English by default with PL switch

README.md

@@ -1,5 +1,5 @@
 ---
-title: Misja – Slayer 1B
+title: Mission – Slayer 1B
 emoji: 📈
 colorFrom: green
 colorTo: blue
@@ -9,8 +9,9 @@ license: apache-2.0
 short_description: Polish 1B LM trained from scratch, version by version
 ---
 
-# Misja – Slayer 1B
+# Mission – Slayer 1B (Misja – Slayer 1B)
 
 Arkadiusz Słota – Fabryka AI (SlayerLab). Polish ~1B-parameter base language model trained from scratch and continued
-version by version; versions, checkpoints, evaluation history and links. Numbers on the page come from the model cards
+version by version; versions, checkpoints, evaluation history and links. English by default, Polish via the PL switch
+(or `?lang=pl`). Numbers on the page come from the model cards
 and `scores-history.csv` (measured values only).

index.html

@@ -1,9 +1,9 @@
 <!DOCTYPE html>
-<html lang="pl">
+<html lang="en">
 <head>
 <meta charset="utf-8">
 <meta name="viewport" content="width=device-width, initial-scale=1">
-<title>Misja – Slayer 1B</title>
+<title>Mission – Slayer 1B</title>
 <style>
   :root { --bg:#0f1115; --card:#171a21; --line:#2a2f3a; --fg:#e8eaee; --muted:#9aa3b2; --acc:#4fb477; --acc2:#e0565b; --acc3:#4c8bd6; }
   * { box-sizing:border-box; }
@@ -38,10 +38,111 @@
   .wrap { overflow-x:auto; }
   footer { margin-top:56px; color:var(--muted); font-size:.9rem; border-top:1px solid var(--line); padding-top:16px; }
   a { color:var(--acc3); }
+  .lang { position:fixed; top:14px; right:16px; display:flex; gap:4px; background:var(--card); border:1px solid var(--line); border-radius:999px; padding:3px; z-index:10; }
+  .lang button { background:none; border:0; color:var(--muted); font:600 .85rem system-ui,sans-serif; padding:4px 10px; border-radius:999px; cursor:pointer; }
+  .lang button[aria-pressed="true"] { background:var(--line); color:var(--fg); }
+  body[data-lang="en"] .l-pl, body[data-lang="pl"] .l-en { display:none; }
 </style>
 </head>
-<body>
+<body data-lang="en">
+<nav class="lang" aria-label="Language"><button type="button" data-set="en" aria-pressed="true">EN</button><button type="button" data-set="pl" aria-pressed="false">PL</button></nav>
 <main>
+<div class="l-en" lang="en">
+<header>
+  <h1>Mission – Slayer 1B</h1>
+  <p class="by">Arkadiusz Słota – Fabryka AI (SlayerLab)</p>
+  <p class="lead">We are building a Polish language model of ~1 billion parameters from scratch and improving it version by
+  version — each version continues training from a checkpoint of the previous one taken before its final learning-rate
+  decay, on new data. All versions, checkpoints and results are public.</p>
+  <div class="pills">
+    <span class="pill"><b>1,073,912,851</b> parameters</span>
+    <span class="pill">tokenizer <b>32,768</b>, Polish</span>
+    <span class="pill">context <b>4096</b> tokens</span>
+    <span class="pill">data: ~82% Polish, ~18% English</span>
+    <span class="pill">weights license <b>Apache-2.0</b></span>
+  </div>
+</header>
+
+<section>
+  <h2>Versions</h2>
+  <div class="cards">
+    <div class="card">
+      <h3>Slayer1B 1.00 <span class="tag dec">decayed</span></h3>
+      <div class="meta">1.5 B training tokens</div>
+      <a href="https://huggingface.co/SlayerLab/Slayer1B-1.00">model</a> ·
+      <a href="https://huggingface.co/SlayerLab/Slayer1B-1.00-checkpoints">checkpoints</a> ·
+      <a href="https://track.fabryka.ai/run/48744ed2-3155-4294-b899-a6fda1f851da">run</a>
+    </div>
+    <div class="card">
+      <h3>Slayer1B 1.01 <span class="tag dec">decayed</span></h3>
+      <div class="meta">3.0 B tokens · continued from 1.00</div>
+      <a href="https://huggingface.co/SlayerLab/Slayer1B-1.01">model</a> ·
+      <a href="https://huggingface.co/SlayerLab/Slayer1B-1.01-checkpoints">checkpoints</a> ·
+      <a href="https://track.fabryka.ai/run/4934429a-d393-42e0-8aac-03eae6b71ff7">run</a>
+    </div>
+    <div class="card">
+      <h3>Slayer1B 1.02 <span class="tag const">constant LR</span></h3>
+      <div class="meta">6.0 B tokens · continued from 1.01</div>
+      <a href="https://huggingface.co/SlayerLab/Slayer1B-1.02">model</a> ·
+      <a href="https://huggingface.co/SlayerLab/Slayer1B-1.02-checkpoints">checkpoints</a> ·
+      <a href="https://track.fabryka.ai/run/21ae1234-82c2-42b7-887b-cbcc69901ae1">run</a>
+    </div>
+  </div>
+</section>
+
+<section>
+  <h2>Results</h2>
+  <div class="wrap">
