tmp

app/public/finetasks

@@ -0,0 +1 @@
+../src/content/assets/finetasks

app/src/content/assets/finetasks/code.js

@@ -1,5 +1,7 @@
+import Plotly from 'plotly.js-basic-dist-min';
 import Papa from 'papaparse';
-import { DataTable } from 'simple-datatables';
+import _ from 'lodash';
+import { getColor } from './colors.mjs';
 
 const languageMap = {
   'Arabic': 'ar',
@@ -10,236 +12,608 @@ const languageMap = {
   'Thai': 'th',
   'Chinese': 'zh',
   'French': 'fr',
-  'Hindi': 'hi',
+  'Hindi': 'hi'
 };
 
-const metricTypes = [
-  { value: 'max_score', label: 'Max Score' },
-  { value: 'avg_snr', label: 'Low Noise' },
-  { value: 'avg_spearman', label: 'Monotonicity' },
-  { value: 'max_n_std', label: 'Non-Randomness' },
-  { value: 'avg_kendall_tau_a', label: 'Ordering Consistency' }
-];
-
-const tableTypes = [
-  { value: 'gen', label: 'Generative' },
-  { value: 'mc', label: 'Multichoice' }
-];
-
-const taskFolders = [
-  { value: 'selected', label: 'FineTasks' },
-  { value: 'non_selected', label: 'Non-Selected' }
-];
-
-function createDropdown(options, onChange) {
+const runNameMap = {
+  "orion": "Dataset-A",
+  "helios": "Dataset-B",
+  "lynx": "Dataset-C",
+  "aquila": "Dataset-D",
+  "commoncrawl": "CommonCrawl",
+  "baseline": "Baseline"
+};
+
+const taskLists = {
+  ar: ['acva_ara:_average', 'alfgahafa_mlqa_ara_cf', 'alghafa_arc_ara_cf:easy', 'alghafa_facts_ara_cf', 'alghafa_meta_dialects_ara_cf', 'alghafa_mmlu_ara_cf:_average', 'alghafa_openbookqa_ara_cf', 'alghafa_piqa_ara_cf', 'alghafa_race_ara_cf', 'alghafa_rating_sentiment_ara_cf', 'alghafa_rating_sentiment_no_neutral_ara_cf', 'alghafa_sciqa_ara_cf', 'alghafa_sentiment_ara_cf', 'arcd_ara', 'belebele_arb_Arab_cf', 'boolq_ara', 'exams_ara_cf:_average', 'mkqa_ara:_average', 'mlmm_arc_ara_cf:challenge', 'mlmm_hellaswag_ara_cf', 'mlmm_mmlu_ara_cf:_average', 'mlmm_truthfulqa_ara_cf:mc1', 'mlmm_truthfulqa_ara_cf:mc2', 'mlqa_ara', 'mmlu_ara_cf:_average', 'soqal_ara_cf', 'toxigen_ara_cf', 'tydiqa_ara', 'xcodah_ara_cf', 'xcopa_ara_cf', 'xcsqa_ara_cf', 'xnli2.0_ara_cf', 'xnli_ara_cf', 'xquad_ara', 'xstory_cloze_ara_cf'],
+  fr: ['belebele_fra_Latn_cf', 'community_boolq_fra_cf', 'exams_fra_cf:_average', 'fquadv2_fra', 'frenchbench_arc_fra_cf:challenge', 'frenchbench_hellaswag_fra_cf', 'meta_mmlu_fra_cf:_average', 'mintaka_fra', 'mkqa_fra:_average', 'mlmm_arc_fra_cf:challenge', 'mlmm_hellaswag_fra_cf', 'mlmm_mmlu_fra_cf:_average', 'mlmm_truthfulqa_fra_cf:mc1', 'mlmm_truthfulqa_fra_cf:mc2', 'pawsx_fra_cf', 'xcodah_fra_cf', 'xcsqa_fra_cf', 'xnli2.0_fra_cf', 'xwinograd_fra_cf'],
+  hi: ['belebele_hin_Deva_cf', 'community_arc_hin_cf:challenge', 'community_arc_hin_cf:easy', 'community_boolq_hin', 'community_hellaswag_hin_cf', 'indicnxnli_hin_cf', 'indicqa_hin', 'indicxcopa_hin_cf', 'meta_mmlu_hin_cf:_average', 'mintaka_hin', 'mlmm_arc_hin_cf:challenge', 'mlmm_hellaswag_hin_cf', 'mlmm_mmlu_hin_cf:_average', 'mlmm_truthfulqa_hin_cf:mc1', 'mlmm_truthfulqa_hin_cf:mc2', 'mlqa_hin', 'xcodah_hin_cf', 'xcsqa_hin_cf', 'xnli2.0_hin_cf', 'xnli_hin_cf', 'xquad_hin', 'xstory_cloze_hin_cf'],
+  ru: ['belebele_rus_Cyrl_cf', 'chegeka_rus', 'mathlogic_qa_rus_cf', 'mera_openbookqa_rus_cf', 'mera_worldtree_rus_cf', 'mkqa_rus:_average', 'mlmm_arc_rus_cf:challenge', 'mlmm_hellaswag_rus_cf', 'mlmm_mmlu_rus_cf:_average', 'mlmm_truthfulqa_rus_cf:mc1', 'mlmm_truthfulqa_rus_cf:mc2', 'parus_rus_cf', 'rcb_rus_cf', 'rummlu_rus_cf:_average', 'sber_squad_rus', 'tydiqa_rus', 'xcodah_rus_cf', 'xcsqa_rus_cf', 'xnli2.0_rus_cf', 'xquad_rus', 'xstory_cloze_rus_cf', 'xwinograd_rus_cf'],
+  sw: ['afric_mmlu_swa_cf:_average', 'afric_xnli_swa_cf', 'belebele_swh_Latn_cf', 'community_arc_swa_cf:challenge', 'community_arc_swa_cf:easy', 'community_mmlu_swa_cf', 'kenswquad_swa', 'm3exams_swa_cf', 'openai_mmlu_swa_cf:_average', 'tydiqa_swa', 'xcodah_swa_cf', 'xcopa_swa_cf', 'xcsqa_swa_cf', 'xnli2.0_swa_cf', 'xnli_swa_cf', 'xstory_cloze_swa_cf'],
+  te: ['belebele_tel_Telu_cf', 'community_hellaswag_tel_cf', 'indicnxnli_tel_cf', 'indicqa_tel', 'indicxcopa_tel_cf', 'mlmm_arc_tel_cf:challenge', 'mlmm_hellaswag_tel_cf', 'mlmm_mmlu_tel_cf:_average', 'mlmm_truthfulqa_tel_cf:mc1', 'mlmm_truthfulqa_tel_cf:mc2', 'tydiqa_tel', 'xstory_cloze_tel_cf'],
+  th: ['belebele_tha_Thai_cf', 'community_hellaswag_tha_cf', 'm3exams_tha_cf', 'meta_mmlu_tha_cf:_average', 'mkqa_tha:_average', 'thai_exams_tha_cf:_average', 'thai_exams_tha_cf:tgat', 'thaiqa_tha', 'wsci_tha_cf', 'xcopa_tha_cf', 'xnli2.0_tha_cf', 'xnli_tha_cf', 'xquad_tha'],
+  tr: ['belebele_tur_Latn_cf', 'community_arc_tur_cf:easy', 'community_hellaswag_tur_cf', 'community_mmlu_tur_cf:_average', 'community_truthfulqa_tur_cf:mc1', 'community_truthfulqa_tur_cf:mc2', 'community_xwinograd_tur_cf', 'exams_tur_cf:_average', 'mkqa_tur:_average', 'tquadv2_tur', 'xcopa_tur_cf', 'xnli2.0_tur_cf', 'xnli_tur_cf', 'xquad_tur'],
