{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "(tut_two)=\n",
    "\n",
    "# **Partial Network Quantization**\n",
    "\n",
    "This example shows how the NVIDIA TensorFlow 2.x Quantization Toolkit can be used to quantize only a few layers in a\n",
    "TensorFlow 2.x model.\n",
    "\n",
    "```{eval-rst}\n",
    "\n",
    ".. admonition:: Goal\n",
    "    :class: note\n",
    "\n",
    "    #. Take a resnet-like model and train on cifar10 dataset.\n",
    "    #. Quantize only layers named 'conv2d_2' and 'dense' in the model.\n",
    "    #. Fine-tune to recover model accuracy.\n",
    "    #. Save both original and quantized model while performing ONNX conversion.\n",
    "\n",
    "```\n",
    "\n",
    "```{eval-rst}\n",
    "\n",
    ".. admonition:: Background\n",
    "    :class: tip\n",
    "\n",
    "    Few/specific layers to quantize are passed to ``quantize_model`` function as an object of ``QuantizationSpec`` class. \n",
    "    quantization mode is set to ``partial`` in ``quantize_model`` function. \n",
    "\n",
    "    Adding layers with single input to ``QuantizationSpec`` is rather simple. However, for multi-input layers, flexibility to quantize specific inputs is also provided.  \n",
    "    \n",
    "    For example, user wants to quantize layers with name ``conv2d_2`` and ``add``. \n",
    "\n",
    "    Default behavior of ``Add`` layer class is NOT to quantize any input. None of inputs to the ``Add`` class layer is quantized when following code snippet is used. \n",
    "    \n",
    "    .. code-block:: python\n",
    "\n",
    "        q_spec = QuantizationSpec()\n",
    "        layer_name = ['conv2d_2']\n",
    "        q_spec.add(name=layer_name, quantization_index=layer_quantization_index)\n",
    "\n",
    "        q_model = quantize_model(model, quantization_spec=q_spec)\n",
    "\n",
    "    \n",
    "    However, when layer of ``Add`` class is passed via ``QuantizationSpec`` object, all inputs are quantized.  \n",
    "\n",
    "    .. code-block:: python\n",
    "\n",
    "        q_spec = QuantizationSpec()\n",
    "        layer_name = ['conv2d_2', 'add']\n",
    "        q_spec.add(name=layer_name, quantization_index=layer_quantization_index)\n",
    "\n",
    "        q_model = quantize_model(model, quantization_spec=q_spec)\n",
    "\n",
    "    \n",
    "    Code to quantize input at specific index (in this case, 1) for layer ``add`` could look as follows.\n",
    "\n",
    "    .. code-block:: python\n",
    "\n",
    "        q_spec = QuantizationSpec()\n",
    "        layer_name = ['conv2d_2', 'add']\n",
    "        layer_quantization_index = [None, [1]]\n",
    "        q_spec.add(name=layer_name, quantization_index=layer_quantization_index)\n",
    "\n",
    "        q_model = quantize_model(model, quantization_spec=q_spec)\n",
    "\n",
    "    Layer name can be found from the putput of ``model.summary()`` function for Functional and Sequetial models.\n",
    "    For subclassed model, use ``KerasModelTravller`` class from tensorflow_quantization.utils.\n",
    "\n",
    "```\n",
    "Refer [Python API](qmodel_api) documentation for more details.\n",
    "\n",
    "---\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "\n",
    "#\n",
    "# SPDX-FileCopyrightText: Copyright (c) 1993-2022 NVIDIA CORPORATION & AFFILIATES. All rights reserved.\n",
    "# SPDX-License-Identifier: Apache-2.0\n",
    "#\n",
    "# Licensed under the Apache License, Version 2.0 (the \"License\");\n",
    "# you may not use this file except in compliance with the License.\n",
    "# You may obtain a copy of the License at\n",
    "#\n",
    "# http://www.apache.org/licenses/LICENSE-2.0\n",
    "#\n",
