These should be fixed now. Clear outputs.
PiperOrigin-RevId: 348060372
Change-Id: Idc312997eb39b3c7c5bdec046cccd19f8f7f6a73
diff --git a/tensorflow/lite/g3doc/performance/post_training_float16_quant.ipynb b/tensorflow/lite/g3doc/performance/post_training_float16_quant.ipynb
index 2ebaaaf..53c57d2 100644
--- a/tensorflow/lite/g3doc/performance/post_training_float16_quant.ipynb
+++ b/tensorflow/lite/g3doc/performance/post_training_float16_quant.ipynb
@@ -11,7 +11,7 @@
},
{
"cell_type": "code",
- "execution_count": 1,
+ "execution_count": null,
"metadata": {
"cellView": "form",
"id": "I9sUhVL_VZNO"
@@ -46,20 +46,20 @@
"id": "CGuqeuPSVNo-"
},
"source": [
- "\u003ctable class=\"tfo-notebook-buttons\" align=\"left\"\u003e\n",
- " \u003ctd\u003e\n",
- " \u003ca target=\"_blank\" href=\"https://www.tensorflow.org/lite/performance/post_training_float16_quant\"\u003e\u003cimg src=\"https://www.tensorflow.org/images/tf_logo_32px.png\" /\u003eView on TensorFlow.org\u003c/a\u003e\n",
- " \u003c/td\u003e\n",
- " \u003ctd\u003e\n",
- " \u003ca target=\"_blank\" href=\"https://colab.research.google.com/github/tensorflow/tensorflow/blob/master/tensorflow/lite/g3doc/performance/post_training_float16_quant.ipynb\"\u003e\u003cimg src=\"https://www.tensorflow.org/images/colab_logo_32px.png\" /\u003eRun in Google Colab\u003c/a\u003e\n",
- " \u003c/td\u003e\n",
- " \u003ctd\u003e\n",
- " \u003ca target=\"_blank\" href=\"https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/g3doc/performance/post_training_float16_quant.ipynb\"\u003e\u003cimg src=\"https://www.tensorflow.org/images/GitHub-Mark-32px.png\" /\u003eView source on GitHub\u003c/a\u003e\n",
- " \u003c/td\u003e\n",
- " \u003ctd\u003e\n",
- " \u003ca href=\"https://storage.googleapis.com/tensorflow_docs/tensorflow/lite/g3doc/performance/post_training_float16_quant.ipynb\"\u003e\u003cimg src=\"https://www.tensorflow.org/images/download_logo_32px.png\" /\u003eDownload notebook\u003c/a\u003e\n",
- " \u003c/td\u003e\n",
- "\u003c/table\u003e"
+ "<table class=\"tfo-notebook-buttons\" align=\"left\">\n",
+ " <td>\n",
+ " <a target=\"_blank\" href=\"https://www.tensorflow.org/lite/performance/post_training_float16_quant\"><img src=\"https://www.tensorflow.org/images/tf_logo_32px.png\" />View on TensorFlow.org</a>\n",
+ " </td>\n",
+ " <td>\n",
+ " <a target=\"_blank\" href=\"https://colab.research.google.com/github/tensorflow/tensorflow/blob/master/tensorflow/lite/g3doc/performance/post_training_float16_quant.ipynb\"><img src=\"https://www.tensorflow.org/images/colab_logo_32px.png\" />Run in Google Colab</a>\n",
+ " </td>\n",
+ " <td>\n",
+ " <a target=\"_blank\" href=\"https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/g3doc/performance/post_training_float16_quant.ipynb\"><img src=\"https://www.tensorflow.org/images/GitHub-Mark-32px.png\" />View source on GitHub</a>\n",
+ " </td>\n",
+ " <td>\n",
+ " <a href=\"https://storage.googleapis.com/tensorflow_docs/tensorflow/lite/g3doc/performance/post_training_float16_quant.ipynb\"><img src=\"https://www.tensorflow.org/images/download_logo_32px.png\" />Download notebook</a>\n",
+ " </td>\n",
+ "</table>"
]
},
{
@@ -97,7 +97,7 @@
},
{
"cell_type": "code",
- "execution_count": 2,
+ "execution_count": null,
"metadata": {
"id": "gyqAw1M9lyab"
},
@@ -114,24 +114,11 @@
},
{
"cell_type": "code",
- "execution_count": 3,
+ "execution_count": null,
"metadata": {
"id": "c6nb7OPlXs_3"
},
- "outputs": [
- {
- "data": {
- "text/plain": [
- "tf.float16"
- ]
- },
- "execution_count": 3,
- "metadata": {
- "tags": []
- },
- "output_type": "execute_result"
