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_community_detection__community_detection__07__h_mid__d_hi__pl_hi__s42" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_4-5__s43", "train_seed": 43, "homophily": "h_mid", "avg_degree": "d_hi", "power_law": "pl_hi", "run_slug": "h_mid__d_hi__pl_hi", - "test_loss": 1.8615974187850952, - "test_best_rerun_accuracy": 0.43505996465682983, + "test_loss": 1.8122713565826416, + "test_best_rerun_accuracy": 0.4466727674007416, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.2612881064414978, + "test_best_rerun_accuracy": 0.258002907037735, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.2546412944793701, + "test_best_rerun_accuracy": 0.24883489310741425, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.2845519185066223, + "test_best_rerun_accuracy": 0.3008633255958557, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.28050270676612854, + "test_best_rerun_accuracy": 0.2982275187969208, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3750477433204651, + "test_best_rerun_accuracy": 0.38899075984954834, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3707311451435089, + "test_best_rerun_accuracy": 0.3786385655403137, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.4139735698699951, + "test_best_rerun_accuracy": 0.43089616298675537, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.5062648057937622, + "test_best_rerun_accuracy": 0.5534800291061401, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.522843599319458, + "test_best_rerun_accuracy": 0.5708610415458679, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.5922530293464661, + "test_best_rerun_accuracy": 0.6460004448890686, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.6482542753219604, + "test_best_rerun_accuracy": 0.6990602612495422, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__07__h_mid__d_hi__pl_hi__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_community_detection__community_detection__07__h_mid__d_hi__pl_hi__s43" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_4-5__s44", "train_seed": 44, "homophily": "h_mid", "avg_degree": "d_hi", "power_law": "pl_hi", "run_slug": "h_mid__d_hi__pl_hi", - "test_loss": 1.8566935062408447, - "test_best_rerun_accuracy": 0.434486985206604, + "test_loss": 1.8095487356185913, + "test_best_rerun_accuracy": 0.4467109739780426, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.26243409514427185, + "test_best_rerun_accuracy": 0.26105889678001404, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.2577736973762512, + "test_best_rerun_accuracy": 0.24956069886684418, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.28581252694129944, + "test_best_rerun_accuracy": 0.3012453317642212, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.2823745012283325, + "test_best_rerun_accuracy": 0.3008251190185547, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.378447562456131, + "test_best_rerun_accuracy": 0.3903277516365051, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.37149515748023987, + "test_best_rerun_accuracy": 0.3787531554698944, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.41523417830467224, + "test_best_rerun_accuracy": 0.43357017636299133, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.5106959939002991, + "test_best_rerun_accuracy": 0.5560775995254517, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.5280770063400269, + "test_best_rerun_accuracy": 0.5717778205871582, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.5913744568824768, + "test_best_rerun_accuracy": 0.6448162794113159, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.6498204469680786, + "test_best_rerun_accuracy": 0.6958515048027039, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__07__h_mid__d_hi__pl_hi__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_community_detection__community_detection__07__h_mid__d_hi__pl_hi__s44" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_1.5-2__s42", "train_seed": 42, "homophily": "h_hi", "avg_degree": "d_lo", "power_law": "pl_lo", "run_slug": "h_hi__d_lo__pl_lo", - "test_loss": 1.5320980548858643, - "test_best_rerun_accuracy": 0.5589044094085693, + "test_loss": 1.3534891605377197, + "test_best_rerun_accuracy": 0.6125754714012146, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.2361142933368683, + "test_best_rerun_accuracy": 0.2691954970359802, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.22996409237384796, + "test_best_rerun_accuracy": 0.26014211773872375, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.23332569003105164, + "test_best_rerun_accuracy": 0.3091527223587036, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.22285889089107513, + "test_best_rerun_accuracy": 0.3112919330596924, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3816945552825928, + "test_best_rerun_accuracy": 0.3944151699542999, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3701963424682617, + "test_best_rerun_accuracy": 0.3784857392311096, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.39090076088905334, + "test_best_rerun_accuracy": 0.37294673919677734, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.4064863622188568, + "test_best_rerun_accuracy": 0.368171751499176, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.5736878514289856, + "test_best_rerun_accuracy": 0.6270150542259216, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.6494766473770142, + "test_best_rerun_accuracy": 0.6884024739265442, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.6999006867408752, + "test_best_rerun_accuracy": 0.7243868708610535, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__08__h_hi__d_lo__pl_lo__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_community_detection__community_detection__08__h_hi__d_lo__pl_lo__s42" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_1.5-2__s43", "train_seed": 43, "homophily": "h_hi", "avg_degree": "d_lo", "power_law": "pl_lo", "run_slug": "h_hi__d_lo__pl_lo", - "test_loss": 1.5363527536392212, - "test_best_rerun_accuracy": 0.5583696365356445, + "test_loss": 1.3625643253326416, + "test_best_rerun_accuracy": 0.6104744672775269, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.2374894917011261, + "test_best_rerun_accuracy": 0.2709527015686035, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.22618229687213898, + "test_best_rerun_accuracy": 0.2596836984157562, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.2348918914794922, + "test_best_rerun_accuracy": 0.310184121131897, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.2233554869890213, + "test_best_rerun_accuracy": 0.3111773133277893, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3791733384132385, + "test_best_rerun_accuracy": 0.39292535185813904, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3693177402019501, + "test_best_rerun_accuracy": 0.3769959509372711, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.38627853989601135, + "test_best_rerun_accuracy": 0.3753151595592499, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.4054931700229645, + "test_best_rerun_accuracy": 0.37294673919677734, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.5715868473052979, + "test_best_rerun_accuracy": 0.6266330480575562, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.6474138498306274, + "test_best_rerun_accuracy": 0.686721682548523, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.6968064904212952, + "test_best_rerun_accuracy": 0.724004864692688, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__08__h_hi__d_lo__pl_lo__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_community_detection__community_detection__08__h_hi__d_lo__pl_lo__s43" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_1.5-2__s44", "train_seed": 44, "homophily": "h_hi", "avg_degree": "d_lo", "power_law": "pl_lo", "run_slug": "h_hi__d_lo__pl_lo", - "test_loss": 1.545002818107605, - "test_best_rerun_accuracy": 0.5550844073295593, + "test_loss": 1.3738481998443604, + "test_best_rerun_accuracy": 0.6108946204185486, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.23928490281105042, + "test_best_rerun_accuracy": 0.2619374990463257, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.2279776930809021, + "test_best_rerun_accuracy": 0.25357168912887573, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.23454809188842773, + "test_best_rerun_accuracy": 0.31029871106147766, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.2241576910018921, + "test_best_rerun_accuracy": 0.3108717203140259, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3764229416847229, + "test_best_rerun_accuracy": 0.39090076088905334, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3692031502723694, + "test_best_rerun_accuracy": 0.37623193860054016, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.38750094175338745, + "test_best_rerun_accuracy": 0.3755825459957123, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.4048437476158142, + "test_best_rerun_accuracy": 0.3733287453651428, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.5676904320716858, + "test_best_rerun_accuracy": 0.6251814365386963, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.644090473651886, + "test_best_rerun_accuracy": 0.6858812570571899, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.6955077052116394, + "test_best_rerun_accuracy": 0.7237374782562256, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__08__h_hi__d_lo__pl_lo__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_community_detection__community_detection__08__h_hi__d_lo__pl_lo__s44" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_4-5__s42", "train_seed": 42, "homophily": "h_hi", "avg_degree": "d_lo", "power_law": "pl_hi", "run_slug": "h_hi__d_lo__pl_hi", - "test_loss": 1.4838958978652954, - "test_best_rerun_accuracy": 0.5757888555526733, + "test_loss": 1.2887213230133057, + "test_best_rerun_accuracy": 0.631790041923523, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.23183588683605194, + "test_best_rerun_accuracy": 0.2641149163246155, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.2233554869890213, + "test_best_rerun_accuracy": 0.25521430373191833, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.23046068847179413, + "test_best_rerun_accuracy": 0.3088853359222412, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.21770188212394714, + "test_best_rerun_accuracy": 0.31236153841018677, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.377645343542099, + "test_best_rerun_accuracy": 0.39414775371551514, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.36668193340301514, + "test_best_rerun_accuracy": 0.3794025480747223, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3863167464733124, + "test_best_rerun_accuracy": 0.3743601441383362, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.40316295623779297, + "test_best_rerun_accuracy": 0.3718007504940033, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.5545114278793335, + "test_best_rerun_accuracy": 0.6168538331985474, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.6455420851707458, + "test_best_rerun_accuracy": 0.6918022632598877, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.6988692879676819, + "test_best_rerun_accuracy": 0.728703498840332, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__09__h_hi__d_lo__pl_hi__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_community_detection__community_detection__09__h_hi__d_lo__pl_hi__s42" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_4-5__s43", "train_seed": 43, "homophily": "h_hi", "avg_degree": "d_lo", "power_law": "pl_hi", "run_slug": "h_hi__d_lo__pl_hi", - "test_loss": 1.4814717769622803, - "test_best_rerun_accuracy": 0.5724654197692871, + "test_loss": 1.3131598234176636, + "test_best_rerun_accuracy": 0.6270914673805237, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.23351669311523438, + "test_best_rerun_accuracy": 0.26273971796035767, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.2237756848335266, + "test_best_rerun_accuracy": 0.25043928623199463, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.2314920872449875, + "test_best_rerun_accuracy": 0.30430132150650024, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.21605928242206573, + "test_best_rerun_accuracy": 0.30854153633117676, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.375773549079895, + "test_best_rerun_accuracy": 0.3867751657962799, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3675605356693268, + "test_best_rerun_accuracy": 0.37393996119499207, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.38517075777053833, + "test_best_rerun_accuracy": 0.37054014205932617, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.4035067558288574, + "test_best_rerun_accuracy": 0.37485674023628235, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.5566506385803223, + "test_best_rerun_accuracy": 0.6096722483634949, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.6483688354492188, + "test_best_rerun_accuracy": 0.6884024739265442, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.7006646990776062, + "test_best_rerun_accuracy": 0.7274047136306763, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__09__h_hi__d_lo__pl_hi__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_community_detection__community_detection__09__h_hi__d_lo__pl_hi__s43" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_4-5__s44", "train_seed": 44, "homophily": "h_hi", "avg_degree": "d_lo", "power_law": "pl_hi", "run_slug": "h_hi__d_lo__pl_hi", - "test_loss": 1.4923393726348877, - "test_best_rerun_accuracy": 0.5687600374221802, + "test_loss": 1.2969404458999634, + "test_best_rerun_accuracy": 0.631866455078125, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.23638169467449188, + "test_best_rerun_accuracy": 0.2638475000858307, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.2240430861711502, + "test_best_rerun_accuracy": 0.2528840899467468, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.23279088735580444, + "test_best_rerun_accuracy": 0.3087325096130371, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.21892428398132324, + "test_best_rerun_accuracy": 0.30582931637763977, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.37783634662628174, + "test_best_rerun_accuracy": 0.3931163549423218, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3678279519081116, + "test_best_rerun_accuracy": 0.37963175773620605, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.38547635078430176, + "test_best_rerun_accuracy": 0.37485674023628235, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.4063717722892761, + "test_best_rerun_accuracy": 0.37256476283073425, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.5565360188484192, + "test_best_rerun_accuracy": 0.6139506697654724, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.644166886806488, + "test_best_rerun_accuracy": 0.6902360916137695, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.6990220546722412, + "test_best_rerun_accuracy": 0.728703498840332, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__09__h_hi__d_lo__pl_hi__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_community_detection__community_detection__09__h_hi__d_lo__pl_hi__s44" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_1.5-2__s42", "train_seed": 42, "homophily": "h_hi", "avg_degree": "d_hi", "power_law": "pl_lo", "run_slug": "h_hi__d_hi__pl_lo", - "test_loss": 1.2372453212738037, - "test_best_rerun_accuracy": 0.6587210893630981, + "test_loss": 1.0757602453231812, + "test_best_rerun_accuracy": 0.704217255115509, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.219153493642807, + "test_best_rerun_accuracy": 0.2512415051460266, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.210176482796669, + "test_best_rerun_accuracy": 0.23997250199317932, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.22014668583869934, + "test_best_rerun_accuracy": 0.2966613173484802, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.20246008038520813, + "test_best_rerun_accuracy": 0.2991825342178345, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3768049478530884, + "test_best_rerun_accuracy": 0.39685994386672974, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.36060813069343567, + "test_best_rerun_accuracy": 0.3777981400489807, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.40205517411231995, + "test_best_rerun_accuracy": 0.40041255950927734, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.41225457191467285, + "test_best_rerun_accuracy": 0.39391854405403137, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.5612727999687195, + "test_best_rerun_accuracy": 0.6141416430473328, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.5721980333328247, + "test_best_rerun_accuracy": 0.6303766369819641, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.7097944617271423, + "test_best_rerun_accuracy": 0.7415004968643188, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__10__h_hi__d_hi__pl_lo__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_community_detection__community_detection__10__h_hi__d_hi__pl_lo__s42" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_1.5-2__s43", "train_seed": 43, "homophily": "h_hi", "avg_degree": "d_hi", "power_law": "pl_lo", "run_slug": "h_hi__d_hi__pl_lo", - "test_loss": 1.237139105796814, - "test_best_rerun_accuracy": 0.6592176556587219, + "test_loss": 1.080308198928833, + "test_best_rerun_accuracy": 0.7037206888198853, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.21976469457149506, + "test_best_rerun_accuracy": 0.2544884979724884, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.20639468729496002, + "test_best_rerun_accuracy": 0.24077469110488892, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.2190006822347641, + "test_best_rerun_accuracy": 0.2994499206542969, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.20177248120307922, + "test_best_rerun_accuracy": 0.2993735074996948, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3734433352947235, + "test_best_rerun_accuracy": 0.39483535289764404, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3631293475627899, + "test_best_rerun_accuracy": 