fix: slow types

Author: load1n9

packages/core/src/core/mod.ts

@@ -12,6 +12,7 @@ import type { Tensor } from "./tensor/tensor.ts";
 import type { NeuralNetwork } from "./api/network.ts";
 import { SGDOptimizer } from "./api/optimizer.ts";
 import { PostProcess, type PostProcessor } from "./api/postprocess.ts";
+import type { DenseLayerConfig } from "./api/layer.ts";
 
 /**
  * Sequential Neural Network
@@ -47,21 +48,22 @@ export class Sequential implements NeuralNetwork {
    */
   async predict(
     data: Tensor<Rank>,
-    config?: { postProcess?: PostProcessor; layers?: [number, number] }
+    config?: { postProcess?: PostProcessor; layers?: [number, number] },
   ): Promise<Tensor<Rank>> {
-    if (!config)
+    if (!config) {
       config = {
         postProcess: PostProcess("none"),
       };
+    }
     if (config.layers) {
       if (
         config.layers[0] < 0 ||
         config.layers[1] > this.config.layers.length
       ) {
         throw new RangeError(
-          `Execution range should be within (0, ${
-            this.config.layers.length
-          }). Received (${(config.layers[0], config.layers[1])})`
+          `Execution range should be within (0, ${this.config.layers.length}). Received (${(config
+            .layers[0],
+            config.layers[1])})`,
         );
       }
       const lastLayer = this.config.layers[config.layers[1] - 1];
@@ -77,9 +79,12 @@ export class Sequential implements NeuralNetwork {
           data,
           {
             postProcess: config.postProcess || PostProcess("none"),
-            outputShape: lastLayer.config.size,
+            outputShape: (lastLayer as {
+              type: LayerType.Dense;
+              config: DenseLayerConfig;
+            }).config.size,
           },
-          layerList
+          layerList,
         );
       } else if (lastLayer.type === LayerType.Activation) {
         const penultimate = this.config.layers[config.layers[1] - 2];
@@ -91,26 +96,29 @@ export class Sequential implements NeuralNetwork {
             data,
             {
               postProcess: config.postProcess || PostProcess("none"),
-              outputShape: penultimate.config.size,
+              outputShape: (penultimate as {
+                type: LayerType.Dense;
+                config: DenseLayerConfig;
+              }).config.size,
             },
-            layerList
+            layerList,
           );
         } else {
           throw new Error(
-            `The penultimate layer must be a dense layer, or a flatten layer if the last layer is an activation layer. Received ${penultimate.type}.`
+            `The penultimate layer must be a dense layer, or a flatten layer if the last layer is an activation layer. Received ${penultimate.type}.`,
           );
         }
       } else {
         throw new Error(
-          `The output layer must be a dense layer, activation layer, or a flatten layer. Received ${lastLayer.type}.`
+          `The output layer must be a dense layer, activation layer, or a flatten layer. Received ${lastLayer.type}.`,
         );
       }
     }
     return await this.backend.predict(
       data,
       config.postProcess
         ? (config as { postProcess: PostProcessor; layers?: [number, number] })
-        : { ...config, postProcess: PostProcess("none") }
+        : { ...config, postProcess: PostProcess("none") },
     );
   }
 

packages/utilities/src/utils/misc/argmax.ts

@@ -1,4 +1,4 @@
-export function argmax(mat: ArrayLike<number | bigint>) {
+export function argmax(mat: ArrayLike<number | bigint>): number {
   let max = mat[0];
   let index = 0;
   for (let i = 0; i < mat.length; i++) {

