Migrate Review classification using Active learning to Keras 3. (kera…

…s-team#1857)

* Migarte to Keras 3

* Migrate to Keras 3

* trim extra epoch outputs

examples/nlp/active_learning_review_classification.py

@@ -2,9 +2,10 @@
 Title: Review Classification using Active Learning
 Author: [Darshan Deshpande](https://twitter.com/getdarshan)
 Date created: 2021/10/29
-Last modified: 2021/10/29
+Last modified: 2024/05/08
 Description: Demonstrating the advantages of active learning through review classification.
 Accelerator: GPU
+Converted to Keras 3 by: [Sachin Prasad](https://github.com/sachinprasadhs)
 """
 
 """
@@ -51,10 +52,14 @@
 ## Importing required libraries
 """
 
+import os
+
+os.environ["KERAS_BACKEND"] = "tensorflow"  # @param ["tensorflow", "jax", "torch"]
+import keras
+from keras import ops
+from keras import layers
 import tensorflow_datasets as tfds
 import tensorflow as tf
-from tensorflow import keras
-from tensorflow.keras import layers
 import matplotlib.pyplot as plt
 import re
 import string
@@ -169,16 +174,8 @@
 """
 
 
-def custom_standardization(input_data):
-    lowercase = tf.strings.lower(input_data)
-    stripped_html = tf.strings.regex_replace(lowercase, "<br />", " ")
-    return tf.strings.regex_replace(
-        stripped_html, f"[{re.escape(string.punctuation)}]", ""
-    )
-
-
 vectorizer = layers.TextVectorization(
-    3000, standardize=custom_standardization, output_sequence_length=150
+    3000, standardize="lower_and_strip_punctuation", output_sequence_length=150
 )
 # Adapting the dataset
 vectorizer.adapt(
@@ -289,7 +286,7 @@ def train_full_model(full_train_dataset, val_dataset, test_dataset):
         callbacks=[
             keras.callbacks.EarlyStopping(patience=4, verbose=1),
             keras.callbacks.ModelCheckpoint(
-                "FullModelCheckpoint.h5", verbose=1, save_best_only=True
+                "FullModelCheckpoint.keras", verbose=1, save_best_only=True
             ),
         ],
     )
@@ -303,7 +300,7 @@ def train_full_model(full_train_dataset, val_dataset, test_dataset):
     )
 
     # Loading the best checkpoint
-    model = keras.models.load_model("FullModelCheckpoint.h5")
+    model = keras.models.load_model("FullModelCheckpoint.keras")
 
     print("-" * 100)
     print(
@@ -370,6 +367,7 @@ def train_active_learning_models(
     num_iterations=3,
     sampling_size=5000,
 ):
+
     # Creating lists for storing metrics
     losses, val_losses, accuracies, val_accuracies = [], [], [], []
 
@@ -389,7 +387,7 @@ def train_active_learning_models(
     # Defining checkpoints.
     # The checkpoint callback is reused throughout the training since it only saves the best overall model.
     checkpoint = keras.callbacks.ModelCheckpoint(
-        "AL_Model.h5", save_best_only=True, verbose=1
+        "AL_Model.keras", save_best_only=True, verbose=1
     )
     # Here, patience is set to 4. This can be set higher if desired.
     early_stopping = keras.callbacks.EarlyStopping(patience=4, verbose=1)
@@ -413,9 +411,9 @@ def train_active_learning_models(
         predictions = model.predict(test_dataset)
 
         # Generating labels from the output probabilities
-        rounded = tf.where(tf.greater(predictions, 0.5), 1, 0)
+        rounded = ops.where(ops.greater(predictions, 0.5), 1, 0)
 
-        # Evaluating the number of zeros and ones incorrectly classified
+        # Evaluating the number of zeros and ones incorrrectly classified
         _, _, false_negatives, false_positives = model.evaluate(test_dataset, verbose=0)
 
         print("-" * 100)
@@ -482,7 +480,7 @@ def train_active_learning_models(
         )
 
         # Loading the best model from this training loop
-        model = keras.models.load_model("AL_Model.h5")
+        model = keras.models.load_model("AL_Model.keras")
 
     # Plotting the overall history and evaluating the final model
     plot_history(losses, val_losses, accuracies, val_accuracies)