0.3779127597808838, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.39796775579452515, + "test_best_rerun_accuracy": 0.39880815148353577, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.4101153612136841, + "test_best_rerun_accuracy": 0.3957521617412567, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.5592482089996338, + "test_best_rerun_accuracy": 0.6109710335731506, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.572350800037384, + "test_best_rerun_accuracy": 0.6293834447860718, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.7077316641807556, + "test_best_rerun_accuracy": 0.7383680939674377, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__10__h_hi__d_hi__pl_lo__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_community_detection__community_detection__10__h_hi__d_hi__pl_lo__s43" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_1.5-2__s44", "train_seed": 44, "homophily": "h_hi", "avg_degree": "d_hi", "power_law": "pl_lo", "run_slug": "h_hi__d_hi__pl_lo", - "test_loss": 1.240233063697815, - "test_best_rerun_accuracy": 0.6589120626449585, + "test_loss": 1.0855813026428223, + "test_best_rerun_accuracy": 0.7029566764831543, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.22045229375362396, + "test_best_rerun_accuracy": 0.2512032985687256, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.20849567651748657, + "test_best_rerun_accuracy": 0.23794789612293243, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.22094888985157013, + "test_best_rerun_accuracy": 0.29532432556152344, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.20184887945652008, + "test_best_rerun_accuracy": 0.2945603132247925, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3726411461830139, + "test_best_rerun_accuracy": 0.39418596029281616, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3610283434391022, + "test_best_rerun_accuracy": 0.37691953778266907, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.3950263559818268, + "test_best_rerun_accuracy": 0.4022461473941803, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.4074031710624695, + "test_best_rerun_accuracy": 0.3946443498134613, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.5573764443397522, + "test_best_rerun_accuracy": 0.6097486615180969, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.5696386098861694, + "test_best_rerun_accuracy": 0.6269004344940186, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.7117044925689697, + "test_best_rerun_accuracy": 0.740927517414093, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__10__h_hi__d_hi__pl_lo__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_community_detection__community_detection__10__h_hi__d_hi__pl_lo__s44" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_4-5__s42", "train_seed": 42, "homophily": "h_hi", "avg_degree": "d_hi", "power_law": "pl_hi", "run_slug": "h_hi__d_hi__pl_hi", - "test_loss": 1.0778677463531494, - "test_best_rerun_accuracy": 0.7097944617271423, + "test_loss": 0.9378951191902161, + "test_best_rerun_accuracy": 0.7414240837097168, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.21006187796592712, + "test_best_rerun_accuracy": 0.23565588891506195, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.2036442756652832, + "test_best_rerun_accuracy": 0.22495989501476288, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.21036748588085175, + "test_best_rerun_accuracy": 0.2869585156440735, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.1941324770450592, + "test_best_rerun_accuracy": 0.28501030802726746, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.3679807484149933, + "test_best_rerun_accuracy": 0.3824203610420227, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3512873351573944, + "test_best_rerun_accuracy": 0.3650393486022949, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.39132094383239746, + "test_best_rerun_accuracy": 0.39823517203330994, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.4114905595779419, + "test_best_rerun_accuracy": 0.40778517723083496, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.5508823990821838, + "test_best_rerun_accuracy": 0.6010008454322815, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.5668882131576538, + "test_best_rerun_accuracy": 0.6173886656761169, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.651654064655304, + "test_best_rerun_accuracy": 0.694438099861145, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__11__h_hi__d_hi__pl_hi__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_community_detection__community_detection__11__h_hi__d_hi__pl_hi__s42" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_4-5__s43", "train_seed": 43, "homophily": "h_hi", "avg_degree": "d_hi", "power_law": "pl_hi", "run_slug": "h_hi__d_hi__pl_hi", - "test_loss": 1.0654218196868896, - "test_best_rerun_accuracy": 0.7102910876274109, + "test_loss": 0.9291359782218933, + "test_best_rerun_accuracy": 0.7421116828918457, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.21136067807674408, + "test_best_rerun_accuracy": 0.23810069262981415, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.20058828592300415, + "test_best_rerun_accuracy": 0.22690808773040771, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.21216288208961487, + "test_best_rerun_accuracy": 0.28256550431251526, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.1924898773431778, + "test_best_rerun_accuracy": 0.28527772426605225, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.36641454696655273, + "test_best_rerun_accuracy": 0.38601115345954895, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.3533119559288025, + "test_best_rerun_accuracy": 0.3704637587070465, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.39338377118110657, + "test_best_rerun_accuracy": 0.39361295104026794, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.4098479747772217, + "test_best_rerun_accuracy": 0.4067537486553192, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.5567651987075806, + "test_best_rerun_accuracy": 0.6050118207931519, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.570135235786438, + "test_best_rerun_accuracy": 0.6179998517036438, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.655512273311615, + "test_best_rerun_accuracy": 0.6978760957717896, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__11__h_hi__d_hi__pl_hi__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_community_detection__community_detection__11__h_hi__d_hi__pl_hi__s43" }, { "experiment": "community_detection", - "wandb_project": "challenge_community_detection", + "wandb_project": "shard_cd", "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_4-5__s44", "train_seed": 44, "homophily": "h_hi", "avg_degree": "d_hi", "power_law": "pl_hi", "run_slug": "h_hi__d_hi__pl_hi", - "test_loss": 1.0614569187164307, - "test_best_rerun_accuracy": 0.7123920917510986, + "test_loss": 0.9276243448257446, + "test_best_rerun_accuracy": 0.7423790693283081, "test_best_rerun_mse": null, "test_triangles_total_structural": null, "test_mse_by_total_triangles": null, "ood_test": { "h_lo__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.21170447766780853, + "test_best_rerun_accuracy": 0.24023990333080292, "test_best_rerun_mse": null }, "h_lo__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.20203988254070282, + "test_best_rerun_accuracy": 0.22912369668483734, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.210176482796669, + "test_best_rerun_accuracy": 0.2887539267539978, "test_best_rerun_mse": null }, "h_lo__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.19550767540931702, + "test_best_rerun_accuracy": 0.2856215238571167, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.368171751499176, + "test_best_rerun_accuracy": 0.3883795440196991, "test_best_rerun_mse": null }, "h_mid__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.35610052943229675, + "test_best_rerun_accuracy": 0.37241193652153015, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.39296355843544006, + "test_best_rerun_accuracy": 0.3981969654560089, "test_best_rerun_mse": null }, "h_mid__d_hi__pl_hi": { - "test_best_rerun_accuracy": 0.4139353632926941, + "test_best_rerun_accuracy": 0.4085109531879425, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_lo": { - "test_best_rerun_accuracy": 0.5557720065116882, + "test_best_rerun_accuracy": 0.6047062277793884, "test_best_rerun_mse": null }, "h_hi__d_lo__pl_hi": { - "test_best_rerun_accuracy": 0.5720834136009216, + "test_best_rerun_accuracy": 0.6205974221229553, "test_best_rerun_mse": null }, "h_hi__d_hi__pl_lo": { - "test_best_rerun_accuracy": 0.6558942794799805, + "test_best_rerun_accuracy": 0.698296308517456, "test_best_rerun_mse": null } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__community_detection__11__h_hi__d_hi__pl_hi__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_community_detection__community_detection__11__h_hi__d_hi__pl_hi__s44" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_1.5-2__s42", "train_seed": 42, "homophily": "h_lo", "avg_degree": "d_lo", "power_law": "pl_lo", "run_slug": "h_lo__d_lo__pl_lo", - "test_loss": 87.79856872558594, + "test_loss": 103.7693099975586, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 84.7706298828125, + "test_best_rerun_mse": 99.37831115722656, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 0.061073940837761166, + "test_mse_by_total_triangles": 0.07159820688560992, "ood_test": { "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 26.20136070251465, + "test_best_rerun_mse": 27.330459594726562, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 0.13790189843428763 + "test_mse_by_total_triangles": 0.14384452418277138 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 10183.357421875, + "test_best_rerun_mse": 10635.9609375, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.7723441351441032 + "test_mse_by_total_triangles": 0.8066712883959044 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 391.2452087402344, + "test_best_rerun_mse": 354.98895263671875, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 0.1318655910819799 + "test_mse_by_total_triangles": 0.11964575417482937 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 7205.154296875, + "test_best_rerun_mse": 7336.1064453125, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 0.962612464512358 + "test_mse_by_total_triangles": 0.9801077415247161 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 163.98965454101562, + "test_best_rerun_mse": 153.3765411376953, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 0.14161455487134336 + "test_mse_by_total_triangles": 0.13244951739006502 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 170734.9375, + "test_best_rerun_mse": 170894.453125, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 3.964679024242987 + "test_mse_by_total_triangles": 3.968383176783392 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 15777.7197265625, + "test_best_rerun_mse": 15196.38671875, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 1.1062768003479526 + "test_mse_by_total_triangles": 1.0655158265846305 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 45563.41015625, + "test_best_rerun_mse": 46248.25390625, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 2.3355072098134193 + "test_mse_by_total_triangles": 2.3706112002793582 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 3589.44384765625, + "test_best_rerun_mse": 3671.532470703125, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.6381233506944445 + "test_mse_by_total_triangles": 0.6527168836805556 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 752714.5625, + "test_best_rerun_mse": 756233.25, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 8.514581660124657 + "test_mse_by_total_triangles": 8.5543844665905 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 261834.453125, + "test_best_rerun_mse": 262601.53125, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 4.357153963440001 + "test_mse_by_total_triangles": 4.369918813339324 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__00__h_lo__d_lo__pl_lo__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_triangle_counting__triangle_counting__00__h_lo__d_lo__pl_lo__s42" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_1.5-2__s43", "train_seed": 43, "homophily": "h_lo", "avg_degree": "d_lo", "power_law": "pl_lo", "run_slug": "h_lo__d_lo__pl_lo", - "test_loss": 94.0346908569336, + "test_loss": 66.18878936767578, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 88.72521209716797, + "test_best_rerun_mse": 65.18682098388672, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 0.0639230634705821, + "test_mse_by_total_triangles": 0.04696456843219504, "ood_test": { "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 39.16325378417969, + "test_best_rerun_mse": 27.961711883544922, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 0.20612238833778782 + "test_mse_by_total_triangles": 0.14716690465023644 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 10529.6162109375, + "test_best_rerun_mse": 8967.6962890625, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.7986057042804323 + "test_mse_by_total_triangles": 0.6801438216960561 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 440.15911865234375, + "test_best_rerun_mse": 183.81419372558594, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 0.1483515735262365 + "test_mse_by_total_triangles": 0.06195287958395212 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 7313.99267578125, + "test_best_rerun_mse": 7049.56494140625, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 0.9771533300976953 + "test_mse_by_total_triangles": 0.9418256434744489 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 176.3068084716797, + "test_best_rerun_mse": 153.53924560546875, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 0.15225112994100146 + "test_mse_by_total_triangles": 0.13259002211180376 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 171964.015625, + "test_best_rerun_mse": 165857.828125, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 3.9932197572218096 + "test_mse_by_total_triangles": 3.8514264379760355 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 15931.330078125, + "test_best_rerun_mse": 14364.3720703125, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 1.117047404159655 + "test_mse_by_total_triangles": 1.007177960336033 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 45739.92578125, + "test_best_rerun_mse": 46258.82421875, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 2.3445551171895023 + "test_mse_by_total_triangles": 2.371153017517556 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 3617.634521484375, + "test_best_rerun_mse": 3738.55419921875, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.6431350260416666 + "test_mse_by_total_triangles": 0.6646318576388889 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 754751.875, + "test_best_rerun_mse": 753396.875, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 8.537627399522641 + "test_mse_by_total_triangles": 8.522299865389183 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 262293.9375, + "test_best_rerun_mse": 261510.90625, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 4.364800184713694 + "test_mse_by_total_triangles": 4.351769860882299 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__00__h_lo__d_lo__pl_lo__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_triangle_counting__triangle_counting__00__h_lo__d_lo__pl_lo__s43" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_1.5-2__s44", "train_seed": 44, "homophily": "h_lo", "avg_degree": "d_lo", "power_law": "pl_lo", "run_slug": "h_lo__d_lo__pl_lo", - "test_loss": 91.22958374023438, + "test_loss": 87.03919219970703, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 93.32726287841797, + "test_best_rerun_mse": 89.05577850341797, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 0.06723866201615128, + "test_mse_by_total_triangles": 0.06416122370563254, "ood_test": { "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 32.397438049316406, + "test_best_rerun_mse": 28.453157424926758, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 0.1705128318385074 + "test_mse_by_total_triangles": 0.14975346013119345 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 10556.314453125, + "test_best_rerun_mse": 9831.416015625, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.8006305994027304 + "test_mse_by_total_triangles": 0.7456515749431172 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 368.49139404296875, + "test_best_rerun_mse": 255.51260375976562, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 0.12419662758441818 + "test_mse_by_total_triangles": 0.08611816776534062 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 7264.04150390625, + "test_best_rerun_mse": 7185.37060546875, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 0.9704798268411824 + "test_mse_by_total_triangles": 0.9599693527680361 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 163.54283142089844, + "test_best_rerun_mse": 167.41357421875, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 0.14122869725466186 + "test_mse_by_total_triangles": 0.14457130761550085 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 171505.5, + "test_best_rerun_mse": 168127.921875, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 3.9825724503065207 + "test_mse_by_total_triangles": 3.904140857212521 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 15698.0771484375, + "test_best_rerun_mse": 14739.2236328125, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 1.1006925500236642 + "test_mse_by_total_triangles": 1.0334611998886902 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 45547.19140625, + "test_best_rerun_mse": 46426.19140625, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 2.3346758627428366 + "test_mse_by_total_triangles": 2.379731990683787 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 3584.070068359375, + "test_best_rerun_mse": 3757.74462890625, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.6371680121527777 + "test_mse_by_total_triangles": 0.6680434895833334 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 