tflite_conv1d_conversion_bug.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "tflite_conv1d_conversion_bug.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Jzl22GNGbDt7"
      },
      "source": [
        "# TF Lite `Conv1D` conversion bug demo"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "L4RoUpQkjuk5"
      },
      "source": [
        "Conversion of `Conv1D` layer seems to have changed in TF Lite `2.4.0`. Minimal example below demonstrates interpreter `RESHAPE` error if we try to resize and allocate input."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PwQ5iEETanLq"
      },
      "source": [
        "## Prepare"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Dh5dvbl6arr5"
      },
      "source": [
        "use_tf_2_3 = False # Using TF Lite 2.3.0 gives no error\n",
        "\n",
        "if use_tf_2_3:\n",
        "    !pip install tensorflow==2.3.0\n",
        "else:\n",
        "    # 2.5.0 and 2.4.0 fail, too\n",
        "    !pip install tensorflow==2.6.0"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "l9vk-50PZkSj"
      },
      "source": [
        "import tensorflow as tf\n",
        "print(tf.__version__)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XVyhL41jaIFs"
      },
      "source": [
        "def generate_model(num_hidden_units: int = 500) -> tf.keras.models.Sequential:\n",
        "    model = tf.keras.models.Sequential()\n",
        "    model.add(tf.keras.layers.Conv1D(filters=num_hidden_units, kernel_size=3, strides=1, padding='SAME', activation='relu'))\n",
        "    optimizer = tf.keras.optimizers.Adam()\n",
        "    model.compile(optimizer=optimizer, loss='categorical_crossentropy', metrics=['accuracy'])\n",
        "    return model"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7WdyoJQ6aj3Q"
      },
      "source": [
        "## Create model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dh-uhJw6aKIU"
      },
      "source": [
        "model = generate_model()\n",
        "_ = model.predict(tf.random.normal((1, 100, 80)))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pCZubKlIafJb"
      },
      "source": [
        "## Conversion"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9sIGNMdraXW6"
      },
      "source": [
        "converter = tf.lite.TFLiteConverter.from_keras_model(model)\n",
        "converter.optimizations = [tf.lite.Optimize.DEFAULT]\n",
        "\n",
        "model_path = f\"model_{tf.__version__}.tflite\"\n",
        "with open(model_path, \"wb\") as f:\n",
        "    f.write(converter.convert())"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LdqKlzo0acqt"
      },
      "source": [
        "## Interpreter test"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qZQ9ja83cf5-"
      },
      "source": [
        "interpreter = tf.lite.Interpreter(model_path)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JF-k3FbTZV5A"
      },
      "source": [
        "interpreter.resize_tensor_input(0, [1, 50, 80])\n",
        "interpreter.allocate_tensors()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wjkVCwgYoz4u"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}