+  zh: ['agieval_zho_cf:_average', 'belebele_zho_Hans_cf', 'c3_zho_cf', 'ceval_zho_cf:_average', 'chinese_squad_zho', 'cmath_zho_cf', 'cmmlu_zho_cf:_average', 'cmnli_zho_cf', 'cmrc2018_zho', 'm3exams_zho_cf', 'mkqa_zho:_average', 'mlmm_arc_zho_cf:challenge', 'mlmm_hellaswag_zho_cf', 'mlmm_mmlu_zho_cf:_average', 'mlmm_truthfulqa_zho_cf:mc1', 'mlmm_truthfulqa_zho_cf:mc2', 'ocnli_zho_cf', 'pawsx_zho_cf', 'xcodah_zho_cf', 'xcopa_zho_cf', 'xcsqa_zho_cf', 'xnli2.0_zho_cf', 'xnli_zho_cf', 'xquad_zho', 'xstory_cloze_zho_cf', 'xwinograd_zho_cf']
+};
+
+const LINE_SETTINGS = {
+  width: 2.5,
+  type: "scatter",
+  mode: "lines+markers",
+};
+
+const DEFAULT_LAYOUT = {
+  font: {
+    family: "apple-system, Arial, sans-serif",
+  },
+  title: {
+    font: {
+      size: 15,
+    },
+  },
+  xaxis: {
+    title: {
+      text: "Training Tokens (billions)",
+      font: {
+        size: 14,
+      },
+    },
+    tickfont: {
+      size: 12,
+    },
+    showgrid: false,
+    mirror: true,
+    ticks: "outside",
+    showline: true,
+  },
+  yaxis: {
+    title: {
+      font: {
+        size: 14,
+      },
+      standoff: 10,
+    },
+    showgrid: false,
+    mirror: true,
+    ticks: "outside",
+    showline: true,
+    tickfont: {
+      size: 12,
+    },
+  },
+  height: 300, // You can adjust this value
+  autosize: true,
+  legend: {
+    orientation: 'h',        // Set to 'h' for horizontal legend (required for columns)
+    yanchor: 'bottom',
+    y: 0,                    // Position at the bottom
+    xanchor: 'right',
+    x: 1,                    // Position at the right
+    traceorder: 'normal',
+    font: { size: 12 },
+    tracegroupgap: 0,        // Space between legend items
+    bgcolor: 'rgba(255, 255, 255, 0.8)' // White background with 70% transparency (1 - 0.3 = 70%)
+  },
+  margin: {
+    t: 25,
+    b: 60,
+    l: 60,
+    r: 40,
+  },
+};
+
+export function initPlotApplets() {
+  const plotContainers = document.querySelectorAll('.task-signal-plot');
+  plotContainers.forEach(container => {
+    initPlotApplet(container);
+  });
+}
+
+function initPlotApplet(container) {
+  const defaultLanguage = container.dataset.language || 'Arabic';
+  const defaultTask = container.dataset.task || '';
+  const defaultMetric = container.dataset.metric || '';
+  const groupSeeds = container.dataset.groupSeeds === 'true';
+  const showControls = container.dataset.showControls === 'true';
+  const taskMetrics = (container.dataset.taskMetrics || 'monotonicity,snr,ordering,randomness').split(",");
+
+  const controls = createControls(container, defaultLanguage, defaultTask, defaultMetric, taskMetrics);
+  if (!showControls)
+    controls.style.display = 'none';
+  container.appendChild(controls);
+
+  const plotContainer = document.createElement('div');
+  plotContainer.className = 'plot-container';
+  container.appendChild(plotContainer);
+
+  const statsContainer = document.createElement('div');
+  statsContainer.className = 'stats-container';
+  container.appendChild(statsContainer);
+
+
+  // Create an initial empty plot
+  Plotly.newPlot(plotContainer, []);
+
+  // Set up the resize function
+  const resizePlot = () => {
+    const width = container.offsetWidth;
+    Plotly.relayout(plotContainer, { width: width });
+  };
+
+  // Add resize listener
+  window.addEventListener('resize', resizePlot);
+
+  // Initial resize
+  resizePlot();
+
+  // Load the initial data
+  updateLanguageTasks(container, defaultTask, defaultMetric, groupSeeds, taskMetrics);
+}
+
+function createControls(container, defaultLanguage, defaultTask, defaultMetric, taskMetrics) {
+  const controls = document.createElement('div');
+  controls.className = 'controls';
+
+  const languageSelect = createSelect('language', Object.keys(languageMap), () => updateLanguageTasks(container, '', '', true, taskMetrics));
+  languageSelect.value = defaultLanguage;
+
+  const taskSelect = createSelect('task', [], () => updateMetrics(container, '', true, taskMetrics));
+  const metricSelect = createSelect('metric', [], () => updatePlot(container, taskMetrics));
+
+  controls.appendChild(createControlGroup('Language:', languageSelect));
+  controls.appendChild(createControlGroup('Task:', taskSelect));
+  controls.appendChild(createControlGroup('Metric:', metricSelect));
+
+  return controls;
+}
+
+function createSelect(id, options, onChangeHandler) {
   const select = document.createElement('select');
+  select.id = id;
   options.forEach(option => {
     const optionElement = document.createElement('option');
-    if (typeof option === 'object' && option.value && option.label) {
-      optionElement.value = option.value;
-      optionElement.textContent = option.label;
-    } else {
-      optionElement.value = option;
-      optionElement.textContent = option;
-    }
+    optionElement.value = option;
+    optionElement.textContent = option;
     select.appendChild(optionElement);
   });
-  select.addEventListener('change', onChange);
+  select.addEventListener('change', onChangeHandler);
   return select;
 }
 
