@genai-fi/nanogpt 0.4.3 → 0.4.5

package/dist/ops/webgl/gatherSub.js

@@ -1,4 +1,4 @@
-import { r as l } from "../../index-C4JCoBvj.js";
+import { r as l } from "../../index--6vO-cOz.js";
 class u {
   variableNames = ["labels", "logits", "values"];
   outputShape;

package/dist/ops/webgl/gelu.js

@@ -1,5 +1,5 @@
-import { r as a } from "../../index-C4JCoBvj.js";
-import { u as s, C as x } from "../../kernel_funcs_utils-CAd1h9X1.js";
+import { r as a } from "../../index--6vO-cOz.js";
+import { u as s, C as x } from "../../kernel_funcs_utils-C6YBCuOt.js";
 const t = 0.7978845608028654, r = 0.044715, c = x + `
     float x3 = x * x * x;
     float inner = x + ${r} * x3;

package/dist/ops/webgl/matMulGelu.d.ts

@@ -0,0 +1,21 @@
+import { TensorInfo } from '@tensorflow/tfjs-core';
+import { MathBackendWebGL } from '@tensorflow/tfjs-backend-webgl';
+export declare const MATMUL_SHARED_DIM_THRESHOLD = 1000;
+type BatchMatMulConfig = {
+    a: TensorInfo;
+    b: TensorInfo;
+    transposeA: boolean;
+    transposeB: boolean;
+    backend: MathBackendWebGL;
+    activationSnippet?: string;
+    multiplier?: TensorInfo;
+};
+export declare function batchMatMulGeluImpl({ a, b, transposeA, transposeB, backend, activationSnippet, multiplier, }: BatchMatMulConfig): TensorInfo;
+export declare function batchMatMulKernel(args: {
+    inputs: {
+        x: TensorInfo;
+        kernel: TensorInfo;
+    };
+    backend: MathBackendWebGL;
+}): TensorInfo;
+export {};

