@genai-fi/nanogpt 0.10.2 → 0.11.0

package/dist/{kernel_funcs_utils-CWfOAPGO.js → kernel_funcs_utils-CSaumNDs.js}

@@ -1,15 +1,94 @@
-import { j as B } from "./index-D6Q1lPZO.js";
-import { u as N, f as G } from "./gpgpu_math-DsCcikas.js";
-import { s as K, H as W, e as O } from "./tensor-CzmOBsdf.js";
-import { I as z, h as V, W as F, X as H, u as S } from "./tensor_util-DfwaWayG.js";
-import { f as v } from "./backend_util-B1XRLuq9.js";
-const he = `
+import { $ as B, V as K, aQ as G, a8 as W, aJ as z, aR as V, ab as N, aK as F, ak as S } from "./index-D0RBWjq8.js";
+import { u as O, f as Y } from "./gpgpu_math-CJcbnKPC.js";
+import { f as v } from "./backend_util-C7NWHpv7.js";
+function H(t, e) {
+  return ["x", "y", "z", "w", "u", "v"].slice(0, e).map((s) => `${t}.${s}`);
+}
+function Z(t, e) {
+  return e === 1 ? [t] : H(t, e);
+}
+function pe(t, e) {
+  if (t === 1)
+    return "rc";
+  let s = "";
+  for (let r = 0; r < t; r++)
+    s += e[r], r < t - 1 && (s += ",");
+  return s;
+}
+class q {
+  constructor(e, s) {
+    this.variableNames = ["A"], this.outputShape = e, this.enableShapeUniforms = O(this.outputShape.length), this.userCode = `
+      float unaryOperation(float x) {
+        ${s}
+      }
+
+      void main() {
+        float x = getAAtOutCoords();
+        float y = unaryOperation(x);
+
+        setOutput(y);
+      }
+    `;
+  }
+}
+const T = "if (isnan(x)) return x;", M = "return x;", de = "return abs(x);", j = "return (x >= 0.0) ? x : (exp(x) - 1.0);", J = T + `
+  return (x < 0.0) ? 0.0 : x;
+`, Q = T + `
+  return (x < 0.0) ? 0.0 : min(6.0, x);
+`, he = "return x;", X = "return 1.0 / (1.0 + exp(-1.0 * x));";
+const ee = "return x;", te = `
+  vec4 result;
+
+  result.r = (x.r >= 0.0) ? x.r : (exp(x.r) - 1.0);
+  result.g = (x.g >= 0.0) ? x.g : (exp(x.g) - 1.0);
+  result.b = (x.b >= 0.0) ? x.b : (exp(x.b) - 1.0);
+  result.a = (x.a >= 0.0) ? x.a : (exp(x.a) - 1.0);
+
+  return result;
+`, se = `
+  vec4 result = x * vec4(greaterThanEqual(x, vec4(0.0)));
+  bvec4 isNaN = isnan(x);
+
+  result.r = isNaN.r ? x.r : result.r;
+  result.g = isNaN.g ? x.g : result.g;
+  result.b = isNaN.b ? x.b : result.b;
+  result.a = isNaN.a ? x.a : result.a;
+
+  return result;
+`, ae = `
+  vec4 result = min(x, vec4(6.)) * vec4(greaterThanEqual(x, vec4(0.0)));
+  bvec4 isNaN = isnan(x);
+
+  result.r = isNaN.r ? x.r : result.r;
+  result.g = isNaN.g ? x.g : result.g;
+  result.b = isNaN.b ? x.b : result.b;
+  result.a = isNaN.a ? x.a : result.a;
+
+  return result;
+`, re = "return 1.0 / (1.0 + exp(-1.0 * x));";
+class ne {
+  constructor(e, s) {
+    this.variableNames = ["A"], this.packedInputs = !0, this.packedOutput = !0, this.outputShape = e, this.enableShapeUniforms = O(this.outputShape.length), this.userCode = `
+      vec4 unaryOperation(vec4 x) {
+        ${s}
+      }
+
+      void main() {
+        vec4 x = getAAtOutCoords();
+        vec4 y = unaryOperation(x);
+
+        setOutput(y);
+      }
+    `;
+  }
+}
+const fe = `
   if (isnan(a)) return a;
   if (isnan(b)) return b;
 `;
 class b {
   constructor(e, s, r) {
-    this.variableNames = ["A", "B"], this.outputShape = B(s, r), this.enableShapeUniforms = N(this.outputShape.length), this.userCode = `