-function createPerTaskResultsTable(data, tableType, metric) {
-  const tableWrapper = document.createElement('div');
-  tableWrapper.className = 'table-wrapper fine-tasks-table-wrapper';
+function createControlGroup(labelText, inputElement) {
+  const group = document.createElement('div');
+  group.className = 'control-group';
+  
+  const label = document.createElement('label');
+  label.textContent = labelText;
+  label.className = 'control-label';
+  
+  group.appendChild(label);
+  group.appendChild(inputElement);
+  
+  return group;
+}
 
-  const table = document.createElement('table');
-  table.className = 'results-table fine-tasks-results-table';
+async function updateLanguageTasks(container, defaultTask = '', defaultMetric = '', groupSeeds, taskMetrics) {
+  const languageSelect = container.querySelector('#language');
+  const taskSelect = container.querySelector('#task');
+  const language = languageSelect.value;
+  const langCode = languageMap[language];
 
-  const columns = ['Task', 'Type', ...(tableType === 'gen' ? ['f1', 'prefix_match'] : ['acc', 'acc_norm', 'acc_norm_token', 'acc_norm_pmi'])];
-  
-  const columnNameMap = {
-    // 'Task': 'Task',
-    // 'Type': 'Type',
-    // 'f1': 'f1',
-    // 'prefix_match': 'prefix_match',
-    // 'acc': 'acc',
-    'acc_norm': 'acc_char',
-    'acc_norm_token': 'acc_token',
-    'acc_norm_pmi': 'acc_pmi',
-    'prefix_match': 'prefix'
-  };
+  taskSelect.innerHTML = '<option value="">Loading tasks...</option>';
 
-  const taskMetricMap = {
-    'max_score': 'score',
-    'avg_snr': 'snr',
-    'avg_spearman': 'monotonicity',
-    'max_n_std': 'non-randomness',
-    'avg_kendall_tau_a': 'ordering'
-    // 'avg_spearman': 'monotonicity',
+  try {
+    const tasks = await getTasksForLanguage(langCode);
+    
+    taskSelect.innerHTML = '';
+    if (tasks.length > 0) {
+      tasks.forEach(task => {
+        const option = document.createElement('option');
+        option.value = task;
+        option.textContent = truncateText(task, 25); // Reduced from 30 to 25
+        option.title = task; // Set full task name as title for tooltip
+        taskSelect.appendChild(option);
+      });
+      
+      if (defaultTask && tasks.includes(defaultTask)) {
+        taskSelect.value = defaultTask;
+      } else {
+        taskSelect.selectedIndex = 0;
+      }
+      
+      await updateMetrics(container, defaultMetric, groupSeeds, taskMetrics);
+    } else {
+      taskSelect.innerHTML = '<option value="">No tasks available</option>';
+      clearPlot(container);
+    }
+  } catch (error) {
+    console.error('Error fetching tasks:', error);
+    taskSelect.innerHTML = '<option value="">Error loading tasks</option>';
+    clearPlot(container);
   }
+}
+
+async function getTasksForLanguage(langCode) {
+  return taskLists[langCode] || [];
+}
+
+async function updateMetrics(container, defaultMetric = '', groupSeeds, taskMetrics) {
+  const language = container.querySelector('#language').value;
+  const task = container.querySelector('#task').value;
+  const langCode = languageMap[language];
+  const metricSelect = container.querySelector('#metric');
 
-  const header = table.createTHead();
-  const headerRow = header.insertRow();
-  columns.forEach(column => {
-    const th = document.createElement('th');
-    th.textContent = columnNameMap[column] || column;
+  metricSelect.innerHTML = '<option value="">Loading metrics...</option>';
 