+  <table>
+    <tr><th>test</th><th>1.00</th><th>1.01</th><th>1.02</th></tr>
+    <tr><td>training tokens</td><td>1.5 B</td><td>3.0 B</td><td>6.0 B</td></tr>
+    <tr><td>ARC-Easy (EN), %</td><td>42.89</td><td>47.98</td><td class="best">48.99</td></tr>
+    <tr><td>BLiMP (EN, grammar), %</td><td>71.51</td><td>73.63</td><td class="best">74.71</td></tr>
+    <tr><td>WikiText-2, byte perplexity ↓</td><td>2.5857</td><td>2.4230</td><td class="best">2.3735</td></tr>
+    <tr><td>MultiBLiMP (PL, grammar), %</td><td>97.89</td><td>98.29</td><td class="best">98.38</td></tr>
+    <tr><td>ARC-Easy-PL, %</td><td>36.91</td><td>40.36</td><td class="best">41.84</td></tr>
+  </table>
+  </div>
+  <p class="note">One instrument for all versions (leaderboard protocol of
+  <a href="https://track.fabryka.ai/models">track.fabryka.ai/models</a>, zero-shot). Item-by-item comparison 1.01 → 1.02:
+  BLiMP +1.08 pp (95% interval [+0.78; +1.37]); on the other tests the direction is positive but the difference is
+  unresolved. 1.00 and 1.01 ended with a learning-rate decay, 1.02 did not — so 1.02 vs 1.01 is not a clean comparison.</p>
+</section>
+
+<section>
+  <h2>How the scores grow during training</h2>
+  <figure><img src="scores-history.png" alt="Evaluation history of Slayer1B 1.00, 1.01 and 1.02 versus training tokens"></figure>
+  <p class="note">Measured values only, no fits: one checkpoint every 150 steps. Data: <a href="scores-history.csv">scores-history.csv</a>.</p>
+</section>
+
+<section>
+  <h2>How we do it</h2>
+  <ol class="steps">
+    <li><b>From scratch.</b> A transformer (20 layers, width 2048), our own Polish tokenizer, Polish data with an English
+      minority (<a href="https://huggingface.co/datasets/SlayerLab/slayer-pl-8x3b">slayer-pl-8x3b</a>).</li>
+    <li><b>Continued training instead of a new run.</b> Each version starts from a checkpoint of the previous one —
+      taken before its final learning-rate decay — and gets a new portion of data the model has not seen yet.</li>
+    <li><b>From 1.02 on, no decay along the way.</b> We keep training on the whole prepared corpus (~24 B tokens) and
+      apply the final decay once, at the end.</li>
+    <li><b>We verify everything.</b> Every checkpoint has a SHA-256 sum recorded off the training machine; files are
+      downloaded back and compared with it. From version 1.01 on, a loss forecast is written down before each run and
+      evaluated afterwards.</li>
+  </ol>
+  <p class="note">Technical report:
+  <a href="https://github.com/slayerlabs/Fabryka-AI-Notes/tree/main/papers/slayerlabs/2026-10-23_slayer1b-continuation">Slayer1B: Continuing a Polish-Dominant Base Model</a>.</p>
+</section>
+
+<section>
+  <h2>What's next</h2>
+  <p><b>Slayer1B 1.03</b> — another 1.5 B tokens (7.5 B in total), continued from 1.02 on our own 4×RTX 4090 cards.
+  Live progress on <a href="https://track.fabryka.ai/models">track.fabryka.ai</a>.</p>
+  <p class="note">These are base models: they predict the continuation of a text, they were not tuned for chat or for
+  following instructions, and they can be wrong.</p>
+</section>
+
+<footer>Arkadiusz Słota – Fabryka AI (SlayerLab) · models: <a href="https://huggingface.co/SlayerLab">huggingface.co/SlayerLab</a></footer>
+</div>
+<div class="l-pl" lang="pl">
 <header>
   <h1>Misja – Slayer 1B</h1>
   <p class="by">Arkadiusz Słota – Fabryka AI (SlayerLab)</p>
@@ -135,6 +236,23 @@
 </section>
 
 <footer>Arkadiusz Słota – Fabryka AI (SlayerLab) · modele: <a href="https://huggingface.co/SlayerLab">huggingface.co/SlayerLab</a></footer>
+</div>
 </main>
+<script>
+(function () {
+  var q = new URLSearchParams(location.search).get("lang");
+  var saved = null; try { saved = localStorage.getItem("slayer-lang"); } catch (e) {}
+  function set(l) {
+    if (l !== "pl") l = "en";
+    document.body.setAttribute("data-lang", l);
+    document.documentElement.lang = l;
+    document.title = l === "pl" ? "Misja – Slayer 1B" : "Mission – Slayer 1B";
+    document.querySelectorAll(".lang button").forEach(function (b) { b.setAttribute("aria-pressed", String(b.dataset.set === l)); });
+    try { localStorage.setItem("slayer-lang", l); } catch (e) {}
+  }
+  document.querySelectorAll(".lang button").forEach(function (b) { b.addEventListener("click", function () { set(b.dataset.set); }); });
+  set(q || saved || "en");
+})();
+</script>
 </body>
 </html>