+    this.variableNames = ["A", "B"], this.outputShape = B(s, r), this.enableShapeUniforms = O(this.outputShape.length), this.userCode = `
       float binaryOperation(float a, float b) {
         ${e}
       }
@@ -22,20 +101,6 @@ class b {
     `;
   }
 }
-function Y(t, e) {
-  return ["x", "y", "z", "w", "u", "v"].slice(0, e).map((s) => `${t}.${s}`);
-}
-function Z(t, e) {
-  return e === 1 ? [t] : Y(t, e);
-}
-function fe(t, e) {
-  if (t === 1)
-    return "rc";
-  let s = "";
-  for (let r = 0; r < t; r++)
-    s += e[r], r < t - 1 && (s += ",");
-  return s;
-}
 const xe = `
   result.r = isNaN.r ? NAN : result.r;
   result.g = isNaN.g ? NAN : result.g;
@@ -46,7 +111,7 @@ class E {
   constructor(e, s, r, u = !1) {
     this.variableNames = ["A", "B"], this.supportsBroadcasting = !0, this.packedInputs = !0, this.packedOutput = !0, this.outputShape = B(s, r);
     const n = this.outputShape.length;
-    this.enableShapeUniforms = N(n);
+    this.enableShapeUniforms = O(n);
     let o = "";
     if (u)
       if (n === 0 || K(this.outputShape) === 1)
@@ -56,7 +121,7 @@ class E {
           result.w = 0.;
         `;
       else if (o = `
-          ${G(n)} coords = getOutputCoords();
+          ${Y(n)} coords = getOutputCoords();
         `, n === 1)
         this.enableShapeUniforms ? o += `
             result.y = (coords + 1) >= outShape ? 0. : result.y;
@@ -109,126 +174,59 @@ function P(t) {
   return s.incRef(r.dataId), { dataId: r.dataId, shape: r.shape, dtype: r.dtype };
 }
 const ge = {
-  kernelName: z,
+  kernelName: G,
   backendName: "webgl",
   kernelFunc: P
 };
-function T(t) {
+function L(t) {
   const { inputs: e, backend: s } = t, { real: r, imag: u } = e, n = s.makeTensorInfo(r.shape, "complex64"), o = s.texData.get(n.dataId), i = P({ inputs: { x: r }, backend: s }), a = P({ inputs: { x: u }, backend: s });
   return o.complexTensorInfos = { real: i, imag: a }, n;
 }
 const me = {
-  kernelName: V,
+  kernelName: W,
   backendName: "webgl",
-  kernelFunc: T
+  kernelFunc: L
 };
-const L = "return (a < 0.) ? b * a : a;", w = `
+const w = "return (a < 0.) ? b * a : a;", R = `
   vec4 aLessThanZero = vec4(lessThan(a, vec4(0.)));
   return (aLessThanZero * (b * a)) + ((vec4(1.0) - aLessThanZero) * a);
 `;
-function q(t) {
-  const { inputs: e, backend: s, attrs: r } = t, { x: u } = e, { alpha: n } = r, o = s.makeTensorInfo([], "float32", W(n, "float32")), i = O().getBool("WEBGL_PACK_BINARY_OPERATIONS") ? new E(w, u.shape, o.shape) : new b(L, u.shape, o.shape), a = s.runWebGLProgram(i, [u, o], "float32");
+function oe(t) {
+  const { inputs: e, backend: s, attrs: r } = t, { x: u } = e, { alpha: n } = r, o = s.makeTensorInfo([], "float32", V(n, "float32")), i = N().getBool("WEBGL_PACK_BINARY_OPERATIONS") ? new E(R, u.shape, o.shape) : new b(w, u.shape, o.shape), a = s.runWebGLProgram(i, [u, o], "float32");
   return s.disposeIntermediateTensorInfo(o), a;
 }
 const be = {
-  kernelName: F,
+  kernelName: z,
   backendName: "webgl",
-  kernelFunc: q
+  kernelFunc: oe
 };
-const R = "return (a < 0.) ? b * a : a;", k = `
+const k = "return (a < 0.) ? b * a : a;", $ = `
   vec4 aLessThanZero = vec4(lessThan(a, vec4(0.)));