-    if (th.textContent !== "Task" && th.textContent !== "Type") {
-        th.textContent += " " + (taskMetricMap[metric] || metric);
+  try {
+    const metrics = await getMetricsForTask(langCode, task);
+    
+    metricSelect.innerHTML = '';
+    metrics.forEach(metric => {
+      const option = document.createElement('option');
+      option.value = metric;
+      option.textContent = metric;
+      metricSelect.appendChild(option);
+    });
+
+    if (defaultMetric && metrics.includes(defaultMetric)) {
+      metricSelect.value = defaultMetric;
+    } else if (metricSelect.options.length > 0) {
+      metricSelect.selectedIndex = 0;
     }
-    th.title = th.textContent;
-    if (column === 'Type')
-      th.style.width = '40px';
-    headerRow.appendChild(th);
+
+    await updatePlot(container, taskMetrics);
+  } catch (error) {
+    console.error('Error fetching metrics:', error);
+    metricSelect.innerHTML = '<option value="">Error loading metrics</option>';
+    clearPlot(container);
+  }
+}
+
+async function getMetricsForTask(langCode, task) {
+  return new Promise((resolve, reject) => {
+    Papa.parse(`data/${langCode}/${task}_stats.csv`, {
+      download: true,
+      header: true,
+      complete: function(results) {
+        const metrics = [...new Set(results.data.map(row => row.metric).filter(metric => metric))];
+        resolve(metrics);
+      },
+      error: function(error) {
+        console.error('Error fetching metrics:', error);
+        reject(error);
+      }
+    });
   });
+}
+
+function updatePlot(container, taskMetrics) {
+  const language = container.querySelector('#language').value;
+  const task = container.querySelector('#task').value;
+  const metric = container.querySelector('#metric').value;
+  const title = container.dataset.title;
+  const langCode = languageMap[language];
+
+  if (!langCode || !task || !metric) {
+    clearPlot(container);
+    return;
+  }
+
+  const dataUrl = `data/${langCode}/${task}_data.csv`;
+  const statsUrl = `data/${langCode}/${task}_stats.csv`;
+
+  Promise.all([
+    new Promise((resolve, reject) => {
+      Papa.parse(dataUrl, {
+        download: true,
+        header: true,
+        dynamicTyping: true,
+        complete: resolve,
+        error: reject
+      });
+    }),
+    new Promise((resolve, reject) => {
+      Papa.parse(statsUrl, {
+        download: true,
+        header: true,
+        dynamicTyping: true,
+        complete: resolve,
+        error: reject
+      });
+    })
+  ]).then(([dataResult, statsResult]) => {
+    const taskData = dataResult.data;
+    const statsData = statsResult.data;
+    plotData(container, taskData, statsData, metric, title, taskMetrics);
+  }).catch(error => {
+    console.error('Error parsing CSV:', error);
+    clearPlot(container);
+  });
+}
+
+function plotData(container, data, stats, metric, title, taskMetrics) {
+  const groupSeeds = container.dataset.groupSeeds === 'true';
+  const sortedData = sortDataByTokens(data);
+  const groupedData = groupDataByRunname(sortedData, groupSeeds, metric);
+  const interpolatedData = interpolateData(groupedData, metric);
+  const smoothedData = smoothData(interpolatedData, metric);
+  const traces = createTraces(smoothedData, metric);
+
+  const plotContainer = container.querySelector('.plot-container');
+
+  const layout = _.merge({}, DEFAULT_LAYOUT, {
+    title: { text: `${title}` },
+    xaxis: { 
+      title: { text: 'Training Tokens (billions)' },
+      tickvals: [0, 5, 10, 15, 20, 25],
+      ticktext: ['0', '5B', '10B', '15B', '20B', '25B'],
+      tickangle: 45,
+      range: [0, 30], // Set the range to start from 0 and end at 30B
+    },
+    yaxis: { 
+      title: { text: 'Score' },
+      range: [Math.min(...traces.flatMap(trace => trace.y)) * 0.95, Math.max(...traces.flatMap(trace => trace.y)) * 1.05], // Add 5% padding to the top and bottom
+    },
+    width: container.offsetWidth,
+  });
+
+  Plotly.newPlot(plotContainer, traces, layout, {responsive: true});
 