753381.3125, + "test_best_rerun_mse": 754443.6875, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 8.522123824983314 + "test_mse_by_total_triangles": 8.534141233894777 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 261664.484375, + "test_best_rerun_mse": 261495.125, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 4.354325535004077 + "test_mse_by_total_triangles": 4.3515072471003275 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__00__h_lo__d_lo__pl_lo__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_triangle_counting__triangle_counting__00__h_lo__d_lo__pl_lo__s44" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_4-5__s42", "train_seed": 42, "homophily": "h_lo", "avg_degree": "d_lo", "power_law": "pl_hi", "run_slug": "h_lo__d_lo__pl_hi", - "test_loss": 3.94545316696167, + "test_loss": 1.9937427043914795, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 3.997734785079956, + "test_best_rerun_mse": 2.0566940307617188, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 0.021040709395157665, + "test_mse_by_total_triangles": 0.010824705425061677, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 214.29408264160156, + "test_best_rerun_mse": 207.88951110839844, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 0.15439054945360342 + "test_mse_by_total_triangles": 0.14977630483314008 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 12880.9287109375, + "test_best_rerun_mse": 12903.9619140625, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.9769380895667426 + "test_mse_by_total_triangles": 0.9786850143392112 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 327.1950378417969, + "test_best_rerun_mse": 345.58203125, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 0.11027807139932487 + "test_mse_by_total_triangles": 0.11647523803505225 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 8220.9306640625, + "test_best_rerun_mse": 8225.1513671875, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 1.09832073000167 + "test_mse_by_total_triangles": 1.0988846181947227 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 182.99462890625, + "test_best_rerun_mse": 178.46937561035156, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 0.15802644983268566 + "test_mse_by_total_triangles": 0.15411863178786836 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 177818.703125, + "test_best_rerun_mse": 178153.328125, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 4.12917293156697 + "test_mse_by_total_triangles": 4.136943343047557 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 15884.2216796875, + "test_best_rerun_mse": 15967.8134765625, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 1.1137443331711891 + "test_mse_by_total_triangles": 1.1196054884702356 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 47834.33203125, + "test_best_rerun_mse": 47949.5390625, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 2.4519110170306013 + "test_mse_by_total_triangles": 2.457816344379517 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 4011.482177734375, + "test_best_rerun_mse": 4009.41455078125, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.7131523871527777 + "test_mse_by_total_triangles": 0.7127848090277777 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 764179.0, + "test_best_rerun_mse": 765415.875, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 8.644265466104091 + "test_mse_by_total_triangles": 8.658256789927943 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 265889.375, + "test_best_rerun_mse": 266707.03125, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 4.424631404656116 + "test_mse_by_total_triangles": 4.438237918725974 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__01__h_lo__d_lo__pl_hi__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_triangle_counting__triangle_counting__01__h_lo__d_lo__pl_hi__s42" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_4-5__s43", "train_seed": 43, "homophily": "h_lo", "avg_degree": "d_lo", "power_law": "pl_hi", "run_slug": "h_lo__d_lo__pl_hi", - "test_loss": 2.637273073196411, + "test_loss": 3.697826862335205, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 2.6410555839538574, + "test_best_rerun_mse": 3.676438808441162, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 0.013900292547125565, + "test_mse_by_total_triangles": 0.01934967793916401, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 217.00003051757812, + "test_best_rerun_mse": 192.7432098388672, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 0.15634007962361537 + "test_mse_by_total_triangles": 0.13886398403376599 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 12980.216796875, + "test_best_rerun_mse": 12582.099609375, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.984468471511187 + "test_mse_by_total_triangles": 0.9542737663538111 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 341.5283203125, + "test_best_rerun_mse": 313.11309814453125, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 0.11510897213094035 + "test_mse_by_total_triangles": 0.10553188343260238 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 8230.470703125, + "test_best_rerun_mse": 8129.0234375, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 1.0995952843186372 + "test_mse_by_total_triangles": 1.0860418754175016 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 179.60137939453125, + "test_best_rerun_mse": 177.4161834716797, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 0.15509618255140867 + "test_mse_by_total_triangles": 0.1532091394401379 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 178184.984375, + "test_best_rerun_mse": 177003.765625, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 4.137678440809029 + "test_mse_by_total_triangles": 4.110249062441947 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 15936.5751953125, + "test_best_rerun_mse": 15768.5087890625, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 1.1174151728588206 + "test_mse_by_total_triangles": 1.1056309626323446 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 47872.2578125, + "test_best_rerun_mse": 47807.42578125, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 2.453855031652058 + "test_mse_by_total_triangles": 2.450531845878825 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 4002.09375, + "test_best_rerun_mse": 4005.360107421875, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.7114833333333334 + "test_mse_by_total_triangles": 0.7120640190972222 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 764736.4375, + "test_best_rerun_mse": 764026.0, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 8.650571106184179 + "test_mse_by_total_triangles": 8.64253475560784 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 266129.59375, + "test_best_rerun_mse": 266085.96875, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 4.428628854442281 + "test_mse_by_total_triangles": 4.427902896344 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__01__h_lo__d_lo__pl_hi__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_triangle_counting__triangle_counting__01__h_lo__d_lo__pl_hi__s43" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0-0.1__deg_1-2.5__gamma_4-5__s44", "train_seed": 44, "homophily": "h_lo", "avg_degree": "d_lo", "power_law": "pl_hi", "run_slug": "h_lo__d_lo__pl_hi", - "test_loss": 3.575599193572998, + "test_loss": 2.7298684120178223, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 3.5890095233917236, + "test_best_rerun_mse": 2.851917266845703, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 0.01888952380732486, + "test_mse_by_total_triangles": 0.01501009087813528, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 219.34445190429688, + "test_best_rerun_mse": 196.38729858398438, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 0.15802914402326865 + "test_mse_by_total_triangles": 0.14148940820171785 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 12829.2109375, + "test_best_rerun_mse": 12713.6474609375, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.9730156190747061 + "test_mse_by_total_triangles": 0.9642508502796738 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 297.767822265625, + "test_best_rerun_mse": 330.8277282714844, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 0.10035989965137344 + "test_mse_by_total_triangles": 0.11150243622227313 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 8212.6416015625, + "test_best_rerun_mse": 8166.9814453125, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 1.0972133068219774 + "test_mse_by_total_triangles": 1.0911130855460922 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 180.3816375732422, + "test_best_rerun_mse": 178.12557983398438, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 0.15576998063319705 + "test_mse_by_total_triangles": 0.15382174424350983 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 177597.78125, + "test_best_rerun_mse": 177430.03125, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 4.124042849015419 + "test_mse_by_total_triangles": 4.120147483977336 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 15731.7119140625, + "test_best_rerun_mse": 15846.5185546875, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 1.103050898475845 + "test_mse_by_total_triangles": 1.1111007260333403 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 47818.60546875, + "test_best_rerun_mse": 47947.03125, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 2.4511048987005997 + "test_mse_by_total_triangles": 2.4576877979394127 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 3994.02001953125, + "test_best_rerun_mse": 4023.205078125, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.7100480034722222 + "test_mse_by_total_triangles": 0.7152364583333334 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 763774.0, + "test_best_rerun_mse": 765036.125, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 8.639684173614018 + "test_mse_by_total_triangles": 8.653961121228917 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 265265.75, + "test_best_rerun_mse": 266491.375, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 4.414253740036277 + "test_mse_by_total_triangles": 4.434649210390561 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__01__h_lo__d_lo__pl_hi__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_triangle_counting__triangle_counting__01__h_lo__d_lo__pl_hi__s44" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_1.5-2__s42", "train_seed": 42, "homophily": "h_lo", "avg_degree": "d_hi", "power_law": "pl_lo", "run_slug": "h_lo__d_hi__pl_lo", - "test_loss": 3781.752685546875, + "test_loss": 3718.912841796875, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 3639.166748046875, + "test_best_rerun_mse": 3540.4287109375, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.2760080961734452, + "test_mse_by_total_triangles": 0.2685194320013273, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 4683.3271484375, + "test_best_rerun_mse": 6885.89794921875, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 3.3741550060788903 + "test_mse_by_total_triangles": 4.961021577246938 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 3266.08447265625, + "test_best_rerun_mse": 4862.27392578125, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 17.18991827713816 + "test_mse_by_total_triangles": 25.590915398848683 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 1367.878662109375, + "test_best_rerun_mse": 1078.793701171875, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 0.4610308938690175 + "test_mse_by_total_triangles": 0.36359747258910513 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 5900.775390625, + "test_best_rerun_mse": 7947.81689453125, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 0.7883467455744823 + "test_mse_by_total_triangles": 1.0618325844397127 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 3303.512451171875, + "test_best_rerun_mse": 5710.18359375, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 2.8527741374541233 + "test_mse_by_total_triangles": 4.931073915155441 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 133401.6875, + "test_best_rerun_mse": 116063.1328125, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 3.0977542146572543 + "test_mse_by_total_triangles": 2.6951312653840795 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 10316.59375, + "test_best_rerun_mse": 7656.59765625, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 0.7233623439910251 + "test_mse_by_total_triangles": 0.5368530119373159 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 33247.765625, + "test_best_rerun_mse": 28981.654296875, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 1.704227055461582 + "test_mse_by_total_triangles": 1.4855530420254754 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 4507.5302734375, + "test_best_rerun_mse": 5319.77734375, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.8013387152777778 + "test_mse_by_total_triangles": 0.9457381944444444 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 672685.25, + "test_best_rerun_mse": 626065.375, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 7.609303417304843 + "test_mse_by_total_triangles": 7.081947162426615 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 221016.734375, + "test_best_rerun_mse": 194029.90625, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 3.677911476794302 + "test_mse_by_total_triangles": 3.2288270888456228 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__02__h_lo__d_hi__pl_lo__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_triangle_counting__triangle_counting__02__h_lo__d_hi__pl_lo__s42" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_1.5-2__s43", "train_seed": 43, "homophily": "h_lo", "avg_degree": "d_hi", "power_law": "pl_lo", "run_slug": "h_lo__d_hi__pl_lo", - "test_loss": 4730.625, + "test_loss": 2281.53662109375, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 4473.701171875, + "test_best_rerun_mse": 2323.655517578125, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.33930232627038304, + "test_mse_by_total_triangles": 0.1762347756979996, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 7426.0400390625, + "test_best_rerun_mse": 1048.82421875, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 5.350172938805835 + "test_mse_by_total_triangles": 0.7556370452089337 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 8192.68359375, + "test_best_rerun_mse": 1730.8511962890625, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 43.11938733552632 + "test_mse_by_total_triangles": 9.109743138363488 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 4050.074951171875, + "test_best_rerun_mse": 1190.6556396484375, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 1.3650404284367628 + "test_mse_by_total_triangles": 0.4012995078019675 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 7471.56494140625, + "test_best_rerun_mse": 4363.033203125, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 0.998205068992151 + "test_mse_by_total_triangles": 0.5829035675517702 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 7668.984375, + "test_best_rerun_mse": 1287.7236328125, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 6.622611722797927 + "test_mse_by_total_triangles": 1.1120238625323835 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 130463.4375, + "test_best_rerun_mse": 108183.5625, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 3.029524370704068 + "test_mse_by_total_triangles": 2.512157776797325 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 8364.19140625, + "test_best_rerun_mse": 10481.5517578125, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 0.5864669335471884 + "test_mse_by_total_triangles": 0.7349286045303954 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 31084.86328125, + "test_best_rerun_mse": 44005.53515625, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 1.5933601558895893 + "test_mse_by_total_triangles": 2.255653039943103 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 6533.189453125, + "test_best_rerun_mse": 5179.33203125, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 1.1614559027777778 + "test_mse_by_total_triangles": 0.9207701388888889 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 657630.0, + "test_best_rerun_mse": 713468.625, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 7.439000938882165 + "test_mse_by_total_triangles": 8.070638157076118 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 206809.140625, + "test_best_rerun_mse": 250656.609375, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 3.4414847091175345 + "test_mse_by_total_triangles": 4.171144881683391 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__02__h_lo__d_hi__pl_lo__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_triangle_counting__triangle_counting__02__h_lo__d_hi__pl_lo__s43" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_1.5-2__s44", "train_seed": 44, "homophily": "h_lo", "avg_degree": "d_hi", "power_law": "pl_lo", "run_slug": "h_lo__d_hi__pl_lo", - "test_loss": 2995.67431640625, + "test_loss": 3504.039794921875, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 3030.6328125, + "test_best_rerun_mse": 3353.7314453125, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.22985459328782706, + "test_mse_by_total_triangles": 0.2543596090491088, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 1809.7470703125, + "test_best_rerun_mse": 7857.9091796875, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 1.3038523561329252 + "test_mse_by_total_triangles": 5.661317852800792 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 856.2131958007812, + "test_best_rerun_mse": 5539.9296875, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 4.506385241056743 + "test_mse_by_total_triangles": 29.15752467105263 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 1417.048095703125, + "test_best_rerun_mse": 840.58740234375, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 0.4776029982147371 + "test_mse_by_total_triangles": 0.2833122353703236 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 