|
|
<!DOCTYPE html> |
|
|
<html> |
|
|
|
|
|
<head> |
|
|
<title>Language Bench</title> |
|
|
<style> |
|
|
body { |
|
|
max-width: 1200px; |
|
|
margin: 0 auto; |
|
|
padding: 20px; |
|
|
font-family: sans-serif; |
|
|
} |
|
|
|
|
|
.language-header { |
|
|
margin-bottom: 10px; |
|
|
} |
|
|
|
|
|
.speaker-count { |
|
|
font-size: 0.8em; |
|
|
color: #666; |
|
|
font-weight: normal; |
|
|
margin: 0; |
|
|
} |
|
|
</style> |
|
|
<link rel="icon" |
|
|
href="data:image/svg+xml,<svg xmlns=%22http://www.w3.org/2000/svg%22 viewBox=%220 0 100 100%22><text y=%22.9em%22 font-size=%2290%22 fill=%22black%22>🌍</text></svg>"> |
|
|
</head> |
|
|
|
|
|
<body> |
|
|
<h1>Language Bench</h1> |
|
|
<div id="charts"></div> |
|
|
|
|
|
<script type="module"> |
|
|
|
|
|
import * as Plot from "https://cdn.jsdelivr.net/npm/@observablehq/[email protected]/+esm"; |
|
|
|
|
|
async function init() { |
|
|
const scoreKey = "bleu" |
|
|
const scoreName = "BLEU Score" |
|
|
const chartsDiv = document.getElementById('charts'); |
|
|
|
|
|
const response = await fetch('results.json'); |
|
|
const data = await response.json(); |
|
|
|
|
|
const formatScore = (score) => score > 0 ? score.toFixed(2) : "No benchmark available!" |
|
|
const formatTitle = d => (d.language_name + "\n" + parseInt(d.speakers / 1_000_00) / 10 + "M speakers\n" + scoreName + ": " + formatScore(d[scoreKey])) |
|
|
|
|
|
|
|
|
const summaryPlot = Plot.plot({ |
|
|
width: chartsDiv.clientWidth, |
|
|
height: 400, |
|
|
marginBottom: 100, |
|
|
x: { label: "Number of speakers", axis: null }, |
|
|
y: { label: `${scoreName} (average across models)` }, |
|
|
|
|
|
marks: [ |
|
|
Plot.rectY(data, Plot.stackX({ |
|
|
x: "speakers", |
|
|
order: scoreKey, |
|
|
reverse: true, |
|
|
y2: scoreKey, |
|
|
title: formatTitle, |
|
|
tip: true, |
|
|
fill: d => d[scoreKey] > 0 ? "black" : "pink" |
|
|
})), |
|
|
Plot.rectY(data, Plot.pointerX(Plot.stackX({ |
|
|
x: "speakers", |
|
|
order: scoreKey, |
|
|
reverse: true, |
|
|
y2: scoreKey, |
|
|
fill: "grey", |
|
|
}))), |
|
|
Plot.text(data, Plot.stackX({ |
|
|
x: "speakers", |
|
|
y2: scoreKey, |
|
|
order: scoreKey, |
|
|
reverse: true, |
|
|
text: "language_name", |
|
|
frameAnchor: "bottom", |
|
|
textAnchor: "end", |
|
|
dy: 10, |
|
|
rotate: 270, |
|
|
opacity: (d) => d.speakers > 50_000_000 ? 1 : 0, |
|
|
})) |
|
|
] |
|
|
}); |
|
|
|
|
|
|
|
|
chartsDiv.insertBefore(summaryPlot, chartsDiv.firstChild); |
|
|
|
|
|
|
|
|
const languageMap = new Map(); |
|
|
data.forEach(r => { |
|
|
if (!languageMap.has(r.language_name)) { |
|
|
languageMap.set(r.language_name, r.speakers); |
|
|
} |
|
|
}); |
|
|
|
|
|
|
|
|
const languages = [...languageMap.entries()] |
|
|
.sort((a, b) => b[1] - a[1]) |
|
|
.map(([lang]) => lang); |
|
|
|
|
|
|
|
|
|
|
|
languages.forEach(language => { |
|
|
const headerDiv = document.createElement('div'); |
|
|
headerDiv.className = 'language-header'; |
|
|
|
|
|
const h2 = document.createElement('h2'); |
|
|
h2.textContent = language; |
|
|
h2.style.marginBottom = '5px'; |
|
|
|
|
|
const speakerP = document.createElement('p'); |
|
|
speakerP.className = 'speaker-count'; |
|
|
const speakerCount = (languageMap.get(language) / 1_000_000).toFixed(1); |
|
|
speakerP.textContent = `${speakerCount}M speakers`; |
|
|
|
|
|
headerDiv.appendChild(h2); |
|
|
headerDiv.appendChild(speakerP); |
|
|
chartsDiv.appendChild(headerDiv); |
|
|
|
|
|
const languageData = data.filter(r => r.language_name === language)[0]["scores"]; |
|
|
|
|
|
const descriptor = code => { |
|
|
let [org, model] = code.split("/") |
|
|
return model.split("-")[0] |
|
|
} |
|
|
|
|
|
|
|
|
if (languageData && languageData.length > 1) { |
|
|
const plot = Plot.plot({ |
|
|
width: 400, |
|
|
height: 200, |
|
|
margin: 30, |
|
|
y: { |
|
|
domain: [0, 1], |
|
|
label: scoreName |
|
|
}, |
|
|
marks: [ |
|
|
Plot.barY(languageData, { |
|
|
x: d => descriptor(d.model), |
|
|
y: scoreKey |
|
|
}) |
|
|
] |
|
|
}); |
|
|
chartsDiv.appendChild(plot); |
|
|
} |
|
|
|
|
|
|
|
|
}); |
|
|
} |
|
|
|
|
|
init(); |
|
|
</script> |
|
|
</body> |
|
|
|
|
|
</html> |
