@hoff97/tensor-js/dist/lib/onnx/nodes/softmax.js

PyTorch like deep learning inferrence library

var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, generator) {
    function adopt(value) { return value instanceof P ? value : new P(function (resolve) { resolve(value); }); }
    return new (P || (P = Promise))(function (resolve, reject) {
        function fulfilled(value) { try { step(generator.next(value)); } catch (e) { reject(e); } }
        function rejected(value) { try { step(generator["throw"](value)); } catch (e) { reject(e); } }
        function step(result) { result.done ? resolve(result.value) : adopt(result.value).then(fulfilled, rejected); }
        step((generator = generator.apply(thisArg, _arguments || [])).next());
    });
};
import { OnnxNode } from '../node';
export class SoftmaxNode extends OnnxNode {
    constructor(attributes, inputs, outputs, constants, onnxVersion, mode) {
        super(attributes, inputs, outputs, constants, onnxVersion, mode);
        //@ts-ignore
        this.axis = this.getAttributeInt('axis');
    }
    forward(inputs) {
        return __awaiter(this, void 0, void 0, function* () {
            const x = inputs[0];
            const shapeX = x.getShape();
            let ax = this.axis;
            if (ax === undefined) {
                if (this.onnxVersion < 13) {
                    ax = 1;
                }
                else {
                    ax = shapeX.length - 1;
                }
            }
            const sh1 = shapeX.slice(0, ax).reduce((x, y) => x * y, 1);
            const reshaped = x.reshape([sh1, -1], false);
            const max = reshaped.max(1, true);
            const normalized = reshaped.subtract(max);
            const exp = normalized.exp();
            const sum = exp.sum(1, true);
            const result = exp.divide(sum);
            max.delete();
            normalized.delete();
            exp.delete();
            sum.delete();
            return [result.reshape(shapeX, false)];
        });
    }
    getType() {
        return 'Softmax';
    }
    delete() { }
}
//# sourceMappingURL=softmax.js.map