5545.8681640625, + "test_best_rerun_mse": 7941.345703125, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 0.7409309504425518 + "test_mse_by_total_triangles": 1.0609680298096191 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 862.7684936523438, + "test_best_rerun_mse": 6454.15771484375, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 0.7450505126531466 + "test_mse_by_total_triangles": 5.573538613854707 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 134738.71875, + "test_best_rerun_mse": 114943.015625, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 3.1288017543655955 + "test_mse_by_total_triangles": 2.669120741802898 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 11481.5048828125, + "test_best_rerun_mse": 7433.1689453125, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 0.8050417110371968 + "test_mse_by_total_triangles": 0.5211869965862082 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 38070.41015625, + "test_best_rerun_mse": 28244.72265625, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 1.9514280668537598 + "test_mse_by_total_triangles": 1.4477791099620687 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 3524.151611328125, + "test_best_rerun_mse": 6064.890625, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.6265158420138889 + "test_mse_by_total_triangles": 1.0782027777777778 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 689925.4375, + "test_best_rerun_mse": 620222.875, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 7.80432154451772 + "test_mse_by_total_triangles": 7.015857776319808 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 228730.859375, + "test_best_rerun_mse": 191844.9375, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 3.806281253640191 + "test_mse_by_total_triangles": 3.1924673006839397 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__02__h_lo__d_hi__pl_lo__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_triangle_counting__triangle_counting__02__h_lo__d_hi__pl_lo__s44" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_4-5__s42", "train_seed": 42, "homophily": "h_lo", "avg_degree": "d_hi", "power_law": "pl_hi", "run_slug": "h_lo__d_hi__pl_hi", - "test_loss": 141.59347534179688, + "test_loss": 133.80874633789062, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 137.14686584472656, + "test_best_rerun_mse": 132.99447631835938, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 0.04622408690418826, + "test_mse_by_total_triangles": 0.04482456229132436, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 162.5874786376953, + "test_best_rerun_mse": 139.5034942626953, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 0.11713795290900239 + "test_mse_by_total_triangles": 0.10050684024689864 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 202.92503356933594, + "test_best_rerun_mse": 117.73936462402344, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 1.0680264924701892 + "test_mse_by_total_triangles": 0.6196808664422286 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 10650.7060546875, + "test_best_rerun_mse": 10639.1103515625, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.8077896135523321 + "test_mse_by_total_triangles": 0.8069101518060295 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 7001.46728515625, + "test_best_rerun_mse": 7254.41162109375, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 0.9353997708959586 + "test_mse_by_total_triangles": 0.9691932693512024 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 209.96044921875, + "test_best_rerun_mse": 159.9431915283203, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 0.181312995871114 + "test_mse_by_total_triangles": 0.13812019993810043 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 169328.984375, + "test_best_rerun_mse": 169733.859375, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 3.9320310323007615 + "test_mse_by_total_triangles": 3.941432736740665 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 13579.9501953125, + "test_best_rerun_mse": 13573.05859375, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 0.9521771277038634 + "test_mse_by_total_triangles": 0.9516939134588417 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 44206.00390625, + "test_best_rerun_mse": 45418.02734375, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 2.2659287460274746 + "test_mse_by_total_triangles": 2.328055120393152 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 3260.388427734375, + "test_best_rerun_mse": 3447.152099609375, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.579624609375 + "test_mse_by_total_triangles": 0.6128270399305555 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 746189.25, + "test_best_rerun_mse": 750142.0, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 8.440768412836668 + "test_mse_by_total_triangles": 8.485481261948124 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 254947.46875, + "test_best_rerun_mse": 256900.21875, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 4.2425485289467995 + "test_mse_by_total_triangles": 4.275043994308821 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__03__h_lo__d_hi__pl_hi__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_triangle_counting__triangle_counting__03__h_lo__d_hi__pl_hi__s42" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_4-5__s43", "train_seed": 43, "homophily": "h_lo", "avg_degree": "d_hi", "power_law": "pl_hi", "run_slug": "h_lo__d_hi__pl_hi", - "test_loss": 136.92222595214844, + "test_loss": 145.23092651367188, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 133.60507202148438, + "test_best_rerun_mse": 140.07705688476562, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 0.04503035794455153, + "test_mse_by_total_triangles": 0.04721168078354082, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 155.73580932617188, + "test_best_rerun_mse": 126.6436996459961, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 0.11220159173355322 + "test_mse_by_total_triangles": 0.09124185853457932 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 176.49093627929688, + "test_best_rerun_mse": 89.44583129882812, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 0.9288996646278783 + "test_mse_by_total_triangles": 0.470767533151727 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 10851.1845703125, + "test_best_rerun_mse": 10464.98046875, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.822994658347554 + "test_mse_by_total_triangles": 0.7937034864429275 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 7126.55322265625, + "test_best_rerun_mse": 7308.833984375, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 0.952111318992151 + "test_mse_by_total_triangles": 0.9764641261690047 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 188.96495056152344, + "test_best_rerun_mse": 153.86282348632812, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 0.1631821680151325 + "test_mse_by_total_triangles": 0.1328694503336167 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 170130.109375, + "test_best_rerun_mse": 169703.109375, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 3.950634157881293 + "test_mse_by_total_triangles": 3.9407186832389 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 13553.0517578125, + "test_best_rerun_mse": 13670.0849609375, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 0.9502911062833053 + "test_mse_by_total_triangles": 0.9584970523725634 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 44547.57421875, + "test_best_rerun_mse": 45648.53125, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 2.2834370915346764 + "test_mse_by_total_triangles": 2.3398703803372802 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 3278.650390625, + "test_best_rerun_mse": 3530.629150390625, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.5828711805555555 + "test_mse_by_total_triangles": 0.6276674045138889 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 748030.75, + "test_best_rerun_mse": 751388.0625, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 8.461599153874868 + "test_mse_by_total_triangles": 8.4995765132405 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 255422.21875, + "test_best_rerun_mse": 257795.140625, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 4.2504487835521605 + "test_mse_by_total_triangles": 4.289936275855757 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__03__h_lo__d_hi__pl_hi__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_triangle_counting__triangle_counting__03__h_lo__d_hi__pl_hi__s43" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0-0.1__deg_4-5__gamma_4-5__s44", "train_seed": 44, "homophily": "h_lo", "avg_degree": "d_hi", "power_law": "pl_hi", "run_slug": "h_lo__d_hi__pl_hi", - "test_loss": 137.71493530273438, + "test_loss": 145.36387634277344, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 136.60125732421875, + "test_best_rerun_mse": 144.36087036132812, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 0.04604019458180612, + "test_mse_by_total_triangles": 0.04865550062734349, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 157.6785125732422, + "test_best_rerun_mse": 153.95831298828125, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 0.11360123384239351 + "test_mse_by_total_triangles": 0.11092097477541876 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 61.32759094238281, + "test_best_rerun_mse": 153.75437927246094, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 0.32277679443359375 + "test_mse_by_total_triangles": 0.8092335751182155 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 11010.056640625, + "test_best_rerun_mse": 10420.1181640625, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.8350441138130451 + "test_mse_by_total_triangles": 0.7903009604901403 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 7674.73095703125, + "test_best_rerun_mse": 7109.88427734375, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 1.025348157252004 + "test_mse_by_total_triangles": 0.949884338990481 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 155.83151245117188, + "test_best_rerun_mse": 186.3960418701172, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 0.13456952715990664 + "test_mse_by_total_triangles": 0.1609637667272169 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 171492.6875, + "test_best_rerun_mse": 169028.328125, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 3.9822749280141183 + "test_mse_by_total_triangles": 3.925049417727104 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 13858.89453125, + "test_best_rerun_mse": 13583.720703125, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 0.9717356984469219 + "test_mse_by_total_triangles": 0.9524415021122563 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 46144.7890625, + "test_best_rerun_mse": 44931.9921875, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 2.3653077585985955 + "test_mse_by_total_triangles": 2.3031417390691478 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 3617.7861328125, + "test_best_rerun_mse": 3381.36962890625, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.6431619791666666 + "test_mse_by_total_triangles": 0.6011323784722222 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 751923.75, + "test_best_rerun_mse": 748325.125, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 8.505636120946122 + "test_mse_by_total_triangles": 8.464929074805154 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 256966.0, + "test_best_rerun_mse": 256234.21875, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 4.276138651756444 + "test_mse_by_total_triangles": 4.263961172682342 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__03__h_lo__d_hi__pl_hi__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_triangle_counting__triangle_counting__03__h_lo__d_hi__pl_hi__s44" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_1.5-2__s42", "train_seed": 42, "homophily": "h_mid", "avg_degree": "d_lo", "power_law": "pl_lo", "run_slug": "h_mid__d_lo__pl_lo", - "test_loss": 5257.2451171875, + "test_loss": 5268.91015625, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 5374.14990234375, + "test_best_rerun_mse": 5439.87353515625, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 0.7179892989103206, + "test_mse_by_total_triangles": 0.7267700113769205, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 1289.4329833984375, + "test_best_rerun_mse": 1550.5223388671875, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 0.9289862992784131 + "test_mse_by_total_triangles": 1.1170910222386077 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 1900.9085693359375, + "test_best_rerun_mse": 2207.710693359375, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 10.004781943873356 + "test_mse_by_total_triangles": 11.619529965049342 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 6492.744140625, + "test_best_rerun_mse": 6727.1220703125, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.49243414035836175 + "test_mse_by_total_triangles": 0.5102102442406143 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 1523.965576171875, + "test_best_rerun_mse": 1256.0379638671875, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 0.5136385494344035 + "test_mse_by_total_triangles": 0.4233360174813574 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 1454.740234375, + "test_best_rerun_mse": 1797.516845703125, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 1.256252361291019 + "test_mse_by_total_triangles": 1.5522597976710923 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 150097.625, + "test_best_rerun_mse": 150585.78125, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 3.485454788222181 + "test_mse_by_total_triangles": 3.4967903875626973 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 9604.4755859375, + "test_best_rerun_mse": 9768.8115234375, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 0.6734311867856893 + "test_mse_by_total_triangles": 0.6849538299984224 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 37404.9296875, + "test_best_rerun_mse": 36949.6328125, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 1.9173166070787842 + "test_mse_by_total_triangles": 1.89397882067251 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 2736.078857421875, + "test_best_rerun_mse": 2867.781005859375, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.4864140190972222 + "test_mse_by_total_triangles": 0.509827734375 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 706084.8125, + "test_best_rerun_mse": 706071.125, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 7.987113700892504 + "test_mse_by_total_triangles": 7.986958870174089 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 230611.890625, + "test_best_rerun_mse": 231655.515625, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 3.837583256369294 + "test_mse_by_total_triangles": 3.854950087780607 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__04__h_mid__d_lo__pl_lo__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_triangle_counting__triangle_counting__04__h_mid__d_lo__pl_lo__s42" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_1.5-2__s43", "train_seed": 43, "homophily": "h_mid", "avg_degree": "d_lo", "power_law": "pl_lo", "run_slug": "h_mid__d_lo__pl_lo", - "test_loss": 5255.0380859375, + "test_loss": 4706.82421875, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 5427.216796875, + "test_best_rerun_mse": 4811.73388671875, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 0.7250790643787575, + "test_mse_by_total_triangles": 0.6428502186664997, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 1388.0482177734375, + "test_best_rerun_mse": 1886.958740234375, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 1.000034739029854 + "test_mse_by_total_triangles": 1.3594803603994057 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 2033.54296875, + "test_best_rerun_mse": 2206.589111328125, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 10.702857730263158 + "test_mse_by_total_triangles": 11.613626901726974 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 7138.91845703125, + "test_best_rerun_mse": 4386.828125, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.5414424313258438 + "test_mse_by_total_triangles": 0.33271354759196053 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 936.6683349609375, + "test_best_rerun_mse": 9888.755859375, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 0.3156954280286274 + "test_mse_by_total_triangles": 3.3329140072042467 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 1641.92431640625, + "test_best_rerun_mse": 1536.0865478515625, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 1.4178966462920986 + "test_mse_by_total_triangles": 1.3264996095436636 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 152554.234375, + "test_best_rerun_mse": 127173.3203125, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 3.542500333805499 + "test_mse_by_total_triangles": 2.953123730087776 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 10032.2578125, + "test_best_rerun_mse": 8226.447265625, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 0.703425733592764 + "test_mse_by_total_triangles": 0.5768088112203759 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 37066.734375, + "test_best_rerun_mse": 38221.07421875, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 1.899981258649854 + "test_mse_by_total_triangles": 1.9591508646650264 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 2791.48876953125, + "test_best_rerun_mse": 2612.12646484375, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.49626467013888886 + "test_mse_by_total_triangles": 0.46437803819444445 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 708865.9375, + "test_best_rerun_mse": 692534.75, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 8.018573323303508 + "test_mse_by_total_triangles": 7.833837652568352 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 233214.046875, + "test_best_rerun_mse": 217828.421875, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 3.8808854088662574 + "test_mse_by_total_triangles": 3.6248551723994473 