package/dist/ops/webgl/matMulGelu.js

@@ -0,0 +1,168 @@
+import { r as C, t as R, e as I, n as G, O as L, j as F, Q as U } from "../../index--6vO-cOz.js";
+import { r as S } from "../../Reshape-CiAY8ltP.js";
+import { u as H } from "../../gpgpu_math-CUzjlO9A.js";
+import { m as B } from "../../mat_mul-BEHRPMh0.js";
+/**
+ * @license
+ * Copyright 2018 Google LLC. All Rights Reserved.
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ * =============================================================================
+ */
+class W {
+  constructor(e, s, n, a = !1, c = !1, o = !1, r = null, u = !1, l = !1) {
+    this.variableNames = ["matrixA", "matrixB"], this.packedInputs = !0, this.packedOutput = !0, this.outputShape = n, this.enableShapeUniforms = H(this.outputShape.length);
+    const h = a ? e[1] : e[2], p = Math.ceil(h / 2), d = a ? "i * 2, rc.y" : "rc.y, i * 2", $ = c ? "rc.z, i * 2" : "i * 2, rc.z", x = a ? ["a.xxyy", "a.zzww"] : ["a.xxzz", "a.yyww"], m = c ? ["b.xzxz", "b.ywyw"] : ["b.xyxy", "b.zwzw"];
+    let i = "", b = "";
+    r && (u ? i = `vec4 activation(vec4 a) {
+          vec4 b = getPreluActivationWeightsAtOutCoords();
+          ${r}
+        }` : l ? i = `vec4 activation(vec4 a) {
+          vec4 b = getLeakyreluAlphaAtOutCoords();
+          ${r}
+        }` : i = `vec4 activation(vec4 x) {
+          ${r}
+        }`, b = "result = activation(result);");
+    const M = o ? "result += getBiasAtOutCoords();" : "";
+    o && this.variableNames.push("bias"), u && this.variableNames.push("preluActivationWeights"), l && this.variableNames.push("leakyreluAlpha");
+    let f = "rc.x", v = "rc.x";
+    e[0] < s[0] ? f = `imod(rc.x, ${e[0]})` : s[0] < e[0] && (v = `imod(rc.x, ${s[0]})`), this.userCode = `
+      ${i}
+      // Don't use uniform for sharedDimensionPacked for performance.
+      const float sharedDimension = ${p}.0;
+
+      vec4 dot2x2ARowBCol(ivec3 rc) {
+        vec4 result = vec4(0);
+        int batchA = ${f};
+        int batchB = ${v};
+        for (int i = 0; i < ${p}; i++) {
+          vec4 a = getMatrixA(batchA, ${d});
+          vec4 b = getMatrixB(batchB, ${$});
+
+          // These swizzled products need to be separately added.
+          // See: https://github.com/tensorflow/tfjs/issues/1735
+          result += (${x[0]} * ${m[0]});
+          result += (${x[1]} * ${m[1]});
+        }
+        return result;
+      }
+
+      void main() {
+        ivec3 rc = getOutputCoords();
+        vec4 result = dot2x2ARowBCol(rc);
+
+        ${M}
+
+        ${b}
+
+        setOutput(result);
+      }
+    `;
+  }
+}
+const g = 0.7978845608028654, w = 0.044715, j = `
+    vec4 x3 = x * x * x;
+    vec4 inner = x + ${w} * x3;
+    inner = ${g} * inner;
+    inner = tanh(inner);
+    inner = 0.5 * (1.0 + inner);
+    vec4 result = x * inner;
+    return result;
+`, q = `
+    vec4 a2 = a * a;
+    vec4 a3 = a2 * a;
+    vec4 u  = ${g} * (a + ${w} * a3);
+    vec4 t  = tanh(u);
+    vec4 sech2 = 1.0 - t * t;
+    vec4 du_dx = ${g} * (1.0 + 3.0 * ${w} * a2);
+    vec4 dgelu = 0.5 * (1.0 + t) + 0.5 * a * sech2 * du_dx;
+    return dgelu * b;
+`, se = 1e3;
+function O({
+  a: t,
+  b: e,
+  transposeA: s,
+  transposeB: n,
+  backend: a,
+  activationSnippet: c,
+  multiplier: o
+}) {
+  const r = t.shape.length, u = e.shape.length, l = s ? t.shape[r - 2] : t.shape[r - 1], h = n ? e.shape[u - 1] : e.shape[u - 2], p = s ? t.shape[r - 1] : t.shape[r - 2], d = n ? e.shape[u - 2] : e.shape[u - 1], $ = t.shape.slice(0, -2), x = e.shape.slice(0, -2), m = G($), i = G(x), M = L(t.shape.slice(0, -2), e.shape.slice(0, -2)).concat([p, d]);
+  F(
+    l === h,
+    () => `Error in matMul: inner shapes (${l}) and (${h}) of Tensors with shapes ${t.shape} and ${e.shape} and transposeA=${s} and transposeB=${n} must match.`
+  );
+  const f = s ? [m, l, p] : [m, p, l], v = n ? [i, d, h] : [i, h, d], A = S({ inputs: { x: t }, backend: a, attrs: { shape: f } }), y = S({ inputs: { x: e }, backend: a, attrs: { shape: v } }), D = [A, y], E = Math.max(m, i), N = c, T = U(t.dtype, e.dtype), _ = new W(
+    f,
+    v,
+    [E, p, d],
+    s,
+    n,
+    !1,
+    N,
+    !!o,
+    !1
+  ), k = [A, y];
+  o && k.push(o);
+  const z = a.runWebGLProgram(_, k, T), K = S({ inputs: { x: z }, backend: a, attrs: { shape: M } });
+  D.push(z);
+  for (const P of D)
+    a.disposeIntermediateTensorInfo(P);
+  return K;
+}
+function Q(t) {
+  const { inputs: e, backend: s } = t, { x: n, kernel: a } = e;
+  if (n === void 0 || a === void 0)
+    throw new Error("BatchMatMul requires two input tensors.");
+  return O({
+    a: n,
+    b: a,
+    transposeA: !1,
+    transposeB: !1,
+    backend: s,
+    activationSnippet: j
+  });
+}
+const J = {
+  kernelName: "MatMulGelu",
+  backendName: "webgl",
+  kernelFunc: Q
+};
+C(J);
+function V(t) {
+  const { dy: e, x: s, kernel: n } = t.inputs, a = t.backend;
+  return R(() => {
+    const c = I().makeTensorFromTensorInfo(
+      O({
+        a: s,
+        b: n,
+        transposeA: !1,
+        transposeB: !1,
+        backend: a,
+        activationSnippet: q,
+        multiplier: e
+      })
+    ), o = B(c, n, !1, !0), r = B(s, c, !0, !1);
+    return [o, r];
+  });
+}
+const X = {
+  kernelName: "MatMulGeluGrad",
+  backendName: "webgl",
+  kernelFunc: V
+};
+C(X);
+export {
+  se as MATMUL_SHARED_DIM_THRESHOLD,
+  O as batchMatMulGeluImpl,
+  Q as batchMatMulKernel
+};