    "# Unless required by applicable law or agreed to in writing, software\n",
    "# distributed under the License is distributed on an \"AS IS\" BASIS,\n",
    "# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n",
    "# See the License for the specific language governing permissions and\n",
    "# limitations under the License.\n",
    "#\n",
    "\n",
    "import tensorflow as tf\n",
    "from tensorflow_quantization import quantize_model, QuantizationSpec\n",
    "from tensorflow_quantization.custom_qdq_cases import ResNetV1QDQCase\n",
    "import tiny_resnet\n",
    "import os\n",
    "from tensorflow_quantization import utils\n",
    "\n",
    "tf.keras.backend.clear_session()\n",
    "\n",
    "# Create folders to save TF and ONNX models\n",
    "assets = utils.CreateAssetsFolders(os.path.join(os.getcwd(), \"tutorials\"))\n",
    "assets.add_folder(\"simple_network_quantize_partial\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "# Load CIFAR10 dataset\n",
    "cifar10 = tf.keras.datasets.cifar10\n",
    "(train_images, train_labels), (test_images, test_labels) = cifar10.load_data()\n",
    "\n",
    "# Normalize the input image so that each pixel value is between 0 and 1.\n",
    "train_images = train_images / 255.0\n",
    "test_images = test_images / 255.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nn_model_original = tiny_resnet.model()\n",
    "tf.keras.utils.plot_model(nn_model_original, to_file = assets.simple_network_quantize_partial.fp32 + \"/model.png\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "1407/1407 [==============================] - 18s 10ms/step - loss: 1.7617 - accuracy: 0.3615 - val_loss: 1.5624 - val_accuracy: 0.4300\n",
      "Epoch 2/10\n",
      "1407/1407 [==============================] - 14s 10ms/step - loss: 1.4876 - accuracy: 0.4645 - val_loss: 1.4242 - val_accuracy: 0.4762\n",
      "Epoch 3/10\n",
      "1407/1407 [==============================] - 14s 10ms/step - loss: 1.3737 - accuracy: 0.5092 - val_loss: 1.3406 - val_accuracy: 0.5202\n",
      "Epoch 4/10\n",
      "1407/1407 [==============================] - 14s 10ms/step - loss: 1.2952 - accuracy: 0.5396 - val_loss: 1.2768 - val_accuracy: 0.5398\n",
      "Epoch 5/10\n",
      "1407/1407 [==============================] - 14s 10ms/step - loss: 1.2370 - accuracy: 0.5649 - val_loss: 1.2466 - val_accuracy: 0.5560\n",
      "Epoch 6/10\n",
      "1407/1407 [==============================] - 14s 10ms/step - loss: 1.1857 - accuracy: 0.5812 - val_loss: 1.2052 - val_accuracy: 0.5718\n",
      "Epoch 7/10\n",
      "1407/1407 [==============================] - 14s 10ms/step - loss: 1.1442 - accuracy: 0.5972 - val_loss: 1.1836 - val_accuracy: 0.5786\n",
      "Epoch 8/10\n",
      "1407/1407 [==============================] - 14s 10ms/step - loss: 1.1079 - accuracy: 0.6091 - val_loss: 1.1356 - val_accuracy: 0.5978\n",
      "Epoch 9/10\n",
      "1407/1407 [==============================] - 14s 10ms/step - loss: 1.0702 - accuracy: 0.6220 - val_loss: 1.1244 - val_accuracy: 0.5940\n",
      "Epoch 10/10\n",
      "1407/1407 [==============================] - 14s 10ms/step - loss: 1.0354 - accuracy: 0.6368 - val_loss: 1.1019 - val_accuracy: 0.6108\n",
      "Baseline FP32 model test accuracy: 61.46\n"
     ]
    }
   ],
   "source": [
    "# Train original classification model\n",
    "nn_model_original.compile(\n",
    "    optimizer=tiny_resnet.optimizer(lr=1e-4),\n",
    "    loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),\n",
    "    metrics=[\"accuracy\"],\n",
    ")\n",
    "\n",
    "nn_model_original.fit(\n",
    "    train_images, train_labels, batch_size=32, epochs=10, validation_split=0.1\n",
    ")\n",
    "\n",
    "# Get baseline model accuracy\n",