- }
- ],
+ "outputs": [],
"source": [
"tf.float16"
]
@@ -147,34 +134,11 @@
},
{
"cell_type": "code",
- "execution_count": 4,
+ "execution_count": null,
"metadata": {
"id": "hWSAjQWagIHl"
},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/mnist.npz\n",
- "11493376/11490434 [==============================] - 0s 0us/step\n",
- "11501568/11490434 [==============================] - 0s 0us/step\n",
- "1875/1875 [==============================] - 12s 6ms/step - loss: 0.2864 - accuracy: 0.9207 - val_loss: 0.1467 - val_accuracy: 0.9560\n"
- ]
- },
- {
- "data": {
- "text/plain": [
- "\u003ctensorflow.python.keras.callbacks.History at 0x7fcd75df46a0\u003e"
- ]
- },
- "execution_count": 4,
- "metadata": {
- "tags": []
- },
- "output_type": "execute_result"
- }
- ],
+ "outputs": [],
"source": [
"# Load MNIST dataset\n",
"mnist = keras.datasets.mnist\n",
@@ -230,7 +194,7 @@
},
{
"cell_type": "code",
- "execution_count": 5,
+ "execution_count": null,
"metadata": {
"id": "_i8B2nDZmAgQ"
},
@@ -251,7 +215,7 @@
},
{
"cell_type": "code",
- "execution_count": 6,
+ "execution_count": null,
"metadata": {
"id": "vptWZq2xnclo"
},
@@ -263,24 +227,11 @@
},
{
"cell_type": "code",
- "execution_count": 7,
+ "execution_count": null,
"metadata": {
"id": "Ie9pQaQrn5ue"
},
- "outputs": [
- {
- "data": {
- "text/plain": [
- "84452"
- ]
- },
- "execution_count": 7,
- "metadata": {
- "tags": []
- },
- "output_type": "execute_result"
- }
- ],
+ "outputs": [],
"source": [
"tflite_model_file = tflite_models_dir/\"mnist_model.tflite\"\n",
"tflite_model_file.write_bytes(tflite_model)"
@@ -297,7 +248,7 @@
},
{
"cell_type": "code",
- "execution_count": 8,
+ "execution_count": null,
"metadata": {
"id": "HEZ6ET1AHAS3"
},
@@ -318,24 +269,11 @@
},
{
"cell_type": "code",
- "execution_count": 9,
+ "execution_count": null,
"metadata": {
"id": "yuNfl3CoHNK3"
},
- "outputs": [
- {
- "data": {
- "text/plain": [
- "44272"
- ]
- },
- "execution_count": 9,
- "metadata": {
- "tags": []
- },
- "output_type": "execute_result"
- }
- ],
+ "outputs": [],
"source": [
"tflite_fp16_model = converter.convert()\n",
"tflite_model_fp16_file = tflite_models_dir/\"mnist_model_quant_f16.tflite\"\n",
@@ -353,21 +291,11 @@
},
{
"cell_type": "code",
- "execution_count": 10,
+ "execution_count": null,
"metadata": {
"id": "JExfcfLDscu4"
},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "total 128K\n",
- "-rw-rw-r-- 1 colaboratory-playground 50844828 44K Jun 23 06:04 mnist_model_quant_f16.tflite\n",
- "-rw-rw-r-- 1 colaboratory-playground 50844828 83K Jun 23 06:04 mnist_model.tflite\n"
- ]
- }
- ],
+ "outputs": [],
"source": [
"!ls -lh {tflite_models_dir}"
]
@@ -401,7 +329,7 @@
},
{
"cell_type": "code",
- "execution_count": 11,
+ "execution_count": null,
"metadata": {
"id": "Jn16Rc23zTss"
},
@@ -413,7 +341,7 @@
},
{
"cell_type": "code",
- "execution_count": 12,
+ "execution_count": null,
"metadata": {
"id": "J8Pztk1mvNVL"
},
@@ -434,7 +362,7 @@
},
{
"cell_type": "code",
- "execution_count": 13,
+ "execution_count": null,
"metadata": {
"id": "AKslvo2kwWac"
},
@@ -452,24 +380,11 @@
},
{
"cell_type": "code",
- "execution_count": 14,
+ "execution_count": null,
"metadata": {
"id": "XZClM2vo3_bm"
},
- "outputs": [
- {
- "data": {
- "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAEICAYAAACQ6CLfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nxfATSYrhJ5LU/wOdAGX9nfSgHgAAAABJRU5ErkJggg==\n",
- "text/plain": [
- "\u003cFigure size 600x400 with 1 Axes\u003e"
- ]
- },
- "metadata": {
- "tags": []
- },
- "output_type": "display_data"
- }
- ],
+ "outputs": [],