package/dist/ops/webgl/mulDropout.js

@@ -1,4 +1,4 @@
-import { r as m } from "../../index-C4JCoBvj.js";
+import { r as m } from "../../index--6vO-cOz.js";
 class f {
   variableNames = ["a", "b"];
   outputShape;

package/dist/ops/webgl/normRMS.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/ops/webgl/normRMS.js

@@ -0,0 +1,78 @@
+import { r as c, e as h } from "../../index--6vO-cOz.js";
+import { s as q } from "../../sum-DdkDf2MG.js";
+class G {
+  variableNames = ["x", "meanSquare", "gamma"];
+  outputShape;
+  userCode;
+  constructor(e, a, o) {
+    this.outputShape = [e, a, o], this.userCode = `
+        void main() {
+            ivec3 coords = getOutputCoords();
+            float x = getXAtOutCoords();
+            float meanSquare = getMeanSquare(coords.x, coords.y, 0);
+            float gamma = getGammaAtOutCoords();
+            float invRms = inversesqrt(meanSquare + 1e-8);
+            float normalized = x * invRms;
+            float outVal = normalized * gamma;
+            setOutput(outVal);
+        }
+        `;
+  }
+}
+function v(t) {
+  const { x: e, gamma: a } = t.inputs, o = t.backend, r = e.shape[0], s = e.shape[1], n = e.shape[2], m = e.square().mean(-1, !0), u = new G(r, s, n);
+  return o.runWebGLProgram(u, [e, m, a], "float32");
+}
+const x = {
+  kernelName: "RMSNorm",
+  backendName: "webgl",
+  kernelFunc: v
+};
+c(x);
+class y {
+  variableNames = ["x", "meanSquare", "dyGamma", "dyXMean"];
+  outputShape;
+  userCode;
+  constructor(e, a, o) {
+    this.outputShape = [e, a, o], this.userCode = `
+        void main() {
+            ivec3 coords = getOutputCoords();
+            float x = getXAtOutCoords();
+            float meanSquare = getMeanSquare(coords.x, coords.y, 0) + 1e-8;
+            float dyGamma = getDyGammaAtOutCoords();
+            float dyXMean = getDyXMean(coords.x, coords.y, 0) / ${o}.0;
+            float invRms = inversesqrt(meanSquare);
+            float dx = dyGamma * invRms - x * dyXMean * invRms / meanSquare;
+            setOutput(dx);
+        }
+        `;
+  }
+}
+class C {
+  variableNames = ["x", "meanSquare", "dy"];
+  outputShape;
+  userCode;
+  constructor(e, a, o) {
+    this.outputShape = [e, a, o], this.userCode = `
+        void main() {
+            ivec3 coords = getOutputCoords();
+            float x = getXAtOutCoords();
+            float meanSquare = getMeanSquare(coords.x, coords.y, 0) + 1e-8;
+            float dy = getDyAtOutCoords();
+            float invRms = inversesqrt(meanSquare);
+            float dGamma = dy * (x * invRms);
+            setOutput(dGamma);
+        }
+        `;
+  }
+}
+function b(t) {
+  const { dy: e, x: a, gamma: o } = t.inputs, r = t.backend, s = a.shape[0], n = a.shape[1], m = a.shape[2], u = a.square().mean(-1, !0), d = e.mul(o), l = d.mul(a).sum(-1, !0), i = new y(s, n, m), g = r.runWebGLProgram(i, [a, u, d, l], "float32"), p = new C(s, n, m), S = r.runWebGLProgram(p, [a, u, e], "float32"), f = q(h().makeTensorFromTensorInfo(S), [0, 1]);
+  return [g, f];
+}
+const N = {
+  kernelName: "RMSNormGrad",
+  backendName: "webgl",
+  kernelFunc: b
+};
+c(N);