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__04__h_mid__d_lo__pl_lo__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_triangle_counting__triangle_counting__04__h_mid__d_lo__pl_lo__s43" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_1.5-2__s44", "train_seed": 44, "homophily": "h_mid", "avg_degree": "d_lo", "power_law": "pl_lo", "run_slug": "h_mid__d_lo__pl_lo", - "test_loss": 5430.181640625, + "test_loss": 5240.5751953125, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 5587.08349609375, + "test_best_rerun_mse": 5442.1123046875, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 0.7464373408274884, + "test_mse_by_total_triangles": 0.7270691121826988, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 1241.96435546875, + "test_best_rerun_mse": 1381.1591796875, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 0.894786999617255 + "test_mse_by_total_triangles": 0.995071455106268 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 1809.4840087890625, + "test_best_rerun_mse": 1802.9298095703125, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 9.523600046258224 + "test_mse_by_total_triangles": 9.489104260896381 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 6959.564453125, + "test_best_rerun_mse": 6106.7958984375, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.5278395489666288 + "test_mse_by_total_triangles": 0.46316237379124003 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 1121.3714599609375, + "test_best_rerun_mse": 1820.029541015625, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 0.3779479137043942 + "test_mse_by_total_triangles": 0.613424179648003 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 1441.8082275390625, + "test_best_rerun_mse": 1440.3955078125, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 1.2450848251632665 + "test_mse_by_total_triangles": 1.24386485994171 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 152518.625, + "test_best_rerun_mse": 147431.1875, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 3.5416734395318596 + "test_mse_by_total_triangles": 3.4235367708526843 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 10085.66015625, + "test_best_rerun_mse": 9446.3076171875, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 0.7071701133256205 + "test_mse_by_total_triangles": 0.662341019295155 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 37927.12890625, + "test_best_rerun_mse": 38090.96875, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 1.9440837001512123 + "test_mse_by_total_triangles": 1.9524818673432776 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 2825.195556640625, + "test_best_rerun_mse": 2799.69140625, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.5022569878472222 + "test_mse_by_total_triangles": 0.4977229166666667 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 710137.125, + "test_best_rerun_mse": 707662.5, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 8.032952784407769 + "test_mse_by_total_triangles": 8.00496023890592 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 233200.609375, + "test_best_rerun_mse": 232313.203125, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 3.8806617971311135 + "test_mse_by_total_triangles": 3.865894582147671 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__04__h_mid__d_lo__pl_lo__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_triangle_counting__triangle_counting__04__h_mid__d_lo__pl_lo__s44" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_4-5__s42", "train_seed": 42, "homophily": "h_mid", "avg_degree": "d_lo", "power_law": "pl_hi", "run_slug": "h_mid__d_lo__pl_hi", - "test_loss": 140.5848846435547, + "test_loss": 131.87379455566406, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 148.2747802734375, + "test_best_rerun_mse": 139.33749389648438, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 0.12804385170417745, + "test_mse_by_total_triangles": 0.12032598782079826, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 157.74961853027344, + "test_best_rerun_mse": 142.73898315429688, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 0.11365246291806444 + "test_mse_by_total_triangles": 0.10283788411692858 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 28.901439666748047, + "test_best_rerun_mse": 40.2872200012207, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 0.1521128403513055 + "test_mse_by_total_triangles": 0.21203800000642475 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 12167.857421875, + "test_best_rerun_mse": 11467.9228515625, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.9228560805365946 + "test_mse_by_total_triangles": 0.8697704096748199 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 231.3026580810547, + "test_best_rerun_mse": 152.51419067382812, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 0.07795842874319335 + "test_mse_by_total_triangles": 0.05140350208083186 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 7747.02490234375, + "test_best_rerun_mse": 7618.44482421875, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 1.0350066669797928 + "test_mse_by_total_triangles": 1.0178282998288244 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 175088.234375, + "test_best_rerun_mse": 172906.84375, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 4.065768028399591 + "test_mse_by_total_triangles": 4.015113406789894 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 15111.373046875, + "test_best_rerun_mse": 14353.3798828125, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 1.0595549745389847 + "test_mse_by_total_triangles": 1.0064072277950147 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 46514.1015625, + "test_best_rerun_mse": 46522.81640625, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 2.384238124070942 + "test_mse_by_total_triangles": 2.38468483296171 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 3693.9208984375, + "test_best_rerun_mse": 3646.09130859375, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.6566970486111111 + "test_mse_by_total_triangles": 0.6481940104166667 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 758421.4375, + "test_best_rerun_mse": 756737.9375, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 8.579136878838954 + "test_mse_by_total_triangles": 8.560093407463548 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 262386.5625, + "test_best_rerun_mse": 260920.59375, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 4.366341545604313 + "test_mse_by_total_triangles": 4.3419465453547 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__05__h_mid__d_lo__pl_hi__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_triangle_counting__triangle_counting__05__h_mid__d_lo__pl_hi__s42" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_4-5__s43", "train_seed": 43, "homophily": "h_mid", "avg_degree": "d_lo", "power_law": "pl_hi", "run_slug": "h_mid__d_lo__pl_hi", - "test_loss": 139.52407836914062, + "test_loss": 139.33627319335938, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 146.0721893310547, + "test_best_rerun_mse": 144.46835327148438, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 0.12614178698709386, + "test_mse_by_total_triangles": 0.1247567817543043, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 170.1674041748047, + "test_best_rerun_mse": 140.94012451171875, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 0.12259899436225122 + "test_mse_by_total_triangles": 0.10154187644936509 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 35.388675689697266, + "test_best_rerun_mse": 46.47947692871094, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 0.18625618784051193 + "test_mse_by_total_triangles": 0.24462882594058388 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 12316.30859375, + "test_best_rerun_mse": 11487.3427734375, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.9341151758627227 + "test_mse_by_total_triangles": 0.8712432896046644 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 228.41705322265625, + "test_best_rerun_mse": 178.04525756835938, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 0.07698586222536442 + "test_mse_by_total_triangles": 0.06000851283058961 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 7770.365234375, + "test_best_rerun_mse": 7571.34912109375, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 1.0381249478122911 + "test_mse_by_total_triangles": 1.01153628872328 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 175380.28125, + "test_best_rerun_mse": 173010.296875, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 4.072549722506038 + "test_mse_by_total_triangles": 4.017515717885009 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 14979.7880859375, + "test_best_rerun_mse": 14542.212890625, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 1.0503287116770088 + "test_mse_by_total_triangles": 1.0196475172223392 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 46556.9296875, + "test_best_rerun_mse": 46360.8046875, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 2.386433424957712 + "test_mse_by_total_triangles": 2.376380372520375 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 3661.864501953125, + "test_best_rerun_mse": 3662.4609375, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.6509981336805556 + "test_mse_by_total_triangles": 0.6511041666666667 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 758924.5, + "test_best_rerun_mse": 756791.8125, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 8.584827437982874 + "test_mse_by_total_triangles": 8.560702832483061 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 262227.5625, + "test_best_rerun_mse": 261337.4375, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 4.363695646747542 + "test_mse_by_total_triangles": 4.348883189389779 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__05__h_mid__d_lo__pl_hi__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_triangle_counting__triangle_counting__05__h_mid__d_lo__pl_hi__s43" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.4-0.6__deg_1-2.5__gamma_4-5__s44", "train_seed": 44, "homophily": "h_mid", "avg_degree": "d_lo", "power_law": "pl_hi", "run_slug": "h_mid__d_lo__pl_hi", - "test_loss": 151.0240478515625, + "test_loss": 143.45443725585938, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 158.71817016601562, + "test_best_rerun_mse": 150.46644592285156, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 0.13706232311400313, + "test_mse_by_total_triangles": 0.12993648179866282, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 164.77513122558594, + "test_best_rerun_mse": 147.7235870361328, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 0.11871407148817431 + "test_mse_by_total_triangles": 0.10642909728828012 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 40.075828552246094, + "test_best_rerun_mse": 35.235042572021484, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 0.2109254134328742 + "test_mse_by_total_triangles": 0.1854475924843236 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 12168.087890625, + "test_best_rerun_mse": 11706.271484375, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.9228735601535836 + "test_mse_by_total_triangles": 0.8878476666192643 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 229.33416748046875, + "test_best_rerun_mse": 196.869384765625, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 0.07729496713194094 + "test_mse_by_total_triangles": 0.06635301138039265 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 7734.0068359375, + "test_best_rerun_mse": 7680.59423828125, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 1.0332674463510354 + "test_mse_by_total_triangles": 1.0261314947603541 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 175163.96875, + "test_best_rerun_mse": 173806.390625, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 4.067526675413339 + "test_mse_by_total_triangles": 4.036002011540963 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 15078.30859375, + "test_best_rerun_mse": 14801.11328125, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 1.0572366143423082 + "test_mse_by_total_triangles": 1.0378006788143317 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 46495.76953125, + "test_best_rerun_mse": 46629.6328125, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 2.3832984535983393 + "test_mse_by_total_triangles": 2.3901600703521453 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 3675.234619140625, + "test_best_rerun_mse": 3709.257080078125, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.6533750434027777 + "test_mse_by_total_triangles": 0.6594234809027778 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 758374.6875, + "test_best_rerun_mse": 758243.3125, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 8.578608050631766 + "test_mse_by_total_triangles": 8.57712195853082 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 261976.296875, + "test_best_rerun_mse": 262258.90625, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 4.359514367313997 + "test_mse_by_total_triangles": 4.364217234120447 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__05__h_mid__d_lo__pl_hi__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_triangle_counting__triangle_counting__05__h_mid__d_lo__pl_hi__s44" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_1.5-2__s42", "train_seed": 42, "homophily": "h_mid", "avg_degree": "d_hi", "power_law": "pl_lo", "run_slug": "h_mid__d_hi__pl_lo", - "test_loss": 114037.5, + "test_loss": 112329.3515625, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 79800.609375, + "test_best_rerun_mse": 78939.2265625, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 1.8530700672255247, + "test_mse_by_total_triangles": 1.8330676797905443, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 68935.5, + "test_best_rerun_mse": 73191.6171875, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 49.66534582132565 + "test_mse_by_total_triangles": 52.73171267110951 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 76215.3828125, + "test_best_rerun_mse": 80139.953125, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 401.13359375 + "test_mse_by_total_triangles": 421.7892269736842 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 35490.3125, + "test_best_rerun_mse": 38812.30859375, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 2.691718809252939 + "test_mse_by_total_triangles": 2.943671489855897 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 67818.5234375, + "test_best_rerun_mse": 71941.7734375, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 22.85760816902595 + "test_mse_by_total_triangles": 24.24731157313785 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 52983.8671875, + "test_best_rerun_mse": 56479.47265625, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 7.078672970941883 + "test_mse_by_total_triangles": 7.545687729625919 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 72574.1484375, + "test_best_rerun_mse": 76361.3828125, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 62.67197619818653 + "test_mse_by_total_triangles": 65.94247220423144 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 42939.69921875, + "test_best_rerun_mse": 45758.6640625, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 3.010776834858365 + "test_mse_by_total_triangles": 3.2084324822956107 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 44125.953125, + "test_best_rerun_mse": 46285.53125, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 2.261825471577221 + "test_mse_by_total_triangles": 2.37252197703624 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 59294.69921875, + "test_best_rerun_mse": 62872.69921875, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 10.54127986111111 + "test_mse_by_total_triangles": 11.17736875 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 479049.65625, + "test_best_rerun_mse": 472597.21875, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 5.41892985814961 + "test_mse_by_total_triangles": 5.345940960714003 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 124289.984375, + "test_best_rerun_mse": 122115.7890625, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 2.068293884063036 + "test_mse_by_total_triangles": 2.0321133753099363 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__06__h_mid__d_hi__pl_lo__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_triangle_counting__triangle_counting__06__h_mid__d_hi__pl_lo__s42" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_1.5-2__s43", "train_seed": 43, "homophily": "h_mid", "avg_degree": "d_hi", "power_law": "pl_lo", "run_slug": "h_mid__d_hi__pl_lo", - "test_loss": 112811.703125, + "test_loss": 113411.0859375, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 79222.84375, + "test_best_rerun_mse": 79427.1328125, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 1.839653625998514, + "test_mse_by_total_triangles": 1.8443974738180382, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 72659.921875, + "test_best_rerun_mse": 67665.515625, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 52.34864688400577 + "test_mse_by_total_triangles": 48.75037148775216 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 80542.1015625, + "test_best_rerun_mse": 73816.90625, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 423.9057976973684 + "test_mse_by_total_triangles": 388.51003289473687 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 37689.73046875, + "test_best_rerun_mse": 35123.19921875, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 2.8585309418847173 + "test_mse_by_total_triangles": 2.6638755569776262 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 71708.2265625, + "test_best_rerun_mse": 65356.1328125, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 24.168596751769464 + "test_mse_by_total_triangles": 22.027682107347488 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 56212.0390625, + "test_best_rerun_mse": 51817.13671875, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 7.509958458583834 + "test_mse_by_total_triangles": 6.922797156813627 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 