Adapted to changed evaluation classes

opcode81 · Sep 27, 2023 · 3e98b09 · 3e98b09
1 parent 27a7049
commit 3e98b09
Showing 1 changed file with 8 additions and 8 deletions.
diff --git a/notebooks/neural_networks.ipynb b/notebooks/neural_networks.ipynb
@@ -2,16 +2,16 @@
  "cells": [
   {
    "cell_type": "markdown",
+   "metadata": {
+    "collapsed": false
+   },
    "source": [
     "# Neural Networks\n",
     "\n",
     "Neural networks being a very powerful class of models, especially in cases where the learning of representations from low-level information (such as pixels, audio samples or text) is key, sensAI provides many useful abstractions for dealing with this class of models, facilitating data handling, learning and evaluation.\n",
     "\n",
     "sensAI mainly provides abstractions for PyTorch, but there is also rudimentary support for TensorFlow."
-   ],
-   "metadata": {
-    "collapsed": false
-   }
+   ]
   },
   {
    "cell_type": "code",
@@ -137,8 +137,8 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "evaluator_params = sensai.evaluation.VectorClassificationModelEvaluatorParams(fractional_split_test_fraction=0.2)\n",
-    "eval_util = sensai.evaluation.ClassificationEvaluationUtil(mnistIoData, evaluator_params=evaluator_params)"
+    "evaluator_params = sensai.evaluation.ClassificationEvaluatorParams(fractional_split_test_fraction=0.2)\n",
+    "eval_util = sensai.evaluation.ClassificationModelEvaluation(mnistIoData, evaluator_params=evaluator_params)"
    ]
   },
   {
@@ -286,7 +286,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "In this case, we derived the class from ``RuleBasedTensorised`` rather than ``Tensoriser``, because our tensoriser does not require fitting. We additionally took care of the normalisation.\n",
+    "In this case, we derived the class from ``RuleBasedTensoriser`` rather than ``Tensoriser``, because our tensoriser does not require fitting. We additionally took care of the normalisation.\n",
     "\n",
     "Now we have all we need to create a sensAI ``TorchVectorClassificationModel`` that will work on the input/output data we loaded earlier."
    ]
@@ -486,7 +486,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.8.13"
+   "version": "3.8.18"
   },
   "orig_nbformat": 4
  },