Model trainer: Make the shore focused loss function even more focused

Dependencies

• [2] ABITXOUU Model trainer: Add provisional colouration
• [3] 7ML3OFE7 Model trainer: initial train and visualize thread pair (total crust mass; todo: altitude instead)
• [4] QMNTEX7P Try a loss function that has higher values for errors near sea level

Change contents

• replacement in trainmodel/src/model.py at line 29
  [3.40]→[2.184:266](∅→∅)
  − y_true = tf.math.asinh(y_true - 0.1)

  − y_pred = tf.math.asinh(y_pred - 0.1)
  [3.40]

  [3.130]
  + y_true = tf.math.asinh((y_true - 0.1) * 5.) / 5.

  + y_pred = tf.math.asinh((y_pred - 0.1) * 5.) / 5.