-  const body = table.createTBody();
-  data.forEach(row => {
-    if (Object.values(row).every(value => value === '' || value === undefined || value === null)) {
-      return;
+  // Display statistics
+  displayStatistics(container, stats, metric, taskMetrics);
+}
+
+function displayStatistics(container, stats, metric, taskMetrics) {
+  const statsContainer = container.querySelector('.stats-container');
+  const metricStats = stats.find(stat => stat.metric === metric);
+  if (metricStats) {
+    statsContainer.innerHTML = `
+      <div class="compact-stats${taskMetrics.length === 1 ? '-single' : ''}">
+        ${taskMetrics.includes('monotonicity') ? '<span title="Average Spearman Correlation">Monotonicity: ' + metricStats.avg_spearman.toFixed(2) + '</span>' : ''}
+        ${taskMetrics.includes('snr') ? '<span title="Average Signal-to-Noise Ratio">Signal-to-Noise: ' + metricStats.avg_snr.toFixed(2) + '</span>' : ''}
+        ${taskMetrics.includes('ordering') ? '<span title="Average Kendall Tau-a">Ordering Consistency: ' + metricStats.avg_kendall_tau_a.toFixed(2) + '</span>' : ''}
+        ${taskMetrics.includes('randomness') ? '<span title="Max N Standard Deviations">Non-Randomness: ' + metricStats.max_n_std.toFixed(2) + '</span>' : ''}
+      </div>
+    `;
+  } else {
+    statsContainer.innerHTML = '<p>No statistics available for this metric.</p>';
+  }
+}
+
+function getReducedTickValues(tokens) {
+  const uniqueTokens = [...new Set(tokens)].sort((a, b) => a - b);
+  const tokenCount = uniqueTokens.length;
+  const targetTickCount = 10; // Adjust this value to increase/decrease the number of ticks
+
+  if (tokenCount <= targetTickCount) {
+    return uniqueTokens;
+  }
+
+  const stride = Math.ceil(tokenCount / targetTickCount);
+  return uniqueTokens.filter((_, index) => index % stride === 0);
+}
+
+function formatTickLabel(value) {
+  if (value >= 1e9) {
+    return (value / 1e9).toFixed(1) + 'B';
+  } else if (value >= 1e6) {
+    return (value / 1e6).toFixed(1) + 'M';
+  } else if (value >= 1e3) {
+    return (value / 1e3).toFixed(1) + 'K';
+  }
+  return value.toString();
+}
+
+function computeStatistics(data, metric) {
+  const stats = {
+    avg_spearman: 0,
+    avg_kendall_tau_a: 0,
+    avg_snr: 0,
+    max_n_std: 0
+  };
+
+  const baselineRun = Object.keys(data).find(key => key.toLowerCase().includes('baseline'));
+  const nonBaselineRuns = Object.keys(data).filter(key => key !== baselineRun);
+
+  // Compute statistics for each non-baseline run
+  nonBaselineRuns.forEach(run => {
+    const runData = data[run];
+    const tokens = runData.map(row => row.tokens);
+    const scores = runData.map(row => row[metric]);
+
+    // Spearman correlation
+    stats.avg_spearman += spearmanCorrelation(tokens, scores);
+
+    // Kendall Tau-a
+    const lastHalf = Math.floor(runData.length / 2);
+    const kendallTauValues = [];
+    for (let i = lastHalf; i < runData.length - 1; i++) {
+      kendallTauValues.push(kendallTauA(scores.slice(0, i + 1), scores.slice(0, i + 2)));
+    }
+    stats.avg_kendall_tau_a += _.mean(kendallTauValues);
+
+    // SNR and max_n_std
+    if (baselineRun) {
+      const baselineScores = data[baselineRun].map(row => row[metric]);
+      const stdDev = standardDeviation(scores);
+      stats.avg_snr += _.mean(scores) / stdDev;
+      stats.max_n_std = Math.max(stats.max_n_std, (_.max(scores) - _.mean(baselineScores)) / stdDev);
     }
+  });
+
+  // Average the statistics
+  const numRuns = nonBaselineRuns.length;
+  stats.avg_spearman /= numRuns;
+  stats.avg_kendall_tau_a /= numRuns;
+  stats.avg_snr /= numRuns;
+
+  return stats;
+}
+
+function spearmanCorrelation(x, y) {
+  const n = x.length;
+  const rankX = rankData(x);
+  const rankY = rankData(y);
+  
+  let sum_d_squared = 0;
+  for (let i = 0; i < n; i++) {
+    const d = rankX[i] - rankY[i];
+    sum_d_squared += d * d;
+  }
+  
+  return 1 - (6 * sum_d_squared) / (n * (n * n - 1));
+}
+
+function rankData(data) {
+  const sorted = [...data].sort((a, b) => a - b);
+  return data.map(x => sorted.indexOf(x) + 1);
+}
+
+function kendallTauA(x, y) {
+  const n = x.length;
+  let concordant = 0;
+  let discordant = 0;
+
+  for (let i = 0; i < n; i++) {
+    for (let j = i + 1; j < n; j++) {
+      const sign_x = Math.sign(x[j] - x[i]);
+      const sign_y = Math.sign(y[j] - y[i]);
+      if (sign_x * sign_y > 0) concordant++;
+      else if (sign_x * sign_y < 0) discordant++;
+    }
+  }
+
+  return (concordant - discordant) / (n * (n - 1) / 2);
+}
+
+function standardDeviation(values) {
+  const mean = _.mean(values);
+  const squareDiffs = values.map(value => {
+    const diff = value - mean;
+    return diff * diff;
+  });
+  const avgSquareDiff = _.mean(squareDiffs);
+  return Math.sqrt(avgSquareDiff);
+}
+
+function interpolateData(data, metric) {
+  return _.mapValues(data, (rows) => {
+    const sortedRows = _.sortBy(rows, 'tokens');
+    const allTokens = _.uniq(_.flatMap(Object.values(data), rows => rows.map(r => r.tokens))).sort((a, b) => a - b);
     
-    const tr = body.insertRow();
-    columns.forEach(column => {
-      const td = tr.insertCell();
-      let value = row[column];
-      if (column === 'Task') {
-        const fullTaskName = value; // Store the full task name
-        const parts = value.split('|');
-        value = parts.length > 1 ? parts[1] : value;
-        value = value.split('_mcf')[0].split('_cf')[0];
-        td.title = fullTaskName; // Set the title attribute to show the full name on hover
-      } else if (column === 'Type') {
-        // Keep the task type as is
-      } else if (typeof value === 'number') {
-        value = value.toFixed(2);
-      } else if (value && !isNaN(parseFloat(value))) {
-        value = parseFloat(value).toFixed(2);
-      } else {
-        value = '';
-      }
-      td.textContent = value;
+    return allTokens.map(token => {
+      const exactMatch = _.find(sortedRows, { tokens: token });
+      if (exactMatch) return exactMatch;
+
+      const lowerRow = _.findLast(sortedRows, r => r.tokens < token);
+      const upperRow = _.find(sortedRows, r => r.tokens > token);
+
+      if (!lowerRow) return { ...upperRow, tokens: token };
+      if (!upperRow) return { ...lowerRow, tokens: token };
+
+      const ratio = (token - lowerRow.tokens) / (upperRow.tokens - lowerRow.tokens);
+      const interpolatedMetric = lowerRow[metric] + (upperRow[metric] - lowerRow[metric]) * ratio;
+
+      return {
+        ...lowerRow,
+        tokens: token,
+        [metric]: interpolatedMetric
+      };
     });
   });
+}
 
-  tableWrapper.appendChild(table);
-  return tableWrapper;
+function smoothData(data, metric, windowSize = 3) {
+  return _.mapValues(data, (rows) => {
+    return rows.map((row, index, array) => {
+      const window = array.slice(Math.max(0, index - windowSize + 1), index + 1);
+      const smoothedMetric = _.meanBy(window, r => r[metric]);
+      return { ...row, [metric]: smoothedMetric };
+    });
+  });
 }
 
