Remove unused code and fix JSDoc (#219)

* Remove unused code and fixes JSDoc
* Adds types to JSDoc.

Author: bileschi

mnist/index.js

@@ -104,10 +104,22 @@ function createDenseModel() {
   return model;
 }
 
+/**
+ * This callback type is used by the `train` function for insertion into
+ * the model.fit callback loop.
+ *
+ * @callback onIterationCallback
+ * @param {string} eventType Selector for which type of event to fire on.
+ * @param {number} batchOrEpochNumber The current epoch / batch number
+ * @param {tf.Logs} logs Logs to append to
+ */
+
 /**
  * Compile and train the given model.
  *
- * @param {*} model The model to
+ * @param {tf.Model} model The model to train.
+ * @param {onIterationCallback} onIteration A callback to execute every 10
+ *     batches & epoch end.
  */
 async function train(model, onIteration) {
   ui.logStatus('Training model...');
@@ -117,17 +129,6 @@ async function train(model, onIteration) {
   // training so that we can decrease our training loss and increase our
   // classification accuracy.
 
-  // The learning rate defines the magnitude by which we update our weights each
-  // training step. The higher the value, the faster our loss values converge,
-  // but also the more likely we are to overshoot optimal parameters
-  // when making an update. A learning rate that is too low will take too long
-  // to find optimal (or good enough) weight parameters while a learning rate
-  // that is too high may overshoot optimal parameters. Learning rate is one of
-  // the most important hyperparameters to set correctly. Finding the right
-  // value takes practice and is often best found empirically by trying many
-  // values.
-  const LEARNING_RATE = 0.01;
-
   // We are using rmsprop as our optimizer.
   // An optimizer is an iterative method for minimizing an loss function.
   // It tries to find the minimum of our loss function with respect to the