    "_, baseline_model_accuracy = nn_model_original.evaluate(\n",
    "    test_images, test_labels, verbose=0\n",
    ")\n",
    "baseline_model_accuracy = round(100 * baseline_model_accuracy, 2)\n",
    "print(\"Baseline FP32 model test accuracy:\", baseline_model_accuracy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: /home/sagar/nvidia/2021/Customers/Waymo/QAT/tensorrt_qat/docs/source/notebooks/tutorials/simple_network_quantize_partial/fp32/saved_model/assets\n",
      "WARNING:tensorflow:From /home/sagar/miniconda3/lib/python3.8/site-packages/tf2onnx/tf_loader.py:711: extract_sub_graph (from tensorflow.python.framework.graph_util_impl) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use `tf.compat.v1.graph_util.extract_sub_graph`\n",
      "ONNX conversion Done!\n"
     ]
    }
   ],
   "source": [
    "# save TF FP32 original model\n",
    "tf.keras.models.save_model(nn_model_original, assets.simple_network_quantize_partial.fp32_saved_model)\n",
    "\n",
    "# Convert FP32 model to ONNX\n",
    "utils.convert_saved_model_to_onnx(saved_model_dir = assets.simple_network_quantize_partial.fp32_saved_model, onnx_model_path = assets.simple_network_quantize_partial.fp32_onnx_model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[I] Layer `conv2d` is not quantized. Partial quantization is enabled and layer name is not in user provided QuantizationSpec class object\n",
      "[I] Layer `re_lu` is not quantized. Partial quantization is enabled and layer name is not in user provided QuantizationSpec class object\n",
      "[I] Layer `conv2d_1` is not quantized. Partial quantization is enabled and layer name is not in user provided QuantizationSpec class object\n",
      "[I] Layer `re_lu_1` is not quantized. Partial quantization is enabled and layer name is not in user provided QuantizationSpec class object\n",
      "[I] Layer `re_lu_2` is not quantized. Partial quantization is enabled and layer name is not in user provided QuantizationSpec class object\n",
      "[I] Layer `conv2d_3` is not quantized. Partial quantization is enabled and layer name is not in user provided QuantizationSpec class object\n",
      "[I] Layer `re_lu_3` is not quantized. Partial quantization is enabled and layer name is not in user provided QuantizationSpec class object\n",
      "[I] Layer `conv2d_4` is not quantized. Partial quantization is enabled and layer name is not in user provided QuantizationSpec class object\n",
      "[I] Layer `re_lu_4` is not quantized. Partial quantization is enabled and layer name is not in user provided QuantizationSpec class object\n",
      "[I] Layer `conv2d_5` is not quantized. Partial quantization is enabled and layer name is not in user provided QuantizationSpec class object\n",
      "[I] Layer `conv2d_6` is not quantized. Partial quantization is enabled and layer name is not in user provided QuantizationSpec class object\n",
      "[I] Layer `re_lu_5` is not quantized. Partial quantization is enabled and layer name is not in user provided QuantizationSpec class object\n",
      "[I] Layer `flatten` is not quantized. Partial quantization is enabled and layer name is not in user provided QuantizationSpec class object\n",
      "[I] Layer `re_lu_6` is not quantized. Partial quantization is enabled and layer name is not in user provided QuantizationSpec class object\n",
      "[I] Layer `dense_1` is not quantized. Partial quantization is enabled and layer name is not in user provided QuantizationSpec class object\n"
     ]
    }
   ],
   "source": [
    "# Quantize model\n",
    "# 1.1 Create a dictionary to quantize only two layers named 'conv2d_2' and 'dense'\n",
    "qspec = QuantizationSpec()\n",
    "layer_name = ['conv2d_2', 'dense']\n",
    "qspec.add(name=layer_name)\n",