-export function initFineTasks(containerId) {
-  const container = document.getElementById(containerId);
-  if (!container) return;
+function sortDataByTokens(data) {
+  return _.sortBy(data, 'tokens');
+}
 
-  const perTaskTitleElement = document.createElement('h3');
-  perTaskTitleElement.textContent = 'Task Results';
-  perTaskTitleElement.className = 'fine-tasks-title';
+function groupDataByRunname(data, groupSeeds, metric) {
+  // Remove null or undefined runs
+  data = data.filter(row => row.runname != null && row.runname !== 'null_undefined');
 
-  const perTaskTableContainer = document.createElement('div');
-  perTaskTableContainer.className = 'table-container';
+  if (!groupSeeds) {
+    return _.groupBy(data, row => `${processRunName(row.runname)}_${row.seed}`);
+  }
 
-  let perTaskDataTable;
+  const grouped = _.groupBy(data, row => processRunName(row.runname));
+  
+  return _.mapValues(grouped, (rows) => {
+    const stepGroups = _.groupBy(rows, 'tokens');
+    return _.map(stepGroups, (stepRows) => {
+      const meanMetric = _.meanBy(stepRows, row => parseFloat(row[metric]) || 0);
+      return {
+        ...stepRows[0],
+        [metric]: meanMetric
+      };
+    });
+  });
+}
 
-  function updatePerTaskResults() {
-    const language = languageDropdownPerTask.value;
-    const metric = metricDropdownPerTask.value;
-    const tableType = tableTypeDropdownPerTask.value;
-    const taskFolder = taskFolderDropdownPerTask.value;
+function processRunName(runname) {
+  for (const [key, value] of Object.entries(runNameMap)) {
+    if (runname.includes(key)) {
+      return value;
+    }
+  }
+  return runname;
+}
 
-    const languageCode = languageMap[language];
+function createTraces(groupedData, metric) {
+  const colorsMapping = new Map();
+  const sortedRunnames = Object.keys(groupedData).sort((a, b) => {
+    if (a.includes('baseline')) return 1;
+    if (b.includes('baseline')) return -1;
+    return a.localeCompare(b);
+  });
 
-    if (!languageCode) {
-      console.error(`Language code not found for ${language}`);
-      perTaskTableContainer.innerHTML = `<p>Error: Language code not found for ${language}</p>`;
-      return;
-    }
+  return sortedRunnames.map((runname, index) => {
+    const color = getColorForTrace(runname, colorsMapping, index);
+    return {
+      x: groupedData[runname].map(row => row.tokens),
+      y: groupedData[runname].map(row => row[metric]),
+      name: runname,
+      line: { 
+        color: color,
+        shape: 'spline',
+        ...LINE_SETTINGS
+      },
+      marker: {
+        color: color,
+        size: 6,
+      },
+      mode: 'lines+markers',
+    };
+  });
+}
 
-    let url = `data/tasks/${taskFolder}/${languageCode}/${metric}/${tableType}_stats.csv`;
-
-    fetch(url)
-      .then(response => {
-        if (!response.ok) {
-          throw new Error(`HTTP error! status: ${response.status}`);
-        }
-        return response.text();
-      })
-      .then(csvText => {
-        const results = Papa.parse(csvText, { header: true }).data;
-        perTaskTableContainer.innerHTML = '';
-        const tableWrapper = createPerTaskResultsTable(results, tableType, metric);
-        perTaskTableContainer.appendChild(tableWrapper);
-
-        if (perTaskDataTable) {
-          perTaskDataTable.destroy();
-        }
-
-        perTaskDataTable = new DataTable('.fine-tasks-results-table', {
-          perPage: 10,
-          perPageSelect: false,
-          searchable: false,
-          sortable: true,
-          fixedHeight: true,
-          labels: {
-            info: ''  // This removes the "Showing 1 to X of Y entries" text
-          }
-        });
-
-      })
-      .catch(error => {
-        console.error('Error fetching CSV:', error);
-        perTaskTableContainer.innerHTML = `<p>Error loading data: ${error.message}</p>`;
-      });
+function getColorForTrace(traceName, colorsMapping, index) {
+  const reusedColor = colorsMapping.get(traceName);
+  if (reusedColor) {
+    return reusedColor;
   }
 