package/dist/ops/webgl/qkv.js

@@ -1,4 +1,4 @@
-import { r as i } from "../../index-C4JCoBvj.js";
+import { r as i } from "../../index--6vO-cOz.js";
 class l {
   variableNames = ["x", "kernel"];
   outputShape;

package/dist/ops/webgl/rope.js

@@ -1,4 +1,4 @@
-import { r as u } from "../../index-C4JCoBvj.js";
+import { r as u } from "../../index--6vO-cOz.js";
 class l {
   variableNames = ["x", "sin", "cos"];
   outputShape;

package/dist/ops/webgl/scatterSub.js

@@ -1,4 +1,4 @@
-import { r as i } from "../../index-C4JCoBvj.js";
+import { r as i } from "../../index--6vO-cOz.js";
 class u {
   variableNames = ["labels", "softmaxProbs", "dy"];
   outputShape;

package/dist/{range-9AzeApCc.js → range-C_vpUjBu.js}

@@ -1,4 +1,4 @@
-import { E as e, R as f } from "./index-C4JCoBvj.js";
+import { E as e, R as f } from "./index--6vO-cOz.js";
 /**
  * @license
  * Copyright 2018 Google LLC. All Rights Reserved.

package/dist/{reshape-Boe4DuIO.js → reshape-z51Eu-re.js}

@@ -1,4 +1,4 @@
-import { o, h as t, E as a, q as p } from "./index-C4JCoBvj.js";
+import { o, h as t, E as a, q as p } from "./index--6vO-cOz.js";
 /**
  * @license
  * Copyright 2020 Google LLC. All Rights Reserved.

package/dist/{sin-KmhiDuMa.js → sin-H567uayl.js}

@@ -1,4 +1,4 @@
-import { o, h as t, E as c, Q as a, T as e } from "./index-C4JCoBvj.js";
+import { o, h as t, E as c, V as a, W as e } from "./index--6vO-cOz.js";
 /**
  * @license
  * Copyright 2018 Google LLC. All Rights Reserved.

package/dist/{slice_util-19zDNNSn.js → slice_util-BdhYwFY_.js}

@@ -1,5 +1,5 @@
-import { o as u, h as p, k as g, w as m, E as w, a1 as x, j as i } from "./index-C4JCoBvj.js";
-import { r as y } from "./reshape-Boe4DuIO.js";
+import { o as u, h as p, k as g, w as m, E as w, a4 as x, j as i } from "./index--6vO-cOz.js";
+import { r as y } from "./reshape-z51Eu-re.js";
 /**
  * @license
  * Copyright 2020 Google LLC. All Rights Reserved.

package/dist/{softmax-Cujsg4ay.js → softmax-Dsxflvdl.js}

@@ -1,4 +1,4 @@
-import { o as r, h as f, E as e, S as i } from "./index-C4JCoBvj.js";
+import { o as r, h as f, E as e, S as i } from "./index--6vO-cOz.js";
 /**
  * @license
  * Copyright 2018 Google LLC. All Rights Reserved.

