Method trainOnBatch

tfjs-layers/src/models.ts:997–1002

* Runs a single gradient update on a single batch of data. * * This method differs from `fit()` and `fitDataset()` in the following * regards: * - It operates on exactly one batch of data. * - It returns only the loss and metric values, instead of * returning the batch-by-b

(
      x: Tensor|Tensor[]|{[inputName: string]: Tensor},
      y: Tensor|Tensor[]|
      {[inputName: string]: Tensor})

   * @doc {heading: 'Models', subheading: 'Classes'}
   */
  override async trainOnBatch(
      x: Tensor|Tensor[]|{[inputName: string]: Tensor},
      y: Tensor|Tensor[]|
      {[inputName: string]: Tensor}): Promise<number|number[]> {
    return this.model.trainOnBatch(x, y);
  }

  /* See parent class for JsDoc */
  /** @nocollapse */