   return (aLessThanZero * (b * a)) + ((vec4(1.0) - aLessThanZero) * a);
 `;
-function j(t) {
-  const { inputs: e, backend: s } = t, { x: r, alpha: u } = e, n = O().getBool("WEBGL_PACK_BINARY_OPERATIONS") ? new E(k, r.shape, u.shape) : new b(R, r.shape, u.shape);
+function ue(t) {
+  const { inputs: e, backend: s } = t, { x: r, alpha: u } = e, n = N().getBool("WEBGL_PACK_BINARY_OPERATIONS") ? new E($, r.shape, u.shape) : new b(k, r.shape, u.shape);
   return s.runWebGLProgram(n, [r, u], "float32");
 }
 const Ne = {
-  kernelName: H,
+  kernelName: F,
   backendName: "webgl",
-  kernelFunc: j
+  kernelFunc: ue
 };
-class M {
-  constructor(e, s) {
-    this.variableNames = ["A"], this.outputShape = e, this.enableShapeUniforms = N(this.outputShape.length), this.userCode = `
-      float unaryOperation(float x) {
-        ${s}
-      }
-
-      void main() {
-        float x = getAAtOutCoords();
-        float y = unaryOperation(x);
-
-        setOutput(y);
-      }
-    `;
-  }
-}
-const U = "if (isnan(x)) return x;", X = "return x;", Oe = "return abs(x);", J = "return (x >= 0.0) ? x : (exp(x) - 1.0);", Q = U + `
-  return (x < 0.0) ? 0.0 : x;
-`, ee = U + `
-  return (x < 0.0) ? 0.0 : min(6.0, x);
-`, ye = "return x;", te = "return 1.0 / (1.0 + exp(-1.0 * x));";
-const se = "return x;", ae = `
-  vec4 result;
-
-  result.r = (x.r >= 0.0) ? x.r : (exp(x.r) - 1.0);
-  result.g = (x.g >= 0.0) ? x.g : (exp(x.g) - 1.0);
-  result.b = (x.b >= 0.0) ? x.b : (exp(x.b) - 1.0);
-  result.a = (x.a >= 0.0) ? x.a : (exp(x.a) - 1.0);
-
-  return result;
-`, re = `
-  vec4 result = x * vec4(greaterThanEqual(x, vec4(0.0)));
-  bvec4 isNaN = isnan(x);
-
-  result.r = isNaN.r ? x.r : result.r;
-  result.g = isNaN.g ? x.g : result.g;
-  result.b = isNaN.b ? x.b : result.b;
-  result.a = isNaN.a ? x.a : result.a;
-
-  return result;
-`, ne = `
-  vec4 result = min(x, vec4(6.)) * vec4(greaterThanEqual(x, vec4(0.0)));
-  bvec4 isNaN = isnan(x);
-
-  result.r = isNaN.r ? x.r : result.r;
-  result.g = isNaN.g ? x.g : result.g;
-  result.b = isNaN.b ? x.b : result.b;
-  result.a = isNaN.a ? x.a : result.a;
-
-  return result;
-`, oe = "return 1.0 / (1.0 + exp(-1.0 * x));";
-class ue {
-  constructor(e, s) {
-    this.variableNames = ["A"], this.packedInputs = !0, this.packedOutput = !0, this.outputShape = e, this.enableShapeUniforms = N(this.outputShape.length), this.userCode = `
-      vec4 unaryOperation(vec4 x) {
-        ${s}
-      }
-
-      void main() {
-        vec4 x = getAAtOutCoords();
-        vec4 y = unaryOperation(x);
-
-        setOutput(y);
-      }
-    `;
-  }
-}
-const Ie = "if (isnan(x)) return x;";
-function Ae({ opSnippet: t, packedOpSnippet: e, cpuKernelImpl: s, dtype: r }) {
+const Oe = "if (isnan(x)) return x;";
+function ye({ opSnippet: t, packedOpSnippet: e, cpuKernelImpl: s, dtype: r }) {
   return ({ inputs: u, backend: n }) => {
     const { x: o } = u, i = n, a = r || o.dtype;
     if (i.shouldExecuteOnCPU([o]) && s != null) {
       const d = i.texData.get(o.dataId), y = s(d.values, a);
       return i.makeTensorInfo(o.shape, a, y);
     }
-    const c = O().getBool("WEBGL_PACK_UNARY_OPERATIONS") && e != null;
+    const c = N().getBool("WEBGL_PACK_UNARY_OPERATIONS") && e != null;