package/dist/{split-DbcNm1-i.js → split-B_k_jwud.js}

@@ -1,4 +1,4 @@
-import { o as p, h as i, E as a, u as c } from "./index-C4JCoBvj.js";
+import { o as p, h as i, E as a, u as c } from "./index--6vO-cOz.js";
 /**
  * @license
  * Copyright 2020 Google LLC. All Rights Reserved.

package/dist/{stack-D1YjmgKN.js → stack-CmqSdsfs.js}

@@ -1,4 +1,4 @@
-import { o as e, i as c, j as n, E as i, P as k } from "./index-C4JCoBvj.js";
+import { o as e, i as c, j as n, E as i, P as k } from "./index--6vO-cOz.js";
 /**
  * @license
  * Copyright 2020 Google LLC. All Rights Reserved.

package/dist/{sum-R28pucR5.js → sum-DdkDf2MG.js}

@@ -1,4 +1,4 @@
-import { o as e, h as u, x as c, E as l, y as m } from "./index-C4JCoBvj.js";
+import { o as e, h as u, x as c, E as l, y as m } from "./index--6vO-cOz.js";
 /**
  * @license
  * Copyright 2018 Google LLC. All Rights Reserved.

package/dist/{tensor-BVeHdl7V.js → tensor-BGYi41cj.js}

@@ -1,4 +1,4 @@
-import { J as t, K as a } from "./index-C4JCoBvj.js";
+import { J as t, K as a } from "./index--6vO-cOz.js";
 /**
  * @license
  * Copyright 2018 Google LLC. All Rights Reserved.

package/dist/{tensor2d-DqFGNs_K.js → tensor2d-DUr_htjt.js}

@@ -1,4 +1,4 @@
-import { I as t, J as s, K as a } from "./index-C4JCoBvj.js";
+import { I as t, J as s, K as a } from "./index--6vO-cOz.js";
 /**
  * @license
  * Copyright 2018 Google LLC. All Rights Reserved.