"source": [
"import matplotlib.pylab as plt\n",
"\n",
@@ -482,7 +397,7 @@
},
{
"cell_type": "code",
- "execution_count": 15,
+ "execution_count": null,
"metadata": {
"id": "3gwhv4lKbYZ4"
},
@@ -500,24 +415,11 @@
},
{
"cell_type": "code",
- "execution_count": 16,
+ "execution_count": null,
"metadata": {
"id": "CIH7G_MwbY2x"
},
- "outputs": [
- {
- "data": {
- "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAEICAYAAACQ6CLfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nxfATSYrhJ5LU/wOdAGX9nfSgHgAAAABJRU5ErkJggg==\n",
- "text/plain": [
- "\u003cFigure size 600x400 with 1 Axes\u003e"
- ]
- },
- "metadata": {
- "tags": []
- },
- "output_type": "display_data"
- }
- ],
+ "outputs": [],
"source": [
"plt.imshow(test_images[0])\n",
"template = \"True:{true}, predicted:{predict}\"\n",
@@ -537,7 +439,7 @@
},
{
"cell_type": "code",
- "execution_count": 17,
+ "execution_count": null,
"metadata": {
"id": "05aeAuWjvjPx"
},
@@ -577,19 +479,11 @@
},
{
"cell_type": "code",
- "execution_count": 18,
+ "execution_count": null,
"metadata": {
"id": "T5mWkSbMcU5z"
},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "0.956\n"
- ]
- }
- ],
+ "outputs": [],
"source": [
"print(evaluate_model(interpreter))"
]
@@ -605,19 +499,11 @@
},
{
"cell_type": "code",
- "execution_count": 19,
+ "execution_count": null,
"metadata": {
"id": "-9cnwiPp6EGm"
},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "0.956\n"
- ]
- }
- ],
+ "outputs": [],
"source": [
"# NOTE: Colab runs on server CPUs. At the time of writing this, TensorFlow Lite\n",
"# doesn't have super optimized server CPU kernels. For this reason this may be\n",
diff --git a/tensorflow/lite/g3doc/performance/post_training_integer_quant.ipynb b/tensorflow/lite/g3doc/performance/post_training_integer_quant.ipynb
index 21c7bd9b..b761387 100644
--- a/tensorflow/lite/g3doc/performance/post_training_integer_quant.ipynb
+++ b/tensorflow/lite/g3doc/performance/post_training_integer_quant.ipynb
@@ -46,20 +46,20 @@
"id": "CIGrZZPTZVeO"
},
"source": [
- "\u003ctable class=\"tfo-notebook-buttons\" align=\"left\"\u003e\n",
- " \u003ctd\u003e\n",
- " \u003ca target=\"_blank\" href=\"https://www.tensorflow.org/lite/performance/post_training_integer_quant\"\u003e\u003cimg src=\"https://www.tensorflow.org/images/tf_logo_32px.png\" /\u003eView on TensorFlow.org\u003c/a\u003e\n",
- " \u003c/td\u003e\n",
- " \u003ctd\u003e\n",
- " \u003ca target=\"_blank\" href=\"https://colab.research.google.com/github/tensorflow/tensorflow/blob/master/tensorflow/lite/g3doc/performance/post_training_integer_quant.ipynb\"\u003e\u003cimg src=\"https://www.tensorflow.org/images/colab_logo_32px.png\" /\u003eRun in Google Colab\u003c/a\u003e\n",
- " \u003c/td\u003e\n",
- " \u003ctd\u003e\n",
- " \u003ca target=\"_blank\" href=\"https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/g3doc/performance/post_training_integer_quant.ipynb\"\u003e\u003cimg src=\"https://www.tensorflow.org/images/GitHub-Mark-32px.png\" /\u003eView source on GitHub\u003c/a\u003e\n",
- " \u003c/td\u003e\n",
- " \u003ctd\u003e\n",
- " \u003ca href=\"https://storage.googleapis.com/tensorflow_docs/tensorflow/lite/g3doc/performance/post_training_integer_quant.ipynb\"\u003e\u003cimg src=\"https://www.tensorflow.org/images/download_logo_32px.png\" /\u003eDownload notebook\u003c/a\u003e\n",
- " \u003c/td\u003e\n",
- "\u003c/table\u003e"
+ "<table class=\"tfo-notebook-buttons\" align=\"left\">\n",
+ " <td>\n",
+ " <a target=\"_blank\" href=\"https://www.tensorflow.org/lite/performance/post_training_integer_quant\"><img src=\"https://www.tensorflow.org/images/tf_logo_32px.png\" />View on TensorFlow.org</a>\n",