    "# 1.2 Call quantize model function\n",
    "q_nn_model = quantize_model(\n",
    "    model=nn_model_original, quantization_mode=\"partial\", quantization_spec=qspec)\n",
    "\n",
    "q_nn_model.compile(\n",
    "    optimizer=tiny_resnet.optimizer(lr=1e-4),\n",
    "    loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),\n",
    "    metrics=[\"accuracy\"],\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tf.keras.utils.plot_model(q_nn_model, to_file = assets.simple_network_quantize_partial.int8 + \"/model.png\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test accuracy immediately after quantization:58.96, diff:2.5\n"
     ]
    }
   ],
   "source": [
    "_, q_model_accuracy = q_nn_model.evaluate(test_images, test_labels, verbose=0)\n",
    "q_model_accuracy = round(100 * q_model_accuracy, 2)\n",
    "print(\n",
    "    \"Test accuracy immediately after quantization:{}, diff:{}\".format(\n",
    "        q_model_accuracy, (baseline_model_accuracy - q_model_accuracy)\n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/2\n",
      "1407/1407 [==============================] - 20s 14ms/step - loss: 1.0074 - accuracy: 0.6480 - val_loss: 1.0854 - val_accuracy: 0.6194\n",
      "Epoch 2/2\n",
      "1407/1407 [==============================] - 19s 14ms/step - loss: 0.9799 - accuracy: 0.6583 - val_loss: 1.0782 - val_accuracy: 0.6242\n",
      "Accuracy after fine tuning for 2 epochs :62.0\n"
     ]
    }
   ],
   "source": [
    "# Fine-tune quantized model\n",
    "fine_tune_epochs = 2\n",
    "q_nn_model.fit(\n",
    "    train_images,\n",
    "    train_labels,\n",
    "    batch_size=32,\n",
    "    epochs=fine_tune_epochs,\n",
    "    validation_split=0.1,\n",
    ")\n",
    "_, q_model_accuracy = q_nn_model.evaluate(test_images, test_labels, verbose=0)\n",
    "q_model_accuracy = round(100 * q_model_accuracy, 2)\n",
    "print(\n",
    "    \"Accuracy after fine tuning for {} epochs :{}\".format(\n",
    "        fine_tune_epochs, q_model_accuracy\n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Found untraced functions such as conv2d_2_layer_call_fn, conv2d_2_layer_call_and_return_conditional_losses, dense_layer_call_fn, dense_layer_call_and_return_conditional_losses while saving (showing 4 of 4). These functions will not be directly callable after loading.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: /home/sagar/nvidia/2021/Customers/Waymo/QAT/tensorrt_qat/docs/source/notebooks/tutorials/simple_network_quantize_partial/int8/saved_model/assets\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: /home/sagar/nvidia/2021/Customers/Waymo/QAT/tensorrt_qat/docs/source/notebooks/tutorials/simple_network_quantize_partial/int8/saved_model/assets\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ONNX conversion Done!\n"
     ]
    }
   ],
   "source": [
    "# Save TF INT8 original model\n",
    "tf.keras.models.save_model(q_nn_model, assets.simple_network_quantize_partial.int8_saved_model)\n",
    "\n",
    "# Convert INT8 model to ONNX\n",
    "utils.convert_saved_model_to_onnx(saved_model_dir = assets.simple_network_quantize_partial.int8_saved_model, onnx_model_path = assets.simple_network_quantize_partial.int8_onnx_model)\n",
    "\n",
    "tf.keras.backend.clear_session()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "```{note}\n",
    "ONNX files can be visualized with [Netron](https://netron.app/).\n",
    "```"
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "4442e1c252d743d7d1ab28567e302ebe8a15da81acb5d7e7894db75e10bdb29d"
  },
  "kernelspec": {
   "display_name": "Python 3.8.10 ('base')",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.10"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}