package/dist/{tfjs_backend-Cug-PH75.js → tfjs_backend-DuKis_xG.js}

@@ -1,18 +1,18 @@
-import { o as f, h as l, E as d, an as nn, ao as tn, j as p, ap as sn, D as Ve, aq as rn, al as P, ar as on, as as an, at as cn, au as un, a1 as ln, aa as pn, p as Y, av as fn, aw as hn, ax as dn, ay as mn, az as gn, aA as $n, aB as bn, aC as kn, x as C, aD as xn, aE as wn, aF as An, aG as yn, aH as _n, aI as In, aJ as En, aK as Sn, I as Tn, J as Nn, K as Dn, aL as Mn, t as K, aM as he, b as I, aN as On, a5 as U, n as Me, c as Oe, aO as Be, aP as Bn, aQ as Pn, aR as Kn, aS as Rn, aT as qn, aU as jn, f as Fn, aV as Gn, aW as Ln, ad as L, s as q, aX as Vn, w as de, a as Cn, ai as Pe, aY as vn, $ as zn } from "./index-C4JCoBvj.js";
-import { r as $ } from "./reshape-Boe4DuIO.js";
-import { s as Ce } from "./split-DbcNm1-i.js";
-import { s as G } from "./sum-R28pucR5.js";
-import { b as me } from "./slice_util-19zDNNSn.js";
-import { r as ue } from "./range-9AzeApCc.js";
-import { t as Wn } from "./tensor-BVeHdl7V.js";
-import { s as le } from "./stack-D1YjmgKN.js";
-import { c as Jn, z as Yn } from "./zeros-dnQxFgAD.js";
-import { n as ve } from "./norm-CZM380I3.js";
-import { c as V } from "./concat-CuRsVY-K.js";
-import { m as y } from "./mat_mul-415y5Qn2.js";
-import { t as ge } from "./tensor2d-DqFGNs_K.js";
-import { r as Un, d as Xn } from "./dropout-DfDdklfL.js";
-import { g as Zn } from "./gather-ZYRWhmXR.js";
+import { o as f, h as l, E as d, ap as nn, aq as tn, j as p, ar as sn, D as Ve, as as rn, O as P, at as on, au as an, av as cn, aw as un, a4 as ln, ag as pn, p as Y, ax as fn, ay as hn, az as dn, aA as mn, aB as gn, aC as $n, aD as bn, aE as kn, x as C, aF as xn, aG as wn, aH as An, aI as yn, aJ as _n, aK as In, aL as En, aM as Sn, I as Tn, J as Nn, K as Dn, aN as Mn, t as K, aO as he, b as I, aP as On, a8 as U, n as Me, c as Oe, aQ as Be, aR as Bn, aS as Pn, aT as Kn, aU as Rn, aV as qn, aW as jn, f as Fn, aX as Gn, aY as Ln, ai as L, s as q, aZ as Vn, w as de, a as Cn, am as Pe, a_ as vn, a2 as zn } from "./index--6vO-cOz.js";
+import { r as $ } from "./reshape-z51Eu-re.js";
+import { s as Ce } from "./split-B_k_jwud.js";
+import { s as G } from "./sum-DdkDf2MG.js";
+import { b as me } from "./slice_util-BdhYwFY_.js";
+import { r as ue } from "./range-C_vpUjBu.js";
+import { t as Wn } from "./tensor-BGYi41cj.js";
+import { s as le } from "./stack-CmqSdsfs.js";
+import { c as Jn, z as Yn } from "./zeros-8xl-W2DC.js";
+import { n as ve } from "./norm-DSva3hI3.js";
+import { c as V } from "./concat-DvWM7HGZ.js";
+import { m as y } from "./mat_mul-BEHRPMh0.js";
+import { t as ge } from "./tensor2d-DUr_htjt.js";
+import { r as Un, d as Zn } from "./dropout-DFEXTPV0.js";
+import { g as Xn } from "./gather-C5D8PxwA.js";
 /**
  * @license
  * Copyright 2018 Google LLC. All Rights Reserved.
@@ -708,7 +708,7 @@ function Ut(e, n = 0) {
   const s = { x: l(e, "x", "step") }, r = { alpha: n };
   return d.runKernel(Sn, s, r);
 }
-const Xt = /* @__PURE__ */ f({ step_: Ut });
+const Zt = /* @__PURE__ */ f({ step_: Ut });
 /**
  * @license
  * Copyright 2018 Google LLC. All Rights Reserved.
@@ -748,13 +748,13 @@ function B(e, n) {
  * limitations under the License.
  * =============================================================================
  */
-function Zt(e, n = 0) {
+function Xt(e, n = 0) {
   const t = l(e, "x", "unstack", "string_or_numeric");
   p(n >= -t.shape.length && n < t.shape.length, () => `Axis = ${n} is not in [-${t.shape.length}, ${t.shape.length})`);