packages/utilities/src/utils/misc/matrix.ts

@@ -27,10 +27,8 @@ export type MatrixLike<DT extends DataType> = {
  * This is a collection of row vectors.
  * A special case of Tensor for 2D data.
  */
-export class Matrix<DT extends DataType>
-  extends Tensor<DT, 2>
-  implements Sliceable, MatrixLike<DT>
-{
+export class Matrix<DT extends DataType> extends Tensor<DT, 2>
+  implements Sliceable, MatrixLike<DT> {
   /**
    * Create a matrix from a typed array
    * @param data Data to move into the matrix.
@@ -42,15 +40,15 @@ export class Matrix<DT extends DataType>
   constructor(dType: DT, shape: Shape<2>);
   constructor(
     data: NDArray<DT>[2] | DType<DT> | DT | TensorLike<DT, 2>,
-    shape?: Shape<2> | DT
+    shape?: Shape<2> | DT,
   ) {
     // @ts-ignore This call will work
     super(data, shape);
   }
-  get head() {
+  get head(): Matrix<DT> {
     return this.slice(0, Math.min(this.nRows, 10));
   }
-  get tail() {
+  get tail(): Matrix<DT> {
     return this.slice(Math.max(this.nRows - 10, 0), this.nRows);
   }
   /** Convert the Matrix into a HTML table */
@@ -87,7 +85,7 @@ export class Matrix<DT extends DataType>
   /** Get the transpose of the matrix. This method clones the matrix. */
   get T(): Matrix<DT> {
     const resArr = new (this.data.constructor as DTypeConstructor<DT>)(
-      this.nRows * this.nCols
+      this.nRows * this.nCols,
     ) as DType<DT>;
     let i = 0;
     for (const col of this.cols()) {
@@ -114,7 +112,7 @@ export class Matrix<DT extends DataType>
   col(n: number): DType<DT> {
     let i = 0;
     const col = new (this.data.constructor as DTypeConstructor<DT>)(
-      this.nRows
+      this.nRows,
     ) as DType<DT>;
     let offset = 0;
     while (i < this.nRows) {
@@ -139,7 +137,7 @@ export class Matrix<DT extends DataType>
   /** Get a column array of all column sums in the matrix */
   colSum(): DType<DT> {
     const sum = new (this.data.constructor as DTypeConstructor<DT>)(
-      this.nRows
+      this.nRows,
     ) as DType<DT>;
     let i = 0;
     while (i < this.nCols) {
@@ -169,8 +167,7 @@ export class Matrix<DT extends DataType>
     while (j < this.nCols) {
       let i = 0;
       while (i < this.nRows) {
-        const adder =
-          (this.item(i, j) as DTypeValue<DT>) *
+        const adder = (this.item(i, j) as DTypeValue<DT>) *
           (rhs.item(i, j) as DTypeValue<DT>);
         // @ts-ignore I'll fix this later
         res += adder as DTypeValue<DT>;
@@ -182,7 +179,7 @@ export class Matrix<DT extends DataType>
   }
   /** Filter the matrix by rows */
   override filter(
-    fn: (value: DType<DT>, row: number, _: DType<DT>[]) => boolean
+    fn: (value: DType<DT>, row: number, _: DType<DT>[]) => boolean,
   ): Matrix<DT> {
     const satisfying: number[] = [];
     let i = 0;
@@ -224,7 +221,7 @@ export class Matrix<DT extends DataType>
   /** Compute the sum of all rows */
   rowSum(): DType<DT> {
     const sum = new (this.data.constructor as DTypeConstructor<DT>)(
-      this.nCols
+      this.nCols,
     ) as DType<DT>;
     let i = 0;
     let offset = 0;
@@ -271,9 +268,9 @@ export class Matrix<DT extends DataType>
     return new Matrix<DT>(
       this.data.slice(
         start ? start * this.nCols : 0,
-        end ? end * this.nCols : undefined
+        end ? end * this.nCols : undefined,
       ) as DType<DT>,
-      [end ? end - start : this.nRows - start, this.nCols]
+      [end ? end - start : this.nRows - start, this.nCols],
     );
   }
   /** Iterate through rows */
@@ -290,7 +287,7 @@ export class Matrix<DT extends DataType>
     while (i < this.nCols) {
       let j = 0;
       const col = new (this.data.constructor as DTypeConstructor<DT>)(
-        this.nRows
+        this.nRows,
       ) as DType<DT>;
       while (j < this.nRows) {
         col[j] = this.data[j * this.nCols + i];