76693.90625, + "test_best_rerun_mse": 70380.8671875, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 66.22962543177893 + "test_mse_by_total_triangles": 60.77795093911917 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 45696.79296875, + "test_best_rerun_mse": 41154.4765625, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 3.2040943043577337 + "test_mse_by_total_triangles": 2.88560346112046 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 46085.1484375, + "test_best_rerun_mse": 43288.67578125, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 2.362250675970065 + "test_mse_by_total_triangles": 2.218907979970783 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 62861.3203125, + "test_best_rerun_mse": 57791.828125, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 11.175345833333333 + "test_mse_by_total_triangles": 10.274102777777777 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 474396.375, + "test_best_rerun_mse": 480181.34375, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 5.366292716310532 + "test_mse_by_total_triangles": 5.431731318507291 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 122311.703125, + "test_best_rerun_mse": 124360.84375, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 2.03537355640424 + "test_mse_by_total_triangles": 2.0694730459454513 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__06__h_mid__d_hi__pl_lo__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_triangle_counting__triangle_counting__06__h_mid__d_hi__pl_lo__s43" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_1.5-2__s44", "train_seed": 44, "homophily": "h_mid", "avg_degree": "d_hi", "power_law": "pl_lo", "run_slug": "h_mid__d_hi__pl_lo", - "test_loss": 113326.296875, + "test_loss": 112798.171875, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 79597.296875, + "test_best_rerun_mse": 79224.0703125, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 1.8483488964099943, + "test_mse_by_total_triangles": 1.8396821083155304, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 70258.875, + "test_best_rerun_mse": 71314.46875, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 50.61878602305475 + "test_mse_by_total_triangles": 51.379300252161386 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 76866.59375, + "test_best_rerun_mse": 77644.546875, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 404.5610197368421 + "test_mse_by_total_triangles": 408.65550986842106 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 36926.33203125, + "test_best_rerun_mse": 37652.328125, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 2.8006319325938565 + "test_mse_by_total_triangles": 2.8556942074326885 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 68707.9765625, + "test_best_rerun_mse": 69081.34375, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 23.157390145770137 + "test_mse_by_total_triangles": 23.283230114593866 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 54116.28125, + "test_best_rerun_mse": 54969.44140625, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 7.2299640948563795 + "test_mse_by_total_triangles": 7.343946747661991 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 73382.203125, + "test_best_rerun_mse": 74252.359375, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 63.36977817357513 + "test_mse_by_total_triangles": 64.12120844127807 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 43582.765625, + "test_best_rerun_mse": 43952.16796875, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 3.055866331860889 + "test_mse_by_total_triangles": 3.0817674918489693 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 44945.671875, + "test_best_rerun_mse": 45301.55859375, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 2.303842937874827 + "test_mse_by_total_triangles": 2.322085119367984 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 60275.953125, + "test_best_rerun_mse": 61146.87890625, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 10.715725 + "test_mse_by_total_triangles": 10.87055625 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 477642.8125, + "test_best_rerun_mse": 475524.75, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 5.403015876158049 + "test_mse_by_total_triangles": 5.379056706220378 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 123579.5625, + "test_best_rerun_mse": 123280.6953125, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 2.0564718436423544 + "test_mse_by_total_triangles": 2.051498432637745 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__06__h_mid__d_hi__pl_lo__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_triangle_counting__triangle_counting__06__h_mid__d_hi__pl_lo__s44" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_4-5__s42", "train_seed": 42, "homophily": "h_mid", "avg_degree": "d_hi", "power_law": "pl_hi", "run_slug": "h_mid__d_hi__pl_hi", - "test_loss": 8595.5, + "test_loss": 7600.79150390625, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 8167.642578125, + "test_best_rerun_mse": 7335.857421875, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 0.5726856386288739, + "test_mse_by_total_triangles": 0.5143638635447343, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 5147.3974609375, + "test_best_rerun_mse": 6417.23779296875, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 3.708499611626441 + "test_mse_by_total_triangles": 4.623370167844921 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 6535.10107421875, + "test_best_rerun_mse": 6146.603515625, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 34.39526881167763 + "test_mse_by_total_triangles": 32.35054481907895 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 5018.2138671875, + "test_best_rerun_mse": 5137.0224609375, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.3806002174582859 + "test_mse_by_total_triangles": 0.38961110814846417 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 4475.4990234375, + "test_best_rerun_mse": 2307.953125, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 1.5084256904069768 + "test_mse_by_total_triangles": 0.7778743259184361 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 6020.1455078125, + "test_best_rerun_mse": 7319.19677734375, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 0.8042946570223781 + "test_mse_by_total_triangles": 0.9778486008475283 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 5679.240234375, + "test_best_rerun_mse": 6568.56494140625, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 4.904352534002591 + "test_mse_by_total_triangles": 5.672335873407815 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 133994.984375, + "test_best_rerun_mse": 119813.5078125, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 3.1115313109557867 + "test_mse_by_total_triangles": 2.782219668690786 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 31297.95703125, + "test_best_rerun_mse": 28872.88671875, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 1.6042829991926804 + "test_mse_by_total_triangles": 1.4799777906991645 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 4724.83642578125, + "test_best_rerun_mse": 5654.92333984375, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.8399709201388889 + "test_mse_by_total_triangles": 1.005319704861111 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 665115.4375, + "test_best_rerun_mse": 622963.0625, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 7.5236749601257875 + "test_mse_by_total_triangles": 7.046854320554733 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 208794.03125, + "test_best_rerun_mse": 188370.5, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 3.4745150225483834 + "test_mse_by_total_triangles": 3.134649626412394 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__07__h_mid__d_hi__pl_hi__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_triangle_counting__triangle_counting__07__h_mid__d_hi__pl_hi__s42" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_4-5__s43", "train_seed": 43, "homophily": "h_mid", "avg_degree": "d_hi", "power_law": "pl_hi", "run_slug": "h_mid__d_hi__pl_hi", - "test_loss": 8443.4091796875, + "test_loss": 8644.5048828125, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 8042.169921875, + "test_best_rerun_mse": 8210.6728515625, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 0.5638879485258028, + "test_mse_by_total_triangles": 0.5757027662012691, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 5105.32861328125, + "test_best_rerun_mse": 4997.88232421875, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 3.6781906435743874 + "test_mse_by_total_triangles": 3.6007797724918946 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 6542.2548828125, + "test_best_rerun_mse": 6317.9521484375, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 34.432920435855266 + "test_mse_by_total_triangles": 33.25237972861842 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 5097.06494140625, + "test_best_rerun_mse": 5031.31298828125, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.38658057955299585 + "test_mse_by_total_triangles": 0.3815937040789723 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 4294.68798828125, + "test_best_rerun_mse": 4128.7373046875, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 1.4474849977355073 + "test_mse_by_total_triangles": 1.391552849574486 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 6060.990234375, + "test_best_rerun_mse": 5992.5927734375, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 0.8097515343186372 + "test_mse_by_total_triangles": 0.8006135969856379 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 5671.640625, + "test_best_rerun_mse": 5513.1572265625, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 4.897789831606218 + "test_mse_by_total_triangles": 4.760930247463299 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 134758.1875, + "test_best_rerun_mse": 134843.890625, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 3.1292538431172208 + "test_mse_by_total_triangles": 3.131243976987739 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 31428.107421875, + "test_best_rerun_mse": 31513.509765625, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 1.6109542991375776 + "test_mse_by_total_triangles": 1.6153318860846275 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 4679.52490234375, + "test_best_rerun_mse": 4650.177734375, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.8319155381944444 + "test_mse_by_total_triangles": 0.8266982638888889 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 666766.875, + "test_best_rerun_mse": 667359.875, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 7.542355745845729 + "test_mse_by_total_triangles": 7.549063662997862 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 209098.015625, + "test_best_rerun_mse": 210298.40625, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 3.4795735880218994 + "test_mse_by_total_triangles": 3.499549136338675 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__07__h_mid__d_hi__pl_hi__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_triangle_counting__triangle_counting__07__h_mid__d_hi__pl_hi__s43" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.4-0.6__deg_4-5__gamma_4-5__s44", "train_seed": 44, "homophily": "h_mid", "avg_degree": "d_hi", "power_law": "pl_hi", "run_slug": "h_mid__d_hi__pl_hi", - "test_loss": 8486.0244140625, + "test_loss": 8358.1484375, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 8143.67822265625, + "test_best_rerun_mse": 8036.90576171875, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 0.5710053444577373, + "test_mse_by_total_triangles": 0.5635188446023524, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 5631.97265625, + "test_best_rerun_mse": 5527.56103515625, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 4.0576171875 + "test_mse_by_total_triangles": 3.9823926766255404 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 7004.40576171875, + "test_best_rerun_mse": 6695.72265625, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 36.865293482730266 + "test_mse_by_total_triangles": 35.24064555921053 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 5033.87353515625, + "test_best_rerun_mse": 5281.51171875, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.3817879055863671 + "test_mse_by_total_triangles": 0.4005697170079636 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 4876.11328125, + "test_best_rerun_mse": 4290.89404296875, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 1.6434490331142568 + "test_mse_by_total_triangles": 1.4462062834407652 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 6416.95166015625, + "test_best_rerun_mse": 6490.30126953125, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 0.8573081710295591 + "test_mse_by_total_triangles": 0.8671077180402471 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 6146.17236328125, + "test_best_rerun_mse": 5944.7890625, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 5.307575443248057 + "test_mse_by_total_triangles": 5.133669311312608 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 133150.296875, + "test_best_rerun_mse": 133170.0, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 3.0919166095810886 + "test_mse_by_total_triangles": 3.0923741408136727 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 31501.736328125, + "test_best_rerun_mse": 31789.173828125, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 1.6147283985916756 + "test_mse_by_total_triangles": 1.6294619830911374 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 5078.74072265625, + "test_best_rerun_mse": 4989.509765625, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.9028872395833333 + "test_mse_by_total_triangles": 0.8870239583333334 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 662870.625, + "test_best_rerun_mse": 663439.75, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 7.498282015316223 + "test_mse_by_total_triangles": 7.504719862448107 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 206839.375, + "test_best_rerun_mse": 207949.40625, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 3.441987835521608 + "test_mse_by_total_triangles": 3.460459724926364 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__07__h_mid__d_hi__pl_hi__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_triangle_counting__triangle_counting__07__h_mid__d_hi__pl_hi__s44" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_1.5-2__s42", "train_seed": 42, "homophily": "h_hi", "avg_degree": "d_lo", "power_law": "pl_lo", "run_slug": "h_hi__d_lo__pl_lo", - "test_loss": 26912.43359375, + "test_loss": 27631.06640625, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 27320.7109375, + "test_best_rerun_mse": 28041.640625, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 1.4004157536265314, + "test_mse_by_total_triangles": 1.4373694512788968, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 13399.8154296875, + "test_best_rerun_mse": 12090.3095703125, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 9.654045698622118 + "test_mse_by_total_triangles": 8.710597673135807 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 16025.248046875, + "test_best_rerun_mse": 14401.25, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 84.34341077302632 + "test_mse_by_total_triangles": 75.79605263157895 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 6593.76513671875, + "test_best_rerun_mse": 5846.75048828125, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.5000959527280053 + "test_mse_by_total_triangles": 0.443439551633011 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 16185.9501953125, + "test_best_rerun_mse": 13098.5966796875, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 5.455325310182845 + "test_mse_by_total_triangles": 4.414761267168015 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 9942.73046875, + "test_best_rerun_mse": 9427.224609375, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 1.3283541040414162 + "test_mse_by_total_triangles": 1.259482245741483 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 14064.2119140625, + "test_best_rerun_mse": 12788.3828125, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 12.145260720261227 + "test_mse_by_total_triangles": 11.043508473661486 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 111503.5625, + "test_best_rerun_mse": 115433.5859375, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 2.589252333735835 + "test_mse_by_total_triangles": 2.680512398697288 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 10522.2880859375, + "test_best_rerun_mse": 9602.6943359375, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 0.7377848889312508 + "test_mse_by_total_triangles": 0.673306291960279 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 10269.8876953125, + "test_best_rerun_mse": 9366.4970703125, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 1.8257578125 + "test_mse_by_total_triangles": 1.6651550347222221 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 609100.375, + "test_best_rerun_mse": 619468.25, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 6.8900419103424095 + "test_mse_by_total_triangles": 7.007321584109136 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 178438.59375, + "test_best_rerun_mse": 183960.1875, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 2.9693740327492386 + "test_mse_by_total_triangles": 3.061258174829015 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__08__h_hi__d_lo__pl_lo__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_triangle_counting__triangle_counting__08__h_hi__d_lo__pl_lo__s42" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_1.5-2__s43", "train_seed": 43, "homophily": "h_hi", "avg_degree": "d_lo", "power_law": "pl_lo", "run_slug": "h_hi__d_lo__pl_lo", - "test_loss": 27183.927734375, + "test_loss": 26880.51171875, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 27604.82421875, + "test_best_rerun_mse": 27414.375, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 1.4149789440130196, + "test_mse_by_total_triangles": 