@genai-fi/nanogpt 0.5.4 → 0.5.6

package/dist/Generator.js

@@ -1,5 +1,5 @@
 import { E as u } from "./index-Dwqa6Zy2.js";
-import "./index-CnHyhpKc.js";
+import "./index-Du-bmOP8.js";
 import "./ops/cpu/attentionMask.js";
 import "./ops/webgl/attentionMask.js";
 import "./ops/grads/attentionMask.js";
@@ -21,21 +21,21 @@ import "./ops/grads/matMulGelu.js";
 import "./ops/cpu/normRMS.js";
 import "./ops/webgl/normRMS.js";
 import "./ops/grads/normRMS.js";
-import "./random_width-DI2h9CMs.js";
+import "./random_width-C-v-35bY.js";
 import "./ops/cpu/gatherSub.js";
 import "./ops/webgl/gatherSub.js";
 import "./ops/cpu/scatterSub.js";
 import "./ops/webgl/scatterSub.js";
 import "./jszip.min-CjP2V1VV.js";
 import f from "./tokeniser/CharTokeniser.js";
-import "./dataset-ZHEPJmED.js";
+import "./dataset-iqT4Otvb.js";
 import "./index-Tf7vU29b.js";
 import "./papaparse.min-C8l2Kvo1.js";
 import "./ops/cpu/gelu.js";
 import "./ops/webgl/gelu.js";
 import "./ops/grads/gelu.js";
-import { t as d } from "./tensor2d-CqtBzOKq.js";
-import { c as g } from "./concat-BRRtq4S2.js";
+import { t as d } from "./tensor2d-CRWjDyUe.js";
+import { c as g } from "./concat-DdKPyAtw.js";
 const k = [
   ...Array.from({ length: 95 }, (a, t) => String.fromCharCode(t + 32)),
   // ASCII

package/dist/NanoGPTModel.d.ts

@@ -9,6 +9,8 @@ export interface TrainingLogEntry {
     time: number;
     example?: string;
     batchSize: number;
+    gradientNorm?: number;
+    learningRate?: number;
 }
 export interface GenerateOptions {
     temperature?: number;

package/dist/NanoGPTModel.js

@@ -1,18 +1,18 @@
 import { defaultConfig as F } from "./config.js";
 import O from "./layers/TransformerBlock.js";
-import { T as N, r as R } from "./TiedEmbedding-DORsPlNL.js";
+import { T as N, r as R } from "./TiedEmbedding-8S8xn8e6.js";
 import A from "./layers/RoPECache.js";
 import G from "./layers/RMSNorm.js";
 import { estimateParameterCount as j } from "./utilities/parameters.js";
 import { createSoftmaxCrossEntropyWithGrad as B } from "./training/sparseCrossEntropy.js";
 import V from "./layers/BaseLayer.js";
-import { E as H, D as W, p as J } from "./random_width-DI2h9CMs.js";
-import { o as x, j as y, u as Q, E as I, aa as U, ab as X, ac as Y, t as z, a9 as Z, f as L, H as tt } from "./index-CnHyhpKc.js";
-import { r as T } from "./reshape-CTIbqjwm.js";
-import { r as et } from "./range-CkOJ7090.js";
-import { s as q } from "./softmax-DX6qXAbm.js";
-import { t as ot } from "./ops-DzQTmLIl.js";
-import { g as st } from "./gather-BWyutxwi.js";
+import { E as H, D as W, p as J } from "./random_width-C-v-35bY.js";
+import { o as x, j as y, u as Q, E as I, a9 as U, aa as X, ab as Y, t as z, a8 as Z, f as L, H as tt } from "./index-Du-bmOP8.js";
+import { r as T } from "./reshape-BH7eBpwq.js";
+import { r as et } from "./range-Bvs1hidm.js";
+import { s as q } from "./softmax-DhWoBa7r.js";
+import { t as ot } from "./ops-CJNniCAV.js";
+import { g as st } from "./gather-D6MsdXqc.js";
 /**
  * @license
  * Copyright 2020 Google LLC. All Rights Reserved.