   const s = { value: t }, r = { axis: n };
   return d.runKernel(Mn, s, r);
 }
-const Ye = /* @__PURE__ */ f({ unstack_: Zt });
+const Ye = /* @__PURE__ */ f({ unstack_: Xt });
 /**
  * @license
  * Copyright 2018 Google LLC. All Rights Reserved.
@@ -804,7 +804,7 @@ function Qt(e, n, t) {
   if (t == null || t === "linear")
     return e;
   if (t === "relu")
-    return I(e, Xt(n));
+    return I(e, Zt(n));
   throw new Error(`Cannot compute gradient for fused activation ${t}.`);
 }
 function es(e, n) {
@@ -862,13 +862,13 @@ function ss({ a: e, b: n, transposeA: t = !1, transposeB: s = !1, bias: r, activ
   a != null && (te = l(a, "prelu weights", "fused matMul"));
   const v = (z, se) => {
     const [j, F, W, ce] = se, R = Qt($(z, W.shape), W, o);
-    let Z, H;
-    if (!t && !s ? (Z = y(R, F, !1, !0), H = y(j, R, !0, !1)) : !t && s ? (Z = y(R, F, !1, !1), H = y(R, j, !0, !1)) : t && !s ? (Z = y(F, R, !1, !0), H = y(j, R, !1, !1)) : (Z = y(F, R, !0, !0), H = y(R, j, !0, !0)), r != null) {
+    let X, H;
+    if (!t && !s ? (X = y(R, F, !1, !0), H = y(j, R, !0, !1)) : !t && s ? (X = y(R, F, !1, !1), H = y(R, j, !0, !1)) : t && !s ? (X = y(F, R, !1, !0), H = y(j, R, !1, !1)) : (X = y(F, R, !0, !0), H = y(R, j, !0, !0)), r != null) {
       const en = es(ce, R);
-      return [Z, H, en];
+      return [X, H, en];
     } else
-      return [Z, H];
-  }, X = {
+      return [X, H];
+  }, Z = {
     a: M,
     b: O,
     bias: N,
@@ -877,13 +877,13 @@ function ss({ a: e, b: n, transposeA: t = !1, transposeB: s = !1, bias: r, activ
   return r == null ? Oe((se, j, F) => {
     const W = (
       // tslint:disable-next-line: no-unnecessary-type-assertion
-      d.runKernel(Be, X, De)
+      d.runKernel(Be, Z, De)
     );
     return F([se, j, W]), { value: $(W, T), gradFunc: v };
   })(M, O) : Oe((se, j, F, W) => {
     const ce = (
       // tslint:disable-next-line: no-unnecessary-type-assertion
-      d.runKernel(Be, X, De)
+      d.runKernel(Be, Z, De)
     );
     return W([se, j, ce, F]), { value: $(ce, T), gradFunc: v };
   })(M, O, N);
@@ -1575,15 +1575,15 @@ function Fe(e, n = !1) {
         if (c === 0)
           o = q(M, y(O, y(N, M)));
         else {
-          const X = q(M, y(O, y(N, M)));
-          o = V([_(o, [0, 0], [c, s]), X], 0);
+          const Z = q(M, y(O, y(N, M)));
+          o = V([_(o, [0, 0], [c, s]), Z], 0);
         }
         const te = xe(O), v = _(r, [0, c], [t, r.shape[1] - c]);
         if (c === 0)
           r = q(v, y(y(v, i), te));
         else {
-          const X = q(v, y(y(v, i), te));
-          r = V([_(r, [0, 0], [t, c]), X], 1);
+          const Z = q(v, y(y(v, i), te));
+          r = V([_(r, [0, 0], [t, c]), Z], 1);
         }
         return [i, o, r];
       }), Cn([m, h, b]);
@@ -1638,7 +1638,7 @@ const Ar = {
  * https://opensource.org/licenses/MIT.
  * =============================================================================
  */
-const Xs = ["channelsFirst", "channelsLast"], Zs = ["nearest", "bilinear"], Hs = ["valid", "same", "causal"], Qs = ["max", "avg"], _r = ["sum", "mul", "concat", "ave"];
+const Zs = ["channelsFirst", "channelsLast"], Xs = ["nearest", "bilinear"], Hs = ["valid", "same", "causal"], Qs = ["max", "avg"], _r = ["sum", "mul", "concat", "ave"];
 /**
  * @license
  * Copyright 2018 Google LLC
@@ -1653,9 +1653,9 @@ class Ue extends Error {