     let l;
-    return c ? l = new ue(o.shape, e) : l = new M(o.shape, t), i.runWebGLProgram(l, [o], a);
+    return c ? l = new ne(o.shape, e) : l = new q(o.shape, t), i.runWebGLProgram(l, [o], a);
   };
 }
-function Ce({ opSnippet: t, packedOpSnippet: e, checkOutOfBounds: s = !1, supportsComplex: r = !1, cpuKernelImpl: u, dtype: n }) {
+function Ie({ opSnippet: t, packedOpSnippet: e, checkOutOfBounds: s = !1, supportsComplex: r = !1, cpuKernelImpl: u, dtype: n }) {
   return ({ inputs: o, backend: i }) => {
     const { a, b: c } = o, l = i;
     if (r && a.dtype === "complex64") {
@@ -236,7 +234,7 @@ function Ce({ opSnippet: t, packedOpSnippet: e, checkOutOfBounds: s = !1, suppor
         [h.complexTensorInfos.real, f.complexTensorInfos.real],
         [h.complexTensorInfos.imag, f.complexTensorInfos.imag]
       ].map((C) => {
-        const [p, x] = C, $ = {
+        const [p, x] = C, U = {
           dataId: p.dataId,
           dtype: p.dtype,
           shape: a.shape
@@ -245,8 +243,8 @@ function Ce({ opSnippet: t, packedOpSnippet: e, checkOutOfBounds: s = !1, suppor
           dtype: x.dtype,
           shape: c.shape
         }, D = new b(t, a.shape, c.shape);
-        return l.runWebGLProgram(D, [$, _], S(p.dtype, x.dtype));
-      }), A = T({ inputs: { real: g, imag: m }, backend: l });
+        return l.runWebGLProgram(D, [U, _], S(p.dtype, x.dtype));
+      }), A = L({ inputs: { real: g, imag: m }, backend: l });
       return l.disposeIntermediateTensorInfo(g), l.disposeIntermediateTensorInfo(m), A;
     }
     const d = n || S(a.dtype, c.dtype);
@@ -260,49 +258,49 @@ function Ce({ opSnippet: t, packedOpSnippet: e, checkOutOfBounds: s = !1, suppor
       ) : f, [A, C] = u(a.shape, c.shape, g, m, d), p = l.makeTensorInfo(C, d), x = l.texData.get(p.dataId);
       return x.values = A, p;
     }
-    const y = O().getBool("WEBGL_PACK_BINARY_OPERATIONS") && e != null;
+    const y = N().getBool("WEBGL_PACK_BINARY_OPERATIONS") && e != null;
     let I;
     return y ? I = new E(e, a.shape, c.shape, s) : I = new b(t, a.shape, c.shape), l.runWebGLProgram(I, [a, c], d);
   };
 }
-function Pe(t, e = !1) {
+function Ae(t, e = !1) {
   if (t === "linear")
-    return e ? se : X;
+    return e ? ee : M;
   if (t === "relu")
-    return e ? re : Q;
+    return e ? se : J;
   if (t === "elu")
-    return e ? ae : J;
+    return e ? te : j;
   if (t === "relu6")
-    return e ? ne : ee;
+    return e ? ae : Q;
   if (t === "prelu")
-    return e ? k : R;
+    return e ? $ : k;
   if (t === "leakyrelu")
-    return e ? w : L;
+    return e ? R : w;
   if (t === "sigmoid")
-    return e ? oe : te;
+    return e ? re : X;
   throw new Error(`Activation ${t} has not been implemented for the WebGL backend.`);
 }
 export {
-  Oe as A,
+  de as A,
   E as B,
-  U as C,
-  ue as U,
+  T as C,
+  ne as U,
   Z as a,
-  Ce as b,
-  fe as c,
-  ye as d,
-  M as e,
-  T as f,
-  Y as g,
+  Ie as b,
+  pe as c,
+  he as d,
+  q as e,
+  L as f,
+  H as g,
   b as h,
   P as i,
-  he as j,
+  fe as j,
   xe as k,
-  Ie as l,
-  Pe as m,
+  Oe as l,
+  Ae as m,
   me as n,
   ge as o,
   be as p,
   Ne as q,
-  Ae as u
+  ye as u
 };