-  const perTaskControls = document.createElement('div');
-  perTaskControls.className = 'controls fine-tasks-controls';
-
-  // Task folder control group
-  const taskFolderControlGroup = document.createElement('div');
-  taskFolderControlGroup.className = 'control-group';
-  const taskFolderLabelPerTask = document.createElement('label');
-  taskFolderLabelPerTask.textContent = 'Task Set: ';
-  const taskFolderDropdownPerTask = createDropdown(taskFolders, updatePerTaskResults);
-  taskFolderDropdownPerTask.value = 'selected'; // Set default to FineTasks
-  taskFolderControlGroup.appendChild(taskFolderLabelPerTask);
-  taskFolderControlGroup.appendChild(taskFolderDropdownPerTask);
-
-  // Language control group
-  const languageControlGroup = document.createElement('div');
-  languageControlGroup.className = 'control-group';
-  const languageLabelPerTask = document.createElement('label');
-  languageLabelPerTask.textContent = 'Language: ';
-  const languageDropdownPerTask = createDropdown(Object.keys(languageMap), updatePerTaskResults);
-  languageControlGroup.appendChild(languageLabelPerTask);
-  languageControlGroup.appendChild(languageDropdownPerTask);
-
-  // Table type control group
-  const tableTypeControlGroup = document.createElement('div');
-  tableTypeControlGroup.className = 'control-group';
-  const tableTypeLabelPerTask = document.createElement('label');
-  tableTypeLabelPerTask.textContent = 'Type: ';
-  const tableTypeDropdownPerTask = createDropdown(tableTypes, updatePerTaskResults);
-  tableTypeControlGroup.appendChild(tableTypeLabelPerTask);
-  tableTypeControlGroup.appendChild(tableTypeDropdownPerTask);
-
-  // Metric control group
-  const metricControlGroup = document.createElement('div');
-  metricControlGroup.className = 'control-group';
-  const metricLabelPerTask = document.createElement('label');
-  metricLabelPerTask.textContent = 'Criteria: ';
-  const metricDropdownPerTask = createDropdown(metricTypes, updatePerTaskResults);
-  metricDropdownPerTask.value = 'max_score'; // Set default to Max Score
-  metricControlGroup.appendChild(metricLabelPerTask);
-  metricControlGroup.appendChild(metricDropdownPerTask);
-
-  perTaskControls.appendChild(taskFolderControlGroup);
-  perTaskControls.appendChild(languageControlGroup);
-  perTaskControls.appendChild(tableTypeControlGroup);
-  perTaskControls.appendChild(metricControlGroup);
-
-  container.appendChild(perTaskControls);
-  // container.appendChild(perTaskTitleElement);
-  container.appendChild(perTaskTableContainer);
-
-  // Initialize with default values
-  updatePerTaskResults();
+  const color = getColor(index);
+  colorsMapping.set(traceName, color);
+  return color;
+}
+
+function clearPlot(container) {
+  const plotContainer = container.querySelector('.plot-container');
+  Plotly.purge(plotContainer);
+}
+
+function truncateText(text, maxLength) {
+  if (text.length <= maxLength) return text;
+  return text.substr(0, maxLength - 2) + '..';
 }
+

app/src/content/assets/finetasks/data/ar/acva_ara:_average_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee86019600943234de0d00cb7f2cfb5f08adea529e281c47fb11ab39e904fa14
+size 26104

app/src/content/assets/finetasks/data/ar/acva_ara:_average_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79551f7eeb2579538604681929741203205e6150f95187ea5319e3e9671f634e
+size 1078

app/src/content/assets/finetasks/data/ar/alfgahafa_mlqa_ara_cf_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0753a9fb838808ff6855bfcce87eb7d716d406dff82985e64bd72abf3e0eeed6
+size 20564

app/src/content/assets/finetasks/data/ar/alfgahafa_mlqa_ara_cf_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b42cd429953188f1e3a2f61a3bbcd3aa669421bac407a5f2843b9ad3bc287b9b
+size 903

app/src/content/assets/finetasks/data/ar/alghafa_arc_ara_cf:easy_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:810a68eb754b4f0a3acae2a34c311676c78d926fd88e34e1c0bb9be949e3aa20
+size 18155

app/src/content/assets/finetasks/data/ar/alghafa_arc_ara_cf:easy_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:642e379750d340963d86ff023426787891d7cb494bf135c33be48c0c9897519f
+size 908

app/src/content/assets/finetasks/data/ar/alghafa_exams_ara_cf:_average_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf7f6df15ba9e2c552c721bed4d292cf75a8bf6b3f3cbd5f65c9903b99e463d0
+size 24386

app/src/content/assets/finetasks/data/ar/alghafa_exams_ara_cf:_average_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9b25491622030909b5b075cd7744fcad61fabe7103253c14355710762cbdc6d6
+size 928

app/src/content/assets/finetasks/data/ar/alghafa_facts_ara_cf_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b3b1f369ae9a64e27702437a049456d90fff09c62133a0232cd146a19bfb1bba
+size 17318

app/src/content/assets/finetasks/data/ar/alghafa_facts_ara_cf_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:19159aa9195f4a9eebbc6a91431d65dce630ec97edf39e32fe3c0f8dc302e546
+size 834

app/src/content/assets/finetasks/data/ar/alghafa_meta_dialects_ara_cf_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:714e3326ff11bfebb268366315fa4b5cf305b9cb8174c451db33773f5ac88d78
+size 18138

app/src/content/assets/finetasks/data/ar/alghafa_meta_dialects_ara_cf_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e7a37090de73b4fc41f1011e031ad56ff95e3883662275daf8c67656e166b5f9
+size 935

app/src/content/assets/finetasks/data/ar/alghafa_mmlu_ara_cf:_average_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6959492567e052a2f9251d092f449dda7ed9118daca1441f5c146e6d2761e10c
+size 23032

app/src/content/assets/finetasks/data/ar/alghafa_mmlu_ara_cf:_average_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3f579eac42e69687634fa0d27ca738a80d5fd854b4dc33ef069210def32a7394
+size 937

app/src/content/assets/finetasks/data/ar/alghafa_openbookqa_ara_cf_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9fe35704efef4b670a7f3ec6f64d64aa2e14f387caaa12b7c5da0eda18c4078a
+size 22998

app/src/content/assets/finetasks/data/ar/alghafa_openbookqa_ara_cf_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e137221596d32ecd3f77c2f451c87e8ff8743a286816f974e7457290a2cfaec
+size 925

app/src/content/assets/finetasks/data/ar/alghafa_piqa_ara_cf_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a436d5d6e494aa37f2f9c4e4f14c2376d97d1c48ed116ca4f9c6f65caf0fbc3f
+size 18478

app/src/content/assets/finetasks/data/ar/alghafa_piqa_ara_cf_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f01a72195ce586f94d8dad8210e5accd5459bf83712ad968b149701dfe4b9e8
+size 880

app/src/content/assets/finetasks/data/ar/alghafa_race_ara_cf_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:80b15aea8264a8bdac278cf2ac0d07cbcc0e8c7ccac150ace4e26dd65471e6fc
+size 18432

app/src/content/assets/finetasks/data/ar/alghafa_race_ara_cf_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:80ae78c9142cf89d1be55977578c2ca041838a5dbaa736b40958a13730e46ae9
+size 893