package/dist/{Reshape-Bt_t7RNz.js → Reshape-Biok_3X1.js}

@@ -1,5 +1,5 @@
-import { ad as $, ae as g, q as p, af as C, l as x } from "./index-CnHyhpKc.js";
-import { u as I } from "./gpgpu_math-Df7gzJWH.js";
+import { ac as f, ad as g, q as p, ae as C, l as x } from "./index-Du-bmOP8.js";
+import { u as I } from "./gpgpu_math-BFbOyvk4.js";
 /**
  * @license
  * Copyright 2018 Google LLC. All Rights Reserved.
@@ -17,7 +17,7 @@ import { u as I } from "./gpgpu_math-Df7gzJWH.js";
  * =============================================================================
  */
 function R(t, e, o = "index") {
-  const s = $(e);
+  const s = f(e);
   return s.map((n, r) => {
     const i = `int ${t[r]} = ${o} / ${n}`, u = r === s.length - 1 ? `int ${t[r + 1]} = ${o} - ${t[r]} * ${n}` : `index -= ${t[r]} * ${n}`;
     return `${i}; ${u};`;
@@ -38,7 +38,7 @@ function S(t, e, o = "index") {
   }).join("");
 }
 function F(t) {
-  const e = $(t).map((o) => o.toString());
+  const e = f(t).map((o) => o.toString());
   return `
   int getFlatIndex(ivec3 coords) {
     return coords.x * ${e[0]} + coords.y * ${e[1]} + coords.z;
@@ -82,7 +82,7 @@ function m(t) {
 function d(t) {
   return t % 2 === 0;
 }
-function f(t, e) {
+function $(t, e) {
   if (t = t.slice(-2), e = e.slice(-2), g(t, e) || !t.length || !e.length || t[0] === 0 || t[1] === 0 || e[0] === 0 || e[1] === 0)
     return !0;
   if (t.length !== e.length) {
@@ -205,7 +205,7 @@ function k(t) {
   const { inputs: e, backend: o, attrs: s } = t, { x: n } = e, { shape: r } = s, i = o, u = p(n.shape), a = C(r, u), c = p(a);
   x(u === c, () => `The new shape (${a}) has ${c} elements and the old shape (${n.shape}) has ${u} elements. The new shape and old shape must have the same number of elements.`);
   const l = i.texData.get(n.dataId);
-  return l.isPacked && !f(n.shape, a) && !(l.texture !== null && f(l.shape, a)) ? D(n, a, i) : (i.incRef(n.dataId), { dataId: n.dataId, shape: a, dtype: n.dtype });
+  return l.isPacked && !$(n.shape, a) && !(l.texture !== null && $(l.shape, a)) ? D(n, a, i) : (i.incRef(n.dataId), { dataId: n.dataId, shape: a, dtype: n.dtype });
 }
 export {
   k as r

package/dist/TeachableLLM.js

@@ -11,7 +11,7 @@ import g from "./tokeniser/bpe.js";
 import "./papaparse.min-C8l2Kvo1.js";
 import "./index-Tf7vU29b.js";
 import "./jszip.min-CjP2V1VV.js";
-import "./index-CnHyhpKc.js";
+import "./index-Du-bmOP8.js";
 import "./ops/cpu/scatterSub.js";
 import "./ops/webgl/scatterSub.js";
 import "./ops/cpu/gatherSub.js";
@@ -91,19 +91,19 @@ class a {
   }
   static loadModel(t) {
     const e = new a();
-    return f(t).then(({ model: i, tokeniser: o }) => {
-      e._model = i, e._tokeniser = o, e._config = i.config, e.setStatus("warmup"), m(i).then(() => {
+    return f(t).then(({ model: i, tokeniser: s }) => {
+      e._model = i, e._tokeniser = s, e._config = i.config, e.setStatus("warmup"), m(i).then(() => {
         e.setStatus("ready"), e.ee.emit("loaded");
-      }).catch((s) => {
-        e.setStatus("error"), e.ee.emit("error", s);
+      }).catch((o) => {
+        e.setStatus("error"), e.ee.emit("error", o);
       });
     }).catch((i) => {
       e.setStatus("error"), e.ee.emit("error", i);
     }), e;
   }
   static create(t, e = {}) {
-    const i = { ...h, ...e }, o = t === "char" ? new c(i.vocabSize) : new g(i.vocabSize), s = new l(i), r = new a(o, s);
-    return r.setStatus("warmup"), m(s).then(() => {
+    const i = { ...h, ...e }, s = t === "char" ? new c(i.vocabSize) : new g(i.vocabSize), o = new l(i), r = new a(s, o);
+    return r.setStatus("warmup"), m(o).then(() => {
       r.tokeniser.trained ? (r.setStatus("ready"), r.ee.emit("loaded")) : (r.setStatus("awaitingTokens"), r.ee.emit("loaded"), r.tokeniser.once("trainStatus", (n) => {
         n === "trained" && r.setStatus("ready");
       }));
@@ -134,10 +134,10 @@ class a {
     if (!this._model || !this._tokeniser)
       throw new Error("Model or tokeniser is not initialized.");
     const t = new p(this._model, this._tokeniser);
-    return t.on("start", () => this.setStatus("training")), t.on("stop", () => this.setStatus("ready")), t.on("log", async (e) => {
-      const i = this.ee.listeners("trainStep");
-      for (const o of i)
-        await o(e);
+    return t.on("start", () => this.setStatus("training")), t.on("stop", () => this.setStatus("ready")), t.on("log", async (e, i) => {
+      const s = this.ee.listeners("trainStep");
+      for (const o of s)
+        await o(e, i);
     }), t;
   }
   train(t, e) {