package/dist/layers/BaseLayer.js

@@ -1,7 +1,6 @@
-import { D as p, w as g, e as o, F as v } from "../index-D6Q1lPZO.js";
-import { v as _ } from "../variable-DzfrwYuP.js";
-import { T as V } from "../tensor-CzmOBsdf.js";
-class F {
+import { x as p, a4 as g, a2 as v, e as o, a5 as _ } from "../index-D0RBWjq8.js";
+import { v as V } from "../variable-DQ9yYgEU.js";
+class T {
   parent;
   config;
   _variables = /* @__PURE__ */ new Map();
@@ -52,9 +51,9 @@ class F {
     const e = this._variables.get(t);
     if (!e && r)
       for (const i of this.children) {
-        const s = i.getVariable(t, !0);
-        if (s)
-          return s;
+        const a = i.getVariable(t, !0);
+        if (a)
+          return a;
       }
     if (!e)
       throw new Error(`Variable ${t} not found`);
@@ -80,7 +79,7 @@ class F {
       const i = t.get(e)?.[0];
       if (!i)
         throw new Error(`Weights for ${e} not found`);
-      r ? r.assign(i) : this._variables.set(e, _(i, this._trainable));
+      r ? r.assign(i) : this._variables.set(e, V(i, this._trainable));
     }), this.children.forEach((r) => {
       r.loadWeights(t);
     });
@@ -98,7 +97,7 @@ class F {
   call(t, ...r) {
     this.build();
     const e = this.forward(t, ...r);
-    if (t.training && e instanceof V) {
+    if (t.training && e instanceof g) {
       const i = this.dropout(e);
       return i !== e && e.dispose(), i;
     } else
@@ -108,22 +107,22 @@ class F {
     return this.build(), this.checkpointingFn(t, ...r);
   }
   checkpointingFn(t, ...r) {
-    const e = this.trainableVariables, s = g((...a) => {
-      const l = a[a.length - 1], n = a.slice(0, r.length), h = this.forward(t, ...n);
+    const e = this.trainableVariables, a = v((...s) => {
+      const l = s[s.length - 1], n = s.slice(0, r.length), h = this.forward(t, ...n);
       return l(n), { value: h, gradFunc: (c, f) => {
         const u = o().state.activeTape;
         o().state.activeTape = [];
-        const b = v((...d) => this.forward(t, ...d.slice(0, n.length)))([...f, ...e], c);
+        const b = _((...d) => this.forward(t, ...d.slice(0, n.length)))([...f, ...e], c);
         return o().state.activeTape = u, b;
       } };
     })(...r, ...e);
     if (t.training) {
-      const a = this.dropout(s);
-      return a !== s && s.dispose(), a;
+      const s = this.dropout(a);
+      return s !== a && a.dispose(), s;
     } else
-      return s;
+      return a;
   }
 }
 export {
-  F as default
+  T as default
 };

package/dist/layers/CausalSelfAttention.js

@@ -1,18 +1,18 @@
 import { attentionMask as P } from "../ops/attentionMask.js";
 import _ from "./BaseLayer.js";
-import { r as k } from "../rope-s4W2XO9B.js";
+import { r as k } from "../rope-BmZmp9uP.js";
 import { appendCache as b } from "../ops/appendCache.js";
-import { k as u, t as S } from "../index-D6Q1lPZO.js";
+import { k as u, t as S } from "../index-D0RBWjq8.js";
 import { softmax16 as C } from "../ops/softmax16.js";
-import { b as R } from "../matMul16-fEAJ4smh.js";
-import { p as V } from "../pack16-CmVZs6af.js";
+import { b as R } from "../matMul16-cDxwemKj.js";
+import { p as V } from "../pack16-k4jq6aMX.js";
 import { transpose16 as J } from "../ops/transpose16.js";
 import { dot16 as L } from "../ops/dot16.js";
 import { reshape16 as N } from "../ops/reshape16.js";
 import { isPackedTensor as l } from "../utilities/packed.js";
 import { qkv as j } from "../ops/qkv.js";
-import { v as T } from "../variable-DzfrwYuP.js";
-import { r as v, d as y } from "../dropout-B1x1kYMa.js";
+import { v as T } from "../variable-DQ9yYgEU.js";
+import { r as v, d as y } from "../dropout-OxuaJz6z.js";
 class X extends _ {
   divisor;
   index;