1.405216823004767, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 12871.2861328125, + "test_best_rerun_mse": 14308.962890625, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 9.273260902602665 + "test_mse_by_total_triangles": 10.30905107393732 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 15470.7685546875, + "test_best_rerun_mse": 16664.80859375, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 81.42509765625 + "test_mse_by_total_triangles": 87.70951891447369 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 6133.82421875, + "test_best_rerun_mse": 7211.45068359375, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.4652123032802427 + "test_mse_by_total_triangles": 0.5469435482437429 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 14489.2333984375, + "test_best_rerun_mse": 17475.765625, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 4.8834625542425005 + "test_mse_by_total_triangles": 5.89004571115605 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 9831.2646484375, + "test_best_rerun_mse": 10412.0283203125, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 1.3134622108800935 + "test_mse_by_total_triangles": 1.3910525478039413 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 13759.1298828125, + "test_best_rerun_mse": 14742.48828125, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 11.88180473472582 + "test_mse_by_total_triangles": 12.730991607297064 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 113593.4296875, + "test_best_rerun_mse": 109808.765625, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 2.6377816665312093 + "test_mse_by_total_triangles": 2.549897028260264 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 9971.8212890625, + "test_best_rerun_mse": 10939.78515625, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 0.6991881425510097 + "test_mse_by_total_triangles": 0.7670582776784463 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 10068.2685546875, + "test_best_rerun_mse": 10624.6435546875, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 1.7899144097222222 + "test_mse_by_total_triangles": 1.8888255208333333 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 613865.5, + "test_best_rerun_mse": 607455.4375, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 6.943944210038121 + "test_mse_by_total_triangles": 6.871434651538975 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 180074.6875, + "test_best_rerun_mse": 176417.859375, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 2.996600061571231 + "test_mse_by_total_triangles": 2.9357472480155757 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__08__h_hi__d_lo__pl_lo__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_triangle_counting__triangle_counting__08__h_hi__d_lo__pl_lo__s43" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_1.5-2__s44", "train_seed": 44, "homophily": "h_hi", "avg_degree": "d_lo", "power_law": "pl_lo", "run_slug": "h_hi__d_lo__pl_lo", - "test_loss": 27479.0625, + "test_loss": 27448.056640625, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 27932.48828125, + "test_best_rerun_mse": 27911.796875, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 1.4317744774847507, + "test_mse_by_total_triangles": 1.430713869239838, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 13285.484375, + "test_best_rerun_mse": 12597.525390625, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 9.571674621757925 + "test_mse_by_total_triangles": 9.076026938490633 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 15702.0595703125, + "test_best_rerun_mse": 14665.5302734375, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 82.64241879111842 + "test_mse_by_total_triangles": 77.18700143914474 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 6248.439453125, + "test_best_rerun_mse": 6016.37548828125, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.4739051538206295 + "test_mse_by_total_triangles": 0.456304549736917 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 14116.6982421875, + "test_best_rerun_mse": 13118.1416015625, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 4.757903013881867 + "test_mse_by_total_triangles": 4.421348702919616 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 10264.9541015625, + "test_best_rerun_mse": 9753.3037109375, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 1.3714033535821644 + "test_mse_by_total_triangles": 1.3030465879676019 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 14139.0400390625, + "test_best_rerun_mse": 13235.19140625, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 12.209879135632557 + "test_mse_by_total_triangles": 11.429353545984457 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 113889.1171875, + "test_best_rerun_mse": 114429.4140625, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 2.6446479005085455 + "test_mse_by_total_triangles": 2.6571942704463125 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 10075.3671875, + "test_best_rerun_mse": 9716.5771484375, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 0.7064484074814191 + "test_mse_by_total_triangles": 0.681291344021701 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 10568.3359375, + "test_best_rerun_mse": 9799.826171875, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 1.8788152777777778 + "test_mse_by_total_triangles": 1.7421913194444445 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 613494.0, + "test_best_rerun_mse": 617749.9375, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 6.939741863963893 + "test_mse_by_total_triangles": 6.9878843195366676 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 180730.046875, + "test_best_rerun_mse": 183773.703125, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 3.0075058139051136 + "test_mse_by_total_triangles": 3.05815491196978 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__08__h_hi__d_lo__pl_lo__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_triangle_counting__triangle_counting__08__h_hi__d_lo__pl_lo__s44" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_4-5__s42", "train_seed": 42, "homophily": "h_hi", "avg_degree": "d_lo", "power_law": "pl_hi", "run_slug": "h_hi__d_lo__pl_hi", - "test_loss": 2638.443115234375, + "test_loss": 2523.91259765625, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 2797.404052734375, + "test_best_rerun_mse": 2642.5126953125, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.49731627604166667, + "test_mse_by_total_triangles": 0.46978003472222224, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 456.7854309082031, + "test_best_rerun_mse": 862.0759887695312, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 0.3290961317782443 + "test_mse_by_total_triangles": 0.6210922109290571 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 750.0567626953125, + "test_best_rerun_mse": 1253.0223388671875, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 3.947667172080592 + "test_mse_by_total_triangles": 6.59485441509046 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 8805.228515625, + "test_best_rerun_mse": 5847.9228515625, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.6678216545790672 + "test_mse_by_total_triangles": 0.4435284680745165 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 311.92974853515625, + "test_best_rerun_mse": 2602.08203125, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 0.1051330463549566 + "test_mse_by_total_triangles": 0.8770077624705089 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 6033.4267578125, + "test_best_rerun_mse": 5652.150390625, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 0.8060690391199065 + "test_mse_by_total_triangles": 0.7551303127087509 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 582.322265625, + "test_best_rerun_mse": 858.4324340820312, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 0.5028689685880829 + "test_mse_by_total_triangles": 0.7413060743368145 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 161415.3125, + "test_best_rerun_mse": 147585.4375, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 3.7482656627345348 + "test_mse_by_total_triangles": 3.4271186489875536 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 11794.5849609375, + "test_best_rerun_mse": 8963.4814453125, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 0.8269937569020824 + "test_mse_by_total_triangles": 0.6284869895745688 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 40758.4453125, + "test_best_rerun_mse": 40217.12109375, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 2.089212430801169 + "test_mse_by_total_triangles": 2.0614650209518683 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 729070.9375, + "test_best_rerun_mse": 711705.4375, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 8.247128915308304 + "test_mse_by_total_triangles": 8.050693273983915 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 244943.390625, + "test_best_rerun_mse": 231420.0, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 4.0760719322550045 + "test_mse_by_total_triangles": 3.8510309021017424 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__09__h_hi__d_lo__pl_hi__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_triangle_counting__triangle_counting__09__h_hi__d_lo__pl_hi__s42" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_4-5__s43", "train_seed": 43, "homophily": "h_hi", "avg_degree": "d_lo", "power_law": "pl_hi", "run_slug": "h_hi__d_lo__pl_hi", - "test_loss": 2535.148193359375, + "test_loss": 2590.84619140625, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 2684.48974609375, + "test_best_rerun_mse": 2738.189208984375, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.4772426215277778, + "test_mse_by_total_triangles": 0.4867891927083333, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 563.5260620117188, + "test_best_rerun_mse": 554.4853515625, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 0.4059986037548406 + "test_mse_by_total_triangles": 0.39948512360410665 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 919.5280151367188, + "test_best_rerun_mse": 880.0433349609375, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 4.83962113229852 + "test_mse_by_total_triangles": 4.631807026110198 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 8234.5439453125, + "test_best_rerun_mse": 7823.33837890625, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.6245387899364808 + "test_mse_by_total_triangles": 0.5933514128863292 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 562.0038452148438, + "test_best_rerun_mse": 641.6730346679688, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 0.18941821544147075 + "test_mse_by_total_triangles": 0.21626998135084893 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 5981.49169921875, + "test_best_rerun_mse": 5860.53662109375, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 0.7991304875375751 + "test_mse_by_total_triangles": 0.7829708244614229 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 677.7760620117188, + "test_best_rerun_mse": 651.82861328125, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 0.5852988445697053 + "test_mse_by_total_triangles": 0.5628917213136874 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 159199.953125, + "test_best_rerun_mse": 157703.140625, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 3.696822244217908 + "test_mse_by_total_triangles": 3.6620643838240756 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 10987.5, + "test_best_rerun_mse": 10877.1396484375, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 0.7704038704249053 + "test_mse_by_total_triangles": 0.7626658006196536 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 40434.38671875, + "test_best_rerun_mse": 40549.3828125, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 2.0726017078656005 + "test_mse_by_total_triangles": 2.078496222897124 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 725228.9375, + "test_best_rerun_mse": 725308.5625, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 8.2036688517358 + "test_mse_by_total_triangles": 8.204569556463015 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 241114.109375, + "test_best_rerun_mse": 241854.109375, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 4.012349348093788 + "test_mse_by_total_triangles": 4.024663594345431 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__09__h_hi__d_lo__pl_hi__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_triangle_counting__triangle_counting__09__h_hi__d_lo__pl_hi__s43" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.9-1__deg_1-2.5__gamma_4-5__s44", "train_seed": 44, "homophily": "h_hi", "avg_degree": "d_lo", "power_law": "pl_hi", "run_slug": "h_hi__d_lo__pl_hi", - "test_loss": 2634.956298828125, + "test_loss": 2676.309326171875, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 2796.9013671875, + "test_best_rerun_mse": 2828.345947265625, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 0.4972269097222222, + "test_mse_by_total_triangles": 0.5028170572916667, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 542.5733032226562, + "test_best_rerun_mse": 686.5233154296875, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 0.3909029562122884 + "test_mse_by_total_triangles": 0.4946133396467489 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 865.4780883789062, + "test_best_rerun_mse": 969.9566650390625, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 4.555147833573191 + "test_mse_by_total_triangles": 5.105035079152961 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 8386.9013671875, + "test_best_rerun_mse": 7182.55908203125, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 0.6360941499573379 + "test_mse_by_total_triangles": 0.5447523004953546 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 465.0816650390625, + "test_best_rerun_mse": 1123.830810546875, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 0.1567514880482179 + "test_mse_by_total_triangles": 0.3787768151489299 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 6042.7861328125, + "test_best_rerun_mse": 5947.642578125, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 0.8073194566215764 + "test_mse_by_total_triangles": 0.7946082268704074 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 675.635498046875, + "test_best_rerun_mse": 726.1038818359375, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 0.583450343736507 + "test_mse_by_total_triangles": 0.6270327131571135 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 159965.765625, + "test_best_rerun_mse": 154475.375, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 3.7146053693340146 + "test_mse_by_total_triangles": 3.5871116245587964 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 11366.46484375, + "test_best_rerun_mse": 10376.126953125, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 0.7969755184230823 + "test_mse_by_total_triangles": 0.7275365974705511 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 40628.6328125, + "test_best_rerun_mse": 40826.7890625, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 2.082558450586909 + "test_mse_by_total_triangles": 2.092715621636168 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 726139.1875, + "test_best_rerun_mse": 722347.4375, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 8.213965448005158 + "test_mse_by_total_triangles": 8.171073804056423 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 242234.234375, + "test_best_rerun_mse": 239842.140625, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 4.0309892063135475 + "test_mse_by_total_triangles": 3.991182677266903 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__09__h_hi__d_lo__pl_hi__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_triangle_counting__triangle_counting__09__h_hi__d_lo__pl_hi__s44" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_1.5-2__s42", "train_seed": 42, "homophily": "h_hi", "avg_degree": "d_hi", "power_law": "pl_lo", "run_slug": "h_hi__d_hi__pl_lo", - "test_loss": 504694.40625, + "test_loss": 506702.65625, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 324194.34375, + "test_best_rerun_mse": 325270.21875, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 3.667232376163705, + "test_mse_by_total_triangles": 3.6794024948248363, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 372864.625, + "test_best_rerun_mse": 365471.6875, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 268.6344560518732 + "test_mse_by_total_triangles": 263.30813220461096 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 400687.8125, + "test_best_rerun_mse": 403677.90625, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 2108.8832236842104 + "test_mse_by_total_triangles": 2124.620559210526 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 278948.46875, + "test_best_rerun_mse": 286011.53125, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 21.15650123246113 + "test_mse_by_total_triangles": 21.69219046264695 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 395364.125, + "test_best_rerun_mse": 431218.21875, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 133.2538338388945 + "test_mse_by_total_triangles": 145.33812563195147 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 324730.53125, + "test_best_rerun_mse": 313512.46875, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 43.384172511690046 + "test_mse_by_total_triangles": 41.885433366733466 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 387715.84375, + "test_best_rerun_mse": 382661.96875, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 334.81506368739207 + "test_mse_by_total_triangles": 330.45075021588946 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 165992.984375, + "test_best_rerun_mse": 167717.125, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 3.854564935328813 + "test_mse_by_total_triangles": 3.8946016394203977 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 311005.09375, + "test_best_rerun_mse": 331581.5, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 21.80655544453793 + "test_mse_by_total_triangles": 23.249298836067872 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 