     super(n), Object.setPrototypeOf(this, Ue.prototype);
   }
 }
-class Xe extends Error {
+class Ze extends Error {
   constructor(n) {
-    super(n), Object.setPrototypeOf(this, Xe.prototype);
+    super(n), Object.setPrototypeOf(this, Ze.prototype);
   }
 }
 class A extends Error {
@@ -1812,10 +1812,10 @@ function Rr(e, n, t = 0, s = 1 / 0) {
   return Ge(t >= 0), Ge(s >= t), Array.isArray(e) && e.length >= t && e.length <= s && e.every((r) => typeof r === n);
 }
 function nr(e, n) {
-  Array.isArray(e) ? (p(e.length > 0, () => `${n} is unexpectedly an empty array.`), e.forEach((t, s) => nr(t, `element ${s + 1} of ${n}`))) : p(Number.isInteger(e) && e > 0, () => `Expected ${n} to be a positive integer, but got ${Ze(e)}.`);
+  Array.isArray(e) ? (p(e.length > 0, () => `${n} is unexpectedly an empty array.`), e.forEach((t, s) => nr(t, `element ${s + 1} of ${n}`))) : p(Number.isInteger(e) && e > 0, () => `Expected ${n} to be a positive integer, but got ${Xe(e)}.`);
 }
-function Ze(e) {
-  return e === null ? "null" : Array.isArray(e) ? "[" + e.map((n) => Ze(n)).join(",") + "]" : typeof e == "string" ? `"${e}"` : `${e}`;
+function Xe(e) {
+  return e === null ? "null" : Array.isArray(e) ? "[" + e.map((n) => Xe(n)).join(",") + "]" : typeof e == "string" ? `"${e}"` : `${e}`;
 }
 function qr(e, n, t) {
   let s = t != null ? t() : Pe(), r;
@@ -1838,10 +1838,10 @@ function jr(e) {
  */
 const Q = /* @__PURE__ */ new Map();
 function tr(e) {
-  fe(Xs, "DataFormat", e);
+  fe(Zs, "DataFormat", e);
 }
 function Fr(e) {
-  fe(Zs, "InterpolationFormat", e);
+  fe(Xs, "InterpolationFormat", e);
 }
 function Gr(e) {
   fe(Hs, "PaddingMode", e);
@@ -1954,14 +1954,14 @@ function Ae() {
  * https://opensource.org/licenses/MIT.
  * =============================================================================
  */
-function Xr(e, n) {
+function Zr(e, n) {
   return C(e, n);
 }
 function or(e, n = -1) {
   const t = e.shape.slice();
   return n < 0 && (n = t.length + n + 1), t.splice(n, 0, 1), $(e, t);
 }
-function Zr(e, n) {
+function Xr(e, n) {
   return K(() => {
     if (e.shape.length !== 2)
       throw new A(`repeat() expects a rank-2 tensor, but received a rank-${e.shape.length} tensor.`);
@@ -2131,7 +2131,7 @@ function ro(e, n, t, s) {
   }
 }
 function oo(e, n, t) {
-  return K(() => (Array.isArray(n) ? n = B(n, "int32") : n = C(n, "int32"), Zn(e, n, t)));
+  return K(() => (Array.isArray(n) ? n = B(n, "int32") : n = C(n, "int32"), Xn(e, n, t)));
 }
 function ao(e) {
   return I(e, e);
@@ -2171,7 +2171,7 @@ function uo(e) {
   return K(() => L(e, U(zn(e), 1)));
 }
 function lo(e, n, t, s) {
-  return K(() => Xn(e, n, t, s));
+  return K(() => Zn(e, n, t, s));
 }
 function po(e) {
   return K(() => {
@@ -2200,8 +2200,8 @@ export {
   ie as N,
   ao as O,
   Ge as P,
-  Xr as Q,
-  Xe as R,
+  Zr as Q,
+  Ze as R,
   Yt as S,
   Dr as T,
   Pr as U,
@@ -2240,7 +2240,7 @@ export {
   lo as ap,
   no as aq,
   Qr as ar,
-  Zr as as,
+  Xr as as,
   Lr as at,
   _r as au,
   es as b,
@@ -2258,7 +2258,7 @@ export {
   Sr as n,
   Tr as o,
   Vr as p,
-  Xt as q,
+  Zt as q,
   so as r,
   ts as s,
   Nr as t,

package/dist/training/AdamExt.js

@@ -1,4 +1,4 @@
-import { A as r, b as c, f as h, s as g, e as o } from "../index-C4JCoBvj.js";
+import { A as r, b as c, f as h, s as g, e as o } from "../index--6vO-cOz.js";
 class u extends r {
   constructor(t, e, s, a, i) {
     super(t, e, s, a), this.config = i, this.startLearningRate = t;