Add validation dataset (#1719)

keras-team · Jan 10, 2024 · 5aff074 · 5aff074
1 parent 1d8d323
commit 5aff074
Show file tree

Hide file tree

Showing 5 changed files with 16,102 additions and 635 deletions.
diff --git a/examples/vision/img/pointnet/pointnet_10_0.png b/examples/vision/img/pointnet/pointnet_10_0.png
diff --git a/examples/vision/img/pointnet/pointnet_28_2.png b/examples/vision/img/pointnet/pointnet_28_2.png
diff --git a/examples/vision/ipynb/pointnet.ipynb b/examples/vision/ipynb/pointnet.ipynb
@@ -10,7 +10,7 @@
     "\n",
     "**Author:** [David Griffiths](https://dgriffiths3.github.io)<br>\n",
     "**Date created:** 2020/05/25<br>\n",
-    "**Last modified:** 2020/05/26<br>\n",
+    "**Last modified:** 2024/01/09<br>\n",
     "**Description:** Implementation of PointNet for ModelNet10 classification."
    ]
   },
@@ -68,7 +68,7 @@
     "from keras import layers\n",
     "from matplotlib import pyplot as plt\n",
     "\n",
-    "keras.random.SeedGenerator(seed=42)"
+    "keras.utils.set_random_seed(seed=42)"
    ]
   },
   {
@@ -259,11 +259,16 @@
     "    return points, label\n",
     "\n",
     "\n",
-    "train_dataset = tf_data.Dataset.from_tensor_slices((train_points, train_labels))\n",
+    "train_size = 0.8\n",
+    "dataset = tf_data.Dataset.from_tensor_slices((train_points, train_labels))\n",
     "test_dataset = tf_data.Dataset.from_tensor_slices((test_points, test_labels))\n",
+    "train_dataset_size = int(len(dataset) * train_size)\n",
     "\n",
-    "train_dataset = train_dataset.shuffle(len(train_points)).map(augment).batch(BATCH_SIZE)\n",
-    "test_dataset = test_dataset.shuffle(len(test_points)).batch(BATCH_SIZE)"
+    "dataset = dataset.shuffle(len(train_points)).map(augment)\n",
+    "test_dataset = test_dataset.shuffle(len(test_points)).batch(BATCH_SIZE)\n",
+    "\n",
+    "train_dataset = dataset.take(train_dataset_size).batch(BATCH_SIZE)\n",
+    "validation_dataset = dataset.skip(train_dataset_size).batch(BATCH_SIZE)"
    ]
   },
   {
@@ -445,7 +450,7 @@
     "    metrics=[\"sparse_categorical_accuracy\"],\n",
     ")\n",
     "\n",
-    "model.fit(train_dataset, epochs=20, validation_data=test_dataset)"
+    "model.fit(train_dataset, epochs=20, validation_data=validation_dataset)"
    ]
   },
   {