package/dist/{TiedEmbedding-DORsPlNL.js → TiedEmbedding-8S8xn8e6.js}

@@ -1,9 +1,9 @@
-import { R as a } from "./random_width-DI2h9CMs.js";
-import "./index-CnHyhpKc.js";
-import { d as s } from "./tfjs_backend-DX9yVvwk.js";
+import { R as a } from "./random_width-C-v-35bY.js";
+import "./index-Du-bmOP8.js";
+import { d as s } from "./tfjs_backend-D9Ytje0G.js";
 import o from "./layers/BaseLayer.js";
-import { v as m } from "./variable-BGvK-VN3.js";
-import { g as d } from "./gather-BWyutxwi.js";
+import { v as m } from "./variable-DZ3fF0R2.js";
+import { g as d } from "./gather-D6MsdXqc.js";
 /**
  * @license
  * Copyright 2018 Google LLC

package/dist/Trainer.d.ts

@@ -9,6 +9,7 @@ export interface ITrainerOptions {
     logInterval?: number;
     prompt?: string;
     validationSplit?: number;
+    advancedMetrics?: boolean;
 }
 export default class Trainer extends EE<'start' | 'stop' | 'log'> {
     private trainer;

package/dist/Trainer.js

@@ -1,10 +1,10 @@
-import { E as m } from "./index-Dwqa6Zy2.js";
-import d from "./training/FullTrainer.js";
-class S extends m {
+import { E as h } from "./index-Dwqa6Zy2.js";
+import m from "./training/FullTrainer.js";
+class p extends h {
   trainer;
   hasTrained = !1;
   constructor(e, t) {
-    super(), this.trainer = new d(e, t, 1e-3);
+    super(), this.trainer = new m(e, t, 1e-3);
   }
   stop() {
     this.trainer.stop();
@@ -25,10 +25,11 @@ class S extends m {
         logInterval: t?.logInterval || 10,
         desiredLoss: t?.desiredLoss || 0.01,
         maxSteps: t?.maxSteps || 1e3,
+        advancedMetrics: t?.advancedMetrics || !1,
         onStep: async (i, a) => {
           const l = this.listeners("log");
-          for (const h of l)
-            await h(i, {
+          for (const d of l)
+            await d(i, {
               ...a,
               progress: a.totalSamples / r,
               remaining: Math.max(
@@ -43,5 +44,5 @@ class S extends m {
   }
 }
 export {
-  S as default
+  p as default
 };

package/dist/{axis_util-CVbf1vmL.js → axis_util-BczFISHz.js}

@@ -1,4 +1,4 @@
-import { l as c } from "./index-CnHyhpKc.js";
+import { l as c } from "./index-Du-bmOP8.js";
 /**
  * @license
  * Copyright 2017 Google LLC. All Rights Reserved.

package/dist/{broadcast_to-BBoMQXbL.js → broadcast_to-B7NGsBSh.js}

@@ -1,5 +1,5 @@
-import { o as h, j as f, n as p, y as g, E as u, L as b } from "./index-CnHyhpKc.js";
-import { r as T } from "./reshape-CTIbqjwm.js";
+import { o as h, j as f, n as p, y as g, E as u, L as b } from "./index-Du-bmOP8.js";
+import { r as T } from "./reshape-BH7eBpwq.js";
 /**
  * @license
  * Copyright 2020 Google LLC. All Rights Reserved.