254248.140625, + "test_best_rerun_mse": 245907.59375, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 13.032351254549182 + "test_mse_by_total_triangles": 12.604828220308576 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 347621.375, + "test_best_rerun_mse": 338567.375, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 61.79935555555556 + "test_mse_by_total_triangles": 60.18975555555556 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 162152.5, + "test_best_rerun_mse": 166656.625, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 2.698359209891335 + "test_mse_by_total_triangles": 2.773311783402393 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__10__h_hi__d_hi__pl_lo__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_triangle_counting__triangle_counting__10__h_hi__d_hi__pl_lo__s42" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_1.5-2__s43", "train_seed": 43, "homophily": "h_hi", "avg_degree": "d_hi", "power_law": "pl_lo", "run_slug": "h_hi__d_hi__pl_lo", - "test_loss": 504446.0625, + "test_loss": 502976.25, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 325006.71875, + "test_best_rerun_mse": 323855.53125, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 3.676421826747961, + "test_mse_by_total_triangles": 3.6633997856407587, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 372370.15625, + "test_best_rerun_mse": 385846.59375, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 268.27821055475505 + "test_mse_by_total_triangles": 277.9874594740634 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 420939.65625, + "test_best_rerun_mse": 411094.25, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 2215.471875 + "test_mse_by_total_triangles": 2163.653947368421 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 319796.6875, + "test_best_rerun_mse": 281455.28125, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 24.254583807356845 + "test_mse_by_total_triangles": 21.346627322715207 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 497963.03125, + "test_best_rerun_mse": 390664.375, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 167.83384942703066 + "test_mse_by_total_triangles": 131.6698264239973 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 324205.96875, + "test_best_rerun_mse": 337288.28125, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 43.31409068136273 + "test_mse_by_total_triangles": 45.06189462257849 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 396425.96875, + "test_best_rerun_mse": 400468.34375, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 342.3367605785838 + "test_mse_by_total_triangles": 345.82758527633854 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 188871.515625, + "test_best_rerun_mse": 168898.21875, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 4.385833076932008 + "test_mse_by_total_triangles": 3.922028115130968 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 386584.84375, + "test_best_rerun_mse": 310284.53125, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 27.105934914458 + "test_mse_by_total_triangles": 21.756032200953584 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 258376.34375, + "test_best_rerun_mse": 266071.96875, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 13.243956315034087 + "test_mse_by_total_triangles": 13.638421689989237 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 355812.21875, + "test_best_rerun_mse": 362176.8125, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 63.25550555555556 + "test_mse_by_total_triangles": 64.3869888888889 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 194173.546875, + "test_best_rerun_mse": 163764.640625, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 3.2312173942888522 + "test_mse_by_total_triangles": 2.7251866377947516 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__10__h_hi__d_hi__pl_lo__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_triangle_counting__triangle_counting__10__h_hi__d_hi__pl_lo__s43" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_1.5-2__s44", "train_seed": 44, "homophily": "h_hi", "avg_degree": "d_hi", "power_law": "pl_lo", "run_slug": "h_hi__d_hi__pl_lo", - "test_loss": 503276.6875, + "test_loss": 504053.34375, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 323960.75, + "test_best_rerun_mse": 323735.75, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 3.6645900026017215, + "test_mse_by_total_triangles": 3.662044840107236, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 389638.25, + "test_best_rerun_mse": 366068.8125, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 280.7192002881844 + "test_mse_by_total_triangles": 263.73833753602304 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 418526.71875, + "test_best_rerun_mse": 394403.25, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 2202.7722039473683 + "test_mse_by_total_triangles": 2075.8065789473685 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 291043.6875, + "test_best_rerun_mse": 279357.1875, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 22.073848122866895 + "test_mse_by_total_triangles": 21.1875 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 411130.34375, + "test_best_rerun_mse": 400534.8125, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 138.5676925345467 + "test_mse_by_total_triangles": 134.99656639703403 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 340389.4375, + "test_best_rerun_mse": 317624.40625, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 45.47621075484302 + "test_mse_by_total_triangles": 42.434790414161654 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 405894.15625, + "test_best_rerun_mse": 381594.9375, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 350.51308829879105 + "test_mse_by_total_triangles": 329.5293069948186 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 173506.9375, + "test_best_rerun_mse": 165287.8125, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 4.029048335036225 + "test_mse_by_total_triangles": 3.8381899614527217 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 325409.6875, + "test_best_rerun_mse": 313355.6875, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 22.816553603982612 + "test_mse_by_total_triangles": 21.971370600196327 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 267784.8125, + "test_best_rerun_mse": 248050.796875, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 13.726219309036855 + "test_mse_by_total_triangles": 12.714685369573017 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 364238.8125, + "test_best_rerun_mse": 340291.71875, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 64.75356666666667 + "test_mse_by_total_triangles": 60.49630555555556 }, "h_hi__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 167508.921875, + "test_best_rerun_mse": 160278.734375, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 2.787494747724361 + "test_mse_by_total_triangles": 2.667178113507397 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__10__h_hi__d_hi__pl_lo__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_triangle_counting__triangle_counting__10__h_hi__d_hi__pl_lo__s44" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_4-5__s42", "train_seed": 42, "homophily": "h_hi", "avg_degree": "d_hi", "power_law": "pl_hi", "run_slug": "h_hi__d_hi__pl_hi", - "test_loss": 112238.046875, + "test_loss": 87454.625, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 108224.453125, + "test_best_rerun_mse": 84453.0390625, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 1.8009494138252375, + "test_mse_by_total_triangles": 1.4053723239395604, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 122463.5703125, + "test_best_rerun_mse": 75107.2265625, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 88.23023797730548 + "test_mse_by_total_triangles": 54.11183469920749 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 132866.03125, + "test_best_rerun_mse": 66173.8984375, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 699.2949013157895 + "test_mse_by_total_triangles": 348.2836759868421 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 77175.2265625, + "test_best_rerun_mse": 38652.94140625, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 5.853259504171407 + "test_mse_by_total_triangles": 2.931584482840349 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 128721.5625, + "test_best_rerun_mse": 29759.92578125, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 43.3844160768453 + "test_mse_by_total_triangles": 10.03030865562858 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 98648.359375, + "test_best_rerun_mse": 78118.1171875, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 13.179473530394121 + "test_mse_by_total_triangles": 10.43662220273881 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 126825.3359375, + "test_best_rerun_mse": 71284.9765625, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 109.52101549006909 + "test_mse_by_total_triangles": 61.5587016947323 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 80075.2265625, + "test_best_rerun_mse": 72043.921875, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 1.859447022164685 + "test_mse_by_total_triangles": 1.6729500714053502 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 87367.6953125, + "test_best_rerun_mse": 36780.6484375, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 6.125907678621512 + "test_mse_by_total_triangles": 2.5789264084630488 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 74841.546875, + "test_best_rerun_mse": 67793.890625, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 3.8362574645035625 + "test_mse_by_total_triangles": 3.4750059267517557 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 107863.3828125, + "test_best_rerun_mse": 70911.265625, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 19.1757125 + "test_mse_by_total_triangles": 12.606447222222222 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 417606.28125, + "test_best_rerun_mse": 385699.25, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 4.723892642218024 + "test_mse_by_total_triangles": 4.3629656233385745 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__11__h_hi__d_hi__pl_hi__s42" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s42_triangle_counting__triangle_counting__11__h_hi__d_hi__pl_hi__s42" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_4-5__s43", "train_seed": 43, "homophily": "h_hi", "avg_degree": "d_hi", "power_law": "pl_hi", "run_slug": "h_hi__d_hi__pl_hi", - "test_loss": 111056.9140625, + "test_loss": 96312.2109375, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 107066.0703125, + "test_best_rerun_mse": 84981.21875, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 1.781672912194432, + "test_mse_by_total_triangles": 1.4141616952057643, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 126469.625, + "test_best_rerun_mse": 84683.34375, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 91.11644452449568 + "test_mse_by_total_triangles": 61.011054574927954 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 136628.953125, + "test_best_rerun_mse": 70101.375, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 719.0997532894737 + "test_mse_by_total_triangles": 368.9546052631579 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 77999.96875, + "test_best_rerun_mse": 32338.546875, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 5.91581105422829 + "test_mse_by_total_triangles": 2.45267704778157 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 127268.8125, + "test_best_rerun_mse": 24611.18359375, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 42.89478008088979 + "test_mse_by_total_triangles": 8.294972562773845 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 102083.4765625, + "test_best_rerun_mse": 94510.46875, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 13.63840702237809 + "test_mse_by_total_triangles": 12.626649131596526 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 130883.796875, + "test_best_rerun_mse": 77806.578125, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 113.02573132556131 + "test_mse_by_total_triangles": 67.19048197322971 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 80019.90625, + "test_best_rerun_mse": 85313.9609375, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 1.8581624152424299 + "test_mse_by_total_triangles": 1.9810969937186513 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 86924.9609375, + "test_best_rerun_mse": 27104.462890625, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 6.094864741095218 + "test_mse_by_total_triangles": 1.9004671778589959 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 76914.8359375, + "test_best_rerun_mse": 92073.78125, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 3.942530931236865 + "test_mse_by_total_triangles": 4.719554116049003 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 111583.9765625, + "test_best_rerun_mse": 74638.0859375, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 19.837151388888888 + "test_mse_by_total_triangles": 13.268993055555555 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 416493.125, + "test_best_rerun_mse": 390714.125, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 4.711300804271349 + "test_mse_by_total_triangles": 4.419693053403165 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__11__h_hi__d_hi__pl_hi__s43" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s43_triangle_counting__triangle_counting__11__h_hi__d_hi__pl_hi__s43" }, { "experiment": "triangle_counting", - "wandb_project": "challenge_triangle_counting", + "wandb_project": "shard_tri", "wandb_run_name": "gread_hom_0.9-1__deg_4-5__gamma_4-5__s44", "train_seed": 44, "homophily": "h_hi", "avg_degree": "d_hi", "power_law": "pl_hi", "run_slug": "h_hi__d_hi__pl_hi", - "test_loss": 110714.2109375, + "test_loss": 77906.46875, "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 106493.765625, + "test_best_rerun_mse": 79139.1953125, "test_triangles_total_structural": 60093.0, - "test_mse_by_total_triangles": 1.7721492623932904, + "test_mse_by_total_triangles": 1.3169453232905663, "ood_test": { "h_lo__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 129045.0390625, + "test_best_rerun_mse": 100867.1953125, "test_triangles_total_structural": 1388.0, - "test_mse_by_total_triangles": 92.97193016030259 + "test_mse_by_total_triangles": 72.67088999459654 }, "h_lo__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 137913.546875, + "test_best_rerun_mse": 74055.8046875, "test_triangles_total_structural": 190.0, - "test_mse_by_total_triangles": 725.8607730263158 + "test_mse_by_total_triangles": 389.76739309210524 }, "h_lo__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 80733.359375, + "test_best_rerun_mse": 60992.74609375, "test_triangles_total_structural": 13185.0, - "test_mse_by_total_triangles": 6.123121681835419 + "test_mse_by_total_triangles": 4.6259193093477435 }, "h_lo__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 128834.28125, + "test_best_rerun_mse": 32375.06640625, "test_triangles_total_structural": 2967.0, - "test_mse_by_total_triangles": 43.42240689248399 + "test_mse_by_total_triangles": 10.911717696747557 }, "h_mid__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 104496.15625, + "test_best_rerun_mse": 120791.3828125, "test_triangles_total_structural": 7485.0, - "test_mse_by_total_triangles": 13.960742317969272 + "test_mse_by_total_triangles": 16.13779329492318 }, "h_mid__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 132674.8125, + "test_best_rerun_mse": 98727.5078125, "test_triangles_total_structural": 1158.0, - "test_mse_by_total_triangles": 114.57237694300518 + "test_mse_by_total_triangles": 85.25691520941278 }, "h_mid__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 81772.6328125, + "test_best_rerun_mse": 77641.0859375, "test_triangles_total_structural": 43064.0, - "test_mse_by_total_triangles": 1.8988629205949286 + "test_mse_by_total_triangles": 1.8029232290892625 }, "h_mid__d_hi__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 88149.8359375, + "test_best_rerun_mse": 34599.98828125, "test_triangles_total_structural": 14262.0, - "test_mse_by_total_triangles": 6.180748558231665 + "test_mse_by_total_triangles": 2.4260263834840834 }, "h_hi__d_lo__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 79145.4453125, + "test_best_rerun_mse": 94788.75, "test_triangles_total_structural": 19509.0, - "test_mse_by_total_triangles": 4.056868384463581 + "test_mse_by_total_triangles": 4.858719052744887 }, "h_hi__d_lo__pl_hi": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 113563.1484375, + "test_best_rerun_mse": 101391.234375, "test_triangles_total_structural": 5625.0, - "test_mse_by_total_triangles": 20.189004166666667 + "test_mse_by_total_triangles": 18.025108333333332 }, "h_hi__d_hi__pl_lo": { "test_best_rerun_accuracy": null, - "test_best_rerun_mse": 416349.90625, + "test_best_rerun_mse": 332930.8125, "test_triangles_total_structural": 88403.0, - "test_mse_by_total_triangles": 4.70968073764465 + "test_mse_by_total_triangles": 3.766057854371458 } }, - "output_dir": "/home/ubuntu/repos/TopoBench/logs/train/runs/notebook_gu_grid_2026-07-25_13-46-02__triangle_counting__11__h_hi__d_hi__pl_hi__s44" + "output_dir": "/private/tmp/claude-501/-Users-hina-Documents-topoLab-tdl-challenge-2026/34bdf9a1-f9d0-47d8-984b-c8715484fb64/scratchpad/exp-gread/logs/train/runs/notebook_gu_grid_2026-08-02_gread-vc_s44_triangle_counting__triangle_counting__11__h_hi__d_hi__pl_hi__s44" } ] } diff --git a/configs/model/graph/gread.yaml b/configs/model/graph/gread.yaml index a059f4b8d..9e0100aae 100644 --- a/configs/model/graph/gread.yaml +++ b/configs/model/graph/gread.yaml @@ -17,7 +17,10 @@ backbone: reaction_term: bspm # GREAD-BS. Options: bspm, fisher, allen-cahn, zeldovich, st, fb, fb3, none time: 3.0 step_size: 1.0 - beta_diag: false + # Per-channel reaction gate (the "(VC)" variants of the paper); every + # tuned dataset configuration in the reference implementation's + # src/gread_params.py sets beta_diag: true. + beta_diag: true add_source: false data_norm: rw self_loop_weight: 1.0