+ " </td>\n",
+ " <td>\n",
+ " <a target=\"_blank\" href=\"https://colab.research.google.com/github/tensorflow/tensorflow/blob/master/tensorflow/lite/g3doc/performance/post_training_integer_quant.ipynb\"><img src=\"https://www.tensorflow.org/images/colab_logo_32px.png\" />Run in Google Colab</a>\n",
+ " </td>\n",
+ " <td>\n",
+ " <a target=\"_blank\" href=\"https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/g3doc/performance/post_training_integer_quant.ipynb\"><img src=\"https://www.tensorflow.org/images/GitHub-Mark-32px.png\" />View source on GitHub</a>\n",
+ " </td>\n",
+ " <td>\n",
+ " <a href=\"https://storage.googleapis.com/tensorflow_docs/tensorflow/lite/g3doc/performance/post_training_integer_quant.ipynb\"><img src=\"https://www.tensorflow.org/images/download_logo_32px.png\" />Download notebook</a>\n",
+ " </td>\n",
+ "</table>"
]
},
{
@@ -110,7 +110,7 @@
"\n",
"import tensorflow as tf\n",
"import numpy as np\n",
- "assert float(tf.__version__[:3]) \u003e= 2.3"
+ "assert float(tf.__version__[:3]) >= 2.3"
]
},
{
@@ -139,38 +139,7 @@
"metadata": {
"id": "eMsw_6HujaqM"
},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/mnist.npz\n",
- "11493376/11490434 [==============================] - 0s 0us/step\n",
- "Epoch 1/5\n",
- "1875/1875 [==============================] - 5s 2ms/step - loss: 0.2793 - accuracy: 0.9227 - val_loss: 0.1392 - val_accuracy: 0.9618\n",
- "Epoch 2/5\n",
- "1875/1875 [==============================] - 5s 2ms/step - loss: 0.1179 - accuracy: 0.9667 - val_loss: 0.0928 - val_accuracy: 0.9719\n",
- "Epoch 3/5\n",
- "1875/1875 [==============================] - 4s 2ms/step - loss: 0.0860 - accuracy: 0.9754 - val_loss: 0.0742 - val_accuracy: 0.9755\n",
- "Epoch 4/5\n",
- "1875/1875 [==============================] - 4s 2ms/step - loss: 0.0691 - accuracy: 0.9796 - val_loss: 0.0686 - val_accuracy: 0.9776\n",
- "Epoch 5/5\n",
- "1875/1875 [==============================] - 4s 2ms/step - loss: 0.0589 - accuracy: 0.9823 - val_loss: 0.0654 - val_accuracy: 0.9787\n"
- ]
- },
- {
- "data": {
- "text/plain": [
- "\u003ctensorflow.python.keras.callbacks.History at 0x7f69e0275a58\u003e"
- ]
- },
- "execution_count": null,
- "metadata": {
- "tags": []
- },
- "output_type": "execute_result"
- }
- ],
+ "outputs": [],
"source": [
"# Load MNIST dataset\n",
"mnist = tf.keras.datasets.mnist\n",
@@ -271,22 +240,7 @@
"metadata": {
"id": "HEZ6ET1AHAS3"
},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "INFO:tensorflow:Assets written to: /tmp/tmpcojyiqri/assets\n"
- ]
- },
- {
- "name": "stderr",
- "output_type": "stream",
- "text": [
- "INFO:tensorflow:Assets written to: /tmp/tmpcojyiqri/assets\n"
- ]
- }
- ],
+ "outputs": [],
"source": [
"converter = tf.lite.TFLiteConverter.from_keras_model(model)\n",
"converter.optimizations = [tf.lite.Optimize.DEFAULT]\n",
@@ -328,22 +282,7 @@
"metadata": {
"id": "FiwiWU3gHdkW"
},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "INFO:tensorflow:Assets written to: /tmp/tmp1bvfr71i/assets\n"
- ]
- },
- {
- "name": "stderr",
- "output_type": "stream",
- "text": [
- "INFO:tensorflow:Assets written to: /tmp/tmp1bvfr71i/assets\n"
- ]
- }
- ],
+ "outputs": [],
"source": [
"def representative_data_gen():\n",
" for input_value in tf.data.Dataset.from_tensor_slices(train_images).batch(1).take(100):\n",
@@ -374,16 +313,7 @@
"metadata": {
"id": "id1OEKFELQwp"
},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "input: \u003cclass 'numpy.float32'\u003e\n",
- "output: \u003cclass 'numpy.float32'\u003e\n"
