diff --git a/TD2 Deep Learning.ipynb b/TD2 Deep Learning.ipynb
index 90f575919287e183369e281825d77792250beeda..f860ee72920401fec866564f1fd7848726d49ca6 100644
--- a/TD2 Deep Learning.ipynb	
+++ b/TD2 Deep Learning.ipynb	
@@ -2410,7 +2410,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "fda7f893",
+   "id": "75eedfd8",
    "metadata": {},
    "source": [
     "# Last layer modification"
@@ -2419,7 +2419,7 @@
   {
    "cell_type": "code",
    "execution_count": 8,
-   "id": "71aec74d",
+   "id": "23ed4bb7",
    "metadata": {
     "scrolled": true
    },
@@ -2673,7 +2673,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "0b216cdf",
+   "id": "d092b026",
    "metadata": {},
    "source": [
     "The output data is really similar to the previous result. This can be because the test data set is small"
@@ -2681,7 +2681,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "79411410",
+   "id": "59a0d2c6",
    "metadata": {},
    "source": [
     "# Quantization  "
@@ -2690,7 +2690,7 @@
   {
    "cell_type": "code",
    "execution_count": 15,
-   "id": "9e3b6a11",
+   "id": "78a660bc",
    "metadata": {},
    "outputs": [
     {
@@ -2732,7 +2732,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "e2f2d2d7",
+   "id": "a481c963",
    "metadata": {},
    "source": [
     "Hence the modification of the last layer doesn't affect that much the size of the model"