package/dist/{concat-BRRtq4S2.js → concat-DdKPyAtw.js}

@@ -1,4 +1,4 @@
-import { o as s, l as a, k as p, y as i, E as l, C as f } from "./index-CnHyhpKc.js";
+import { o as s, l as a, k as p, y as i, E as l, C as f } from "./index-Du-bmOP8.js";
 /**
  * @license
  * Copyright 2020 Google LLC. All Rights Reserved.

package/dist/{dataset-ZHEPJmED.js → dataset-iqT4Otvb.js}

@@ -1,7 +1,7 @@
-import { ag as S, T as h, N, a as v, ah as o, ai as p, aj as g, l as k, t as y } from "./index-CnHyhpKc.js";
+import { af as S, T as h, N, a as v, ag as o, ah as p, ai as g, l as k, t as y } from "./index-Du-bmOP8.js";
 import { s as R } from "./index-C4L8Cm77.js";
-import { s as $ } from "./stack-S2-D2JAQ.js";
-import { t as B } from "./tensor-IZex6Bwp.js";
+import { s as $ } from "./stack-BV1v7l3S.js";
+import { t as B } from "./tensor-DgTOPY6h.js";
 /**
  * @license
  * Copyright 2018 Google LLC. All Rights Reserved.
@@ -381,7 +381,7 @@ class i {
    *   throw an `Error`.
    */
   handleErrors(t) {
-    return new j(this, t);
+    return new Z(this, t);
   }
   // TODO(soergel): Implement reduce() etc.
   /**
@@ -708,7 +708,7 @@ class K extends i {
     return { value: r, done: !1 };
   }
 }
-class j extends i {
+class Z extends i {
   constructor(t, e) {
     super(), this.upstream = t, this.handler = e, this.count = 0, this.lastRead = Promise.resolve({ value: null, done: !1 });
   }
@@ -745,7 +745,7 @@ class w extends i {
     return { value: r, done: !1 };
   }
 }
-class Z extends i {
+class j extends i {
   constructor() {
     super(), this.outputQueue = new f(), this.lastRead = Promise.resolve({ value: null, done: !1 });
   }
@@ -759,7 +759,7 @@ class Z extends i {
     return { value: this.outputQueue.shift(), done: !1 };
   }
 }
-class tt extends Z {
+class tt extends j {
   constructor(t, e) {
     super(), this.upstream = t, this.transform = e;
   }

package/dist/{dropout-lQm_YyX3.js → dropout-B09InSJS.js}

@@ -1,4 +1,4 @@
-import { o as l, j as h, E as m, ak as p, n as c, al as d, ae as g, l as u, T as V, am as v, a9 as N, b as w } from "./index-CnHyhpKc.js";
+import { o as l, j as h, E as m, aj as p, n as c, ak as d, ad as g, l as u, T as V, al as v, a8 as N, b as w } from "./index-Du-bmOP8.js";
 import { s as f } from "./index-C4L8Cm77.js";
 /**
  * @license

package/dist/{gather-BWyutxwi.js → gather-D6MsdXqc.js}

@@ -1,4 +1,4 @@
-import { o as g, j as t, E as h, G as p } from "./index-CnHyhpKc.js";
+import { o as g, j as t, E as h, G as p } from "./index-Du-bmOP8.js";
 /**
  * @license
  * Copyright 2018 Google LLC. All Rights Reserved.

package/dist/{gpgpu_math-Df7gzJWH.js → gpgpu_math-BFbOyvk4.js}

@@ -1,4 +1,4 @@
-import { N as e } from "./index-CnHyhpKc.js";
+import { N as e } from "./index-Du-bmOP8.js";
 /**
  * @license
  * Copyright 2017 Google LLC. All Rights Reserved.