package/dist/layers/MLP.js

@@ -1,9 +1,9 @@
-import { t as p } from "../index-D6Q1lPZO.js";
+import { t as p } from "../index-D0RBWjq8.js";
 import u from "./BaseLayer.js";
-import { c, b as M } from "../matMul16-fEAJ4smh.js";
+import { c, b as M } from "../matMul16-cDxwemKj.js";
 import { reshape16 as o } from "../ops/reshape16.js";
-import { v as d } from "../variable-DzfrwYuP.js";
-import { r as h, d as b } from "../dropout-B1x1kYMa.js";
+import { v as d } from "../variable-DQ9yYgEU.js";
+import { r as h, d as b } from "../dropout-OxuaJz6z.js";
 class O extends u {
   index;
   hiddenUnits;

package/dist/layers/PositionEmbedding.js

@@ -1,9 +1,9 @@
-import { t as c, I as u, b as i } from "../index-D6Q1lPZO.js";
+import { t as c, al as u, b as i } from "../index-D0RBWjq8.js";
 import f from "./BaseLayer.js";
-import { E as g, D as h } from "../random_width-BVV9HveY.js";
-import { r as b } from "../exports_initializers-CZSUJoVE.js";
-import { m as l } from "../mod-DESSvHIU.js";
-import { r as w } from "../range-ZZZD60Fx.js";
+import { E as g, D as h } from "../random_width-UGQn4OWb.js";
+import { r as l } from "../exports_initializers-eS9QJ6ut.js";
+import { m as b } from "../mod-PrOKlFxH.js";
+import { r as w } from "../range-CuGvVN2c.js";
 function E(t) {
   return new h(t);
 }
@@ -20,13 +20,13 @@ class q extends f {
       inputDim: this.config.blockSize,
       outputDim: this.config.nEmbed,
       name: n,
-      embeddingsInitializer: b({ mean: 0, stddev: 0.02 })
+      embeddingsInitializer: l({ mean: 0, stddev: 0.02 })
     }), this.drop = E({ rate: this.config.dropout });
   }
   forward(o, n) {
     const r = o.cache?.[0]?.length ?? 0;
     return c(() => {
-      const [, s] = n.shape, e = this.config.blockSize, a = w(0, s, 1, "int32"), m = l(u(a, i(r, "int32")), i(e, "int32")), d = this.wpe.apply(m), p = n.add(d);
+      const [, s] = n.shape, e = this.config.blockSize, a = w(0, s, 1, "int32"), m = b(u(a, i(r, "int32")), i(e, "int32")), d = this.wpe.apply(m), p = n.add(d);
       return this.drop.apply(p, { training: o.training });
     });
   }

package/dist/layers/RMSNorm.js

@@ -1,8 +1,8 @@
-import { t as s } from "../index-D6Q1lPZO.js";
+import { t as s } from "../index-D0RBWjq8.js";
 import e from "./BaseLayer.js";
 import { normRMS as a } from "../ops/normRMS.js";
-import { v as i } from "../variable-DzfrwYuP.js";
-import { o as m } from "../ones-jU9jlQvM.js";
+import { v as i } from "../variable-DQ9yYgEU.js";
+import { o as m } from "../ones-BX_wEgzB.js";
 class l extends e {
   GAMMA;
   constructor(r, t = "", o) {