app/src/content/assets/finetasks/data/ar/alghafa_rating_sentiment_ara_cf_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bdc28b21863e88fcc8fac6245d25da05db90d7da234708b341636400b2584769
+size 18023

app/src/content/assets/finetasks/data/ar/alghafa_rating_sentiment_ara_cf_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3291cd3da2406bdcc358764cf42ae1a21cf9d1c07b4daf5998e7695a09317c37
+size 936

app/src/content/assets/finetasks/data/ar/alghafa_rating_sentiment_no_neutral_ara_cf_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab37a3551307dadb3260bcd768b554954b7451493ce1e07bbaf8465d25f3e09b
+size 16661

app/src/content/assets/finetasks/data/ar/alghafa_rating_sentiment_no_neutral_ara_cf_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:20a08683aa5cdd6a8a64efcffcf942a4a26dcf2550a2d770b0ff46b8c40ecffe
+size 970

app/src/content/assets/finetasks/data/ar/alghafa_sciqa_ara_cf_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a29f2dcacf4f8d53096dddfbeb1f5cb8eb8e8a5354dc2f979a1908c5541ebcd6
+size 23819

app/src/content/assets/finetasks/data/ar/alghafa_sciqa_ara_cf_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e6cac24425904dec66f4bb38aa34d65d0c1a6bc539baf5ea5300c5f7bc362626
+size 894

app/src/content/assets/finetasks/data/ar/alghafa_sentiment_ara_cf_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c69085a0ca2df0adc4f9ec3c3b9857adad82d6749f41d88e3d43ba16e6d936d3
+size 17942

app/src/content/assets/finetasks/data/ar/alghafa_sentiment_ara_cf_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1db01058cac603bcb5b0a991b39d9499a3537ab0da1ad36eb1c3b317c8d5ff4
+size 903

app/src/content/assets/finetasks/data/ar/arcd_ara_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f5b91b32d5c9a58a34ac035fd9e880de1256f5d0c47edfed7fe591abed789fa
+size 15849

app/src/content/assets/finetasks/data/ar/arcd_ara_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d15a351c2e4cf3dcc3372637baafd4821397bb7ab00c81704d91ec8b55e6a31
+size 478

app/src/content/assets/finetasks/data/ar/belebele_arb_Arab_cf_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6464f54a9a933b4b47c9c513c907ae358909518998ad5db01d8580578b77a1c6
+size 23912

app/src/content/assets/finetasks/data/ar/belebele_arb_Arab_cf_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fda4577c55a3298b9dc6f6a2e148aaf6a4038f8604811ca339f88d3c3f6e7573
+size 903

app/src/content/assets/finetasks/data/ar/boolq_ara_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54675b0af158ca756a5c08ea1b6315f757df93827a3a107b4208b135bdf6d8db
+size 18834

app/src/content/assets/finetasks/data/ar/boolq_ara_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:32457679a2dc38045522dab504c13590c7199bcb452f65acfcf337fbbb3bdc2c
+size 1042

app/src/content/assets/finetasks/data/ar/community_arc_hin_cf:challenge_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2649f9802e39e19da555d2e42851281cca18826534d23246bffd8b15a43e326a
+size 14390

app/src/content/assets/finetasks/data/ar/community_arc_hin_cf:challenge_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aabcf0d879390556fa664c0fb532afa47580407e37f2552026dbfffab89ebf57
+size 469

app/src/content/assets/finetasks/data/ar/community_arc_hin_cf:easy_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2649f9802e39e19da555d2e42851281cca18826534d23246bffd8b15a43e326a
+size 14390

app/src/content/assets/finetasks/data/ar/community_arc_hin_cf:easy_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d213b818d04764e3bd9a0d0ad57ab9e0e38fe26d8db51942c12c1b8eb92f3636
+size 449

app/src/content/assets/finetasks/data/ar/community_arc_swa_cf:challenge_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2649f9802e39e19da555d2e42851281cca18826534d23246bffd8b15a43e326a
+size 14390

app/src/content/assets/finetasks/data/ar/community_arc_swa_cf:challenge_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a7f3601c8b4750a98708960e1df4c574b2038821e4a5740837d52ad770bbbd3c
+size 469

app/src/content/assets/finetasks/data/ar/community_arc_swa_cf:easy_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2649f9802e39e19da555d2e42851281cca18826534d23246bffd8b15a43e326a
+size 14390

app/src/content/assets/finetasks/data/ar/community_arc_swa_cf:easy_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f25f52a5fe47096977b9dd294e354c65f54225ec87a7cde264933b5229ca0a67
+size 449

app/src/content/assets/finetasks/data/ar/community_arc_tur_cf:easy_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2649f9802e39e19da555d2e42851281cca18826534d23246bffd8b15a43e326a
+size 14390

app/src/content/assets/finetasks/data/ar/community_arc_tur_cf:easy_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ad546f00bd725bea998a5fc4c6a870f43a1a4e7457bda42b110096b13a029fd8
+size 449

app/src/content/assets/finetasks/data/ar/exams_ara_cf:_average_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af146e0409fc3332f8f250a36caabb270e01ad48ad5d04dd539de86bdc8529ff
+size 36571

app/src/content/assets/finetasks/data/ar/exams_ara_cf:_average_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9dbd86d8c6c5922af78b295bbc8c89f483049a17500be20cb625565a2e599242
+size 1717

app/src/content/assets/finetasks/data/ar/frenchbench_arc_fra_cf:challenge_data.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2649f9802e39e19da555d2e42851281cca18826534d23246bffd8b15a43e326a
+size 14390

app/src/content/assets/finetasks/data/ar/frenchbench_arc_fra_cf:challenge_stats.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2da7b002e63a530df289c706c5c56113d16427105d18cff9dc556f25feb7e5e5
+size 477