examples/nlp/ipynb/active_learning_review_classification.ipynb

@@ -10,7 +10,7 @@
     "\n",
     "**Author:** [Darshan Deshpande](https://twitter.com/getdarshan)<br>\n",
     "**Date created:** 2021/10/29<br>\n",
-    "**Last modified:** 2021/10/29<br>\n",
+    "**Last modified:** 2024/05/08<br>\n",
     "**Description:** Demonstrating the advantages of active learning through review classification."
    ]
   },
@@ -33,7 +33,7 @@
     "ensure consistency in labeling of new data.\n",
     "\n",
     "The process starts with annotating a small subset of the full dataset and training an\n",
-    "initial model. The best model checkpoint is saved and then tested on a balanced test\n",
+    "initial model. The best model checkpoint is saved and then tested on\u00a0a balanced test\n",
     "set. The test set must be carefully sampled because the full training process will be\n",
     "dependent on it. Once we have the initial evaluation scores, the oracle is tasked with\n",
     "labeling more samples; the number of data points to be sampled is usually determined by\n",
@@ -70,16 +70,20 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 0,
    "metadata": {
     "colab_type": "code"
    },
    "outputs": [],
    "source": [
+    "import os\n",
+    "\n",
+    "os.environ[\"KERAS_BACKEND\"] = \"tensorflow\"  # @param [\"tensorflow\", \"jax\", \"torch\"]\n",
+    "import keras\n",
+    "from keras import ops\n",
+    "from keras import layers\n",
     "import tensorflow_datasets as tfds\n",
     "import tensorflow as tf\n",
-    "from tensorflow import keras\n",
-    "from tensorflow.keras import layers\n",
     "import matplotlib.pyplot as plt\n",
     "import re\n",
     "import string\n",
@@ -102,7 +106,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 0,
    "metadata": {
     "colab_type": "code"
    },
@@ -133,7 +137,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 0,
    "metadata": {
     "colab_type": "code"
    },
@@ -229,23 +233,15 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 0,
    "metadata": {
     "colab_type": "code"
    },
    "outputs": [],
    "source": [
-    "\n",
-    "def custom_standardization(input_data):\n",
-    "    lowercase = tf.strings.lower(input_data)\n",
-    "    stripped_html = tf.strings.regex_replace(lowercase, \"<br />\", \" \")\n",
-    "    return tf.strings.regex_replace(\n",
-    "        stripped_html, f\"[{re.escape(string.punctuation)}]\", \"\"\n",
-    "    )\n",
-    "\n",
     "\n",
     "vectorizer = layers.TextVectorization(\n",
-    "    3000, standardize=custom_standardization, output_sequence_length=150\n",
+    "    3000, standardize=\"lower_and_strip_punctuation\", output_sequence_length=150\n",
     ")\n",
     "# Adapting the dataset\n",
     "vectorizer.adapt(\n",
@@ -283,12 +279,13 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 0,
    "metadata": {
     "colab_type": "code"
    },
    "outputs": [],
    "source": [
+    "\n",
     "# Helper function for merging new history objects with older ones\n",
     "def append_history(losses, val_losses, accuracy, val_accuracy, history):\n",
     "    losses = losses + history.history[\"loss\"]\n",
@@ -312,7 +309,8 @@
     "    plt.legend([\"train_accuracy\", \"val_accuracy\"])\n",
     "    plt.xlabel(\"Epochs\")\n",
     "    plt.ylabel(\"Accuracy\")\n",
-    "    plt.show()\n"
+    "    plt.show()\n",
+    ""
    ]
   },
   {
@@ -331,7 +329,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 0,
    "metadata": {
     "colab_type": "code"
    },
@@ -351,7 +349,8 @@
     "        ]\n",
     "    )\n",
     "    model.summary()\n",
-    "    return model\n"
+    "    return model\n",
+    ""
    ]
   },
   {
@@ -368,7 +367,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 0,
    "metadata": {
     "colab_type": "code"
    },
@@ -395,7 +394,7 @@
     "        callbacks=[\n",
     "            keras.callbacks.EarlyStopping(patience=4, verbose=1),\n",
     "            keras.callbacks.ModelCheckpoint(\n",
-    "                \"FullModelCheckpoint.h5\", verbose=1, save_best_only=True\n",
+    "                \"FullModelCheckpoint.keras\", verbose=1, save_best_only=True\n",
     "            ),\n",
     "        ],\n",
     "    )\n",
@@ -409,7 +408,7 @@
     "    )\n",
     "\n",
     "    # Loading the best checkpoint\n",
-    "    model = keras.models.load_model(\"FullModelCheckpoint.h5\")\n",
+    "    model = keras.models.load_model(\"FullModelCheckpoint.keras\")\n",
     "\n",
     "    print(\"-\" * 100)\n",
     "    print(\n",
@@ -474,7 +473,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 0,
    "metadata": {
     "colab_type": "code"
    },
@@ -510,7 +509,7 @@
     "    # Defining checkpoints.\n",
     "    # The checkpoint callback is reused throughout the training since it only saves the best overall model.\n",
     "    checkpoint = keras.callbacks.ModelCheckpoint(\n",
-    "        \"AL_Model.h5\", save_best_only=True, verbose=1\n",
+    "        \"AL_Model.keras\", save_best_only=True, verbose=1\n",
     "    )\n",
     "    # Here, patience is set to 4. This can be set higher if desired.\n",
     "    early_stopping = keras.callbacks.EarlyStopping(patience=4, verbose=1)\n",
@@ -534,9 +533,9 @@
     "        predictions = model.predict(test_dataset)\n",
     "\n",
     "        # Generating labels from the output probabilities\n",
-    "        rounded = tf.where(tf.greater(predictions, 0.5), 1, 0)\n",
+    "        rounded = ops.where(ops.greater(predictions, 0.5), 1, 0)\n",
     "\n",
-    "        # Evaluating the number of zeros and ones  classified\n",
+    "        # Evaluating the number of zeros and ones incorrrectly classified\n",
     "        _, _, false_negatives, false_positives = model.evaluate(test_dataset, verbose=0)\n",
     "\n",
     "        print(\"-\" * 100)\n",
@@ -603,7 +602,7 @@
     "        )\n",
     "\n",
     "        # Loading the best model from this training loop\n",
-    "        model = keras.models.load_model(\"AL_Model.h5\")\n",
+    "        model = keras.models.load_model(\"AL_Model.keras\")\n",
     "\n",
     "    # Plotting the overall history and evaluating the final model\n",
     "    plot_history(losses, val_losses, accuracies, val_accuracies)\n",
@@ -682,4 +681,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 0
-}
+}