package/dist/{index-CnHyhpKc.js → index-Du-bmOP8.js}

@@ -4030,82 +4030,82 @@ export {
   ne as Z,
   dr as _,
   M as a,
-  xs as a$,
-  aa as a0,
+  $s as a$,
+  oa as a0,
   xe as a1,
-  V as a2,
-  oa as a3,
-  ns as a4,
-  nt as a5,
-  Ca as a6,
-  Fr as a7,
-  qr as a8,
-  S as a9,
-  _a as aA,
-  er as aB,
-  Pa as aC,
-  Ar as aD,
-  Rr as aE,
-  _r as aF,
-  Or as aG,
-  Gr as aH,
-  jr as aI,
-  Kr as aJ,
-  ha as aK,
-  Jr as aL,
-  ia as aM,
-  Ta as aN,
-  $a as aO,
-  Ds as aP,
-  no as aQ,
-  eo as aR,
-  yr as aS,
-  $r as aT,
-  ao as aU,
-  da as aV,
-  ma as aW,
-  ga as aX,
-  Na as aY,
-  va as aZ,
-  to as a_,
-  la as aa,
-  ua as ab,
-  Za as ac,
-  $t as ad,
-  Rt as ae,
-  Rs as af,
-  F as ag,
-  pe as ah,
-  fo as ai,
-  dt as aj,
-  xr as ak,
-  Wn as al,
-  x as am,
-  jt as an,
-  ue as ao,
-  za as ap,
-  rr as aq,
-  Br as ar,
-  Wr as as,
-  Sa as at,
-  Aa as au,
-  Ra as av,
-  ro as aw,
-  Io as ax,
-  oo as ay,
-  yo as az,
+  ns as a2,
+  nt as a3,
+  V as a4,
+  Ca as a5,
+  Fr as a6,
+  qr as a7,
+  S as a8,
+  la as a9,
+  er as aA,
+  Pa as aB,
+  Ar as aC,
+  Rr as aD,
+  _r as aE,
+  Or as aF,
+  Gr as aG,
+  jr as aH,
+  Kr as aI,
+  ha as aJ,
+  Jr as aK,
+  ia as aL,
+  Ta as aM,
+  $a as aN,
+  Ds as aO,
+  no as aP,
+  eo as aQ,
+  yr as aR,
+  $r as aS,
+  ao as aT,
+  da as aU,
+  ma as aV,
+  ga as aW,
+  Na as aX,
+  va as aY,
+  to as aZ,
+  xs as a_,
+  ua as aa,
+  Za as ab,
+  $t as ac,
+  Rt as ad,
+  Rs as ae,
+  F as af,
+  pe as ag,
+  fo as ah,
+  dt as ai,
+  xr as aj,
+  Wn as ak,
+  x as al,
+  jt as am,
+  ue as an,
+  za as ao,
+  rr as ap,
+  Br as aq,
+  Wr as ar,
+  Sa as as,
+  Aa as at,
+  Ra as au,
+  ro as av,
+  Io as aw,
+  oo as ax,
+  yo as ay,
+  _a as az,
   b,
   Vs as b$,
-  $s as b0,
-  ko as b1,
-  Ps as b2,
-  Cs as b3,
-  Lt as b4,
-  te as b5,
-  uo as b6,
-  dn as b7,
-  Re as b8,
-  $e as b9,
+  ko as b0,
+  Ps as b1,
+  Cs as b2,
+  Lt as b3,
+  te as b4,
+  uo as b5,
+  dn as b6,
+  Re as b7,
+  $e as b8,
+  ho as b9,
   ya as bA,
   pa as bB,
   wa as bC,
@@ -4133,31 +4133,31 @@ export {
   _s as bY,
   ae as bZ,
   Os as b_,
-  ho as ba,
-  Gs as bb,
-  Us as bc,
-  zs as bd,
-  Js as be,
-  Ys as bf,
-  Ya as bg,
-  tr as bh,
-  Cr as bi,
-  ir as bj,
-  lr as bk,
-  ur as bl,
-  fr as bm,
-  gr as bn,
-  mr as bo,
-  pr as bp,
-  br as bq,
-  wr as br,
-  Nr as bs,
-  Mr as bt,
-  Hr as bu,
-  Va as bv,
-  Xr as bw,
-  ea as bx,
-  sa as by,
+  Gs as ba,
+  Us as bb,
+  zs as bc,
+  Js as bd,
+  Ys as be,
+  Ya as bf,
+  tr as bg,
+  Cr as bh,
+  ir as bi,
+  lr as bj,
+  ur as bk,
+  fr as bl,
+  gr as bm,
+  mr as bn,
+  pr as bo,
+  br as bp,
+  wr as bq,
+  Nr as br,
+  Mr as bs,
+  Hr as bt,
+  Va as bu,
+  Xr as bv,
+  ea as bw,
+  sa as bx,
+  aa as by,
   fa as bz,
   Eo as c,
   Ws as c0,