package/dist/layers/RoPECache.js

@@ -1,16 +1,16 @@
-import { b as n, p as a, t as p, k as r } from "../index-D6Q1lPZO.js";
-import { r as c } from "../reciprocal-CrYlsAGD.js";
-import { c as l, s as f } from "../sin-Djs4aQiu.js";
-import { r as h } from "../range-ZZZD60Fx.js";
-class g {
+import { l as a, b as n, p as c, t as p, k as r } from "../index-D0RBWjq8.js";
+import { r as l } from "../zeros_like-BAz3iKru.js";
+import { c as f, s as m } from "../unsorted_segment_sum-LAbmE9G4.js";
+import { r as h } from "../range-CuGvVN2c.js";
+class D {
   constructor(s) {
     this.config = s;
     const i = this.config.nEmbed / this.config.nHead;
     if (this.rotaryDim = i, this.rotaryDim % 2 !== 0)
       throw new Error("rotaryDim must be even");
     this.ropeBase = 1e4;
-    const o = h(0, this.rotaryDim, 2, "float32"), e = o.div(n(this.rotaryDim, "float32")), t = a(n(this.ropeBase, "float32"), e);
-    this.ropeInvFreq = c(t), e.dispose(), t.dispose(), o.dispose(), this.config.useRope === !1 ? (this.ropeCos = null, this.ropeSin = null, this.ropeNegSin = null, this.ropeCacheLen = 0) : p(() => {
+    const o = h(0, this.rotaryDim, 2, "float32"), e = a(o, n(this.rotaryDim, "float32")), t = c(n(this.ropeBase, "float32"), e);
+    this.ropeInvFreq = l(t), e.dispose(), t.dispose(), o.dispose(), this.config.useRope === !1 ? (this.ropeCos = null, this.ropeSin = null, this.ropeNegSin = null, this.ropeCacheLen = 0) : p(() => {
       this.ensureRopeCache(this.config.blockSize * 4);
     });
   }
@@ -29,7 +29,7 @@ class g {
       if (s <= this.ropeCacheLen) return;
       this.ropeCos && this.ropeCos.dispose(), this.ropeSin && this.ropeSin.dispose();
       const i = Math.max(s, this.ropeCacheLen + this.config.blockSize * 4), e = h(0, i, 1, "float32").expandDims(1).mul(this.ropeInvFreq.expandDims(0));
-      this.ropeCos = r(l(e).expandDims(-1)), this.ropeSin = r(f(e).expandDims(-1)), this.ropeNegSin = r(this.ropeSin.neg()), this.ropeCacheLen = i;
+      this.ropeCos = r(f(e).expandDims(-1)), this.ropeSin = r(m(e).expandDims(-1)), this.ropeNegSin = r(this.ropeSin.neg()), this.ropeCacheLen = i;
     });
   }
   getCos() {
@@ -46,5 +46,5 @@ class g {
   }
 }
 export {
-  g as default
+  D as default
 };

package/dist/layers/TiedEmbedding.js

@@ -1,13 +1,13 @@
-import "../random_width-BVV9HveY.js";
-import "../index-D6Q1lPZO.js";
-import { r as s } from "../exports_initializers-CZSUJoVE.js";
+import "../random_width-UGQn4OWb.js";
+import "../index-D0RBWjq8.js";
+import { r as s } from "../exports_initializers-eS9QJ6ut.js";
 import a from "./BaseLayer.js";
 import { dot16 as o } from "../ops/dot16.js";
 import { isPackedTensor as r } from "../utilities/packed.js";
-import { p as m } from "../pack16-CmVZs6af.js";
+import { p as m } from "../pack16-k4jq6aMX.js";
 import { transpose16 as d } from "../ops/transpose16.js";
-import { v as p } from "../variable-DzfrwYuP.js";
-import { g as h } from "../gather-CH9sdacz.js";
+import { v as p } from "../variable-DQ9yYgEU.js";
+import { g as h } from "../gather-BcO5UQNJ.js";
 class g extends a {
   vocabSize;
   embedDim;

package/dist/layers/TransformerBlock.js

@@ -2,7 +2,7 @@ import p from "./CausalSelfAttention.js";
 import m from "./MLP.js";
 import d from "./RMSNorm.js";
 import h from "./BaseLayer.js";
-import { k as n, t as u } from "../index-D6Q1lPZO.js";
+import { k as n, t as u } from "../index-D0RBWjq8.js";
 import { add16 as l } from "../ops/add16.js";
 class g extends h {
   ln1;

package/dist/loader/loadTransformers.js

@@ -1,7 +1,7 @@
 import l from "../tokeniser/CharTokeniser.js";
 import c from "../tokeniser/bpe.js";
 import { load_safetensors as b } from "../utilities/safetensors.js";
-import { H as y } from "../index-D6Q1lPZO.js";
+import { aa as y } from "../index-D0RBWjq8.js";
 import { dummyPassAsync as u } from "../utilities/dummy.js";
 import _ from "../models/factory.js";
 async function L(e, a, r, t) {