- ]
- }
- ],
+ "outputs": [],
"source": [
"interpreter = tf.lite.Interpreter(model_content=tflite_model_quant)\n",
"input_type = interpreter.get_input_details()[0]['dtype']\n",
@@ -429,22 +359,7 @@
"metadata": {
"id": "kzjEjcDs3BHa"
},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "INFO:tensorflow:Assets written to: /tmp/tmpvnuxq9pa/assets\n"
- ]
- },
- {
- "name": "stderr",
- "output_type": "stream",
- "text": [
- "INFO:tensorflow:Assets written to: /tmp/tmpvnuxq9pa/assets\n"
- ]
- }
- ],
+ "outputs": [],
"source": [
"def representative_data_gen():\n",
" for input_value in tf.data.Dataset.from_tensor_slices(train_images).batch(1).take(100):\n",
@@ -477,16 +392,7 @@
"metadata": {
"id": "PaNkOS-twz4k"
},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "input: \u003cclass 'numpy.uint8'\u003e\n",
- "output: \u003cclass 'numpy.uint8'\u003e\n"
- ]
- }
- ],
+ "outputs": [],
"source": [
"interpreter = tf.lite.Interpreter(model_content=tflite_model_quant)\n",
"input_type = interpreter.get_input_details()[0]['dtype']\n",
@@ -528,20 +434,7 @@
"metadata": {
"id": "BEY59dC14uRv"
},
- "outputs": [
- {
- "data": {
- "text/plain": [
- "24720"
- ]
- },
- "execution_count": null,
- "metadata": {
- "tags": []
- },
- "output_type": "execute_result"
- }
- ],
+ "outputs": [],
"source": [
"import pathlib\n",
"\n",
@@ -677,21 +570,7 @@
"metadata": {
"id": "iTK0x980coto"
},
- "outputs": [
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- "\u003cFigure size 432x288 with 1 Axes\u003e"
- ]
- },
- "metadata": {
- "needs_background": "light",
- "tags": []
- },
- "output_type": "display_data"
- }
- ],
+ "outputs": [],
"source": [
"test_model(tflite_model_file, test_image_index, model_type=\"Float\")"
]
@@ -711,21 +590,7 @@
"metadata": {
"id": "rc1i9umMcp0t"
},
- "outputs": [
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- "\u003cFigure size 432x288 with 1 Axes\u003e"
- ]
- },
- "metadata": {
- "needs_background": "light",
- "tags": []
- },
- "output_type": "display_data"
- }
- ],
+ "outputs": [],
"source": [
"test_model(tflite_model_quant_file, test_image_index, model_type=\"Quantized\")"
]
@@ -785,15 +650,7 @@
"metadata": {
"id": "T5mWkSbMcU5z"
},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "Float model accuracy is 97.8700% (Number of test samples=10000)\n"
- ]
- }
- ],
+ "outputs": [],
"source": [
"evaluate_model(tflite_model_file, model_type=\"Float\")"
]
@@ -813,15 +670,7 @@
"metadata": {
"id": "-9cnwiPp6EGm"
},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "Quantized model accuracy is 97.8100% (Number of test samples=10000)\n"
- ]
- }
- ],
+ "outputs": [],
"source": [
"evaluate_model(tflite_model_quant_file, model_type=\"Quantized\")"
]
diff --git a/tensorflow/lite/g3doc/performance/post_training_quant.ipynb b/tensorflow/lite/g3doc/performance/post_training_quant.ipynb
index 311c818..0e58e05 100644
--- a/tensorflow/lite/g3doc/performance/post_training_quant.ipynb
+++ b/tensorflow/lite/g3doc/performance/post_training_quant.ipynb
@@ -11,7 +11,7 @@
},
{
"cell_type": "code",
- "execution_count": 1,
+ "execution_count": null,
"metadata": {
"cellView": "form",
"id": "R3yYtBPkM2qZ"
@@ -46,20 +46,20 @@
"id": "CIGrZZPTZVeO"
},
"source": [
- "\u003ctable class=\"tfo-notebook-buttons\" align=\"left\"\u003e\n",
- " \u003ctd\u003e\n",
- " \u003ca target=\"_blank\" href=\"https://www.tensorflow.org/lite/performance/post_training_quant\"\u003e\u003cimg src=\"https://www.tensorflow.org/images/tf_logo_32px.png\" /\u003eView on TensorFlow.org\u003c/a\u003e\n",
- " \u003c/td\u003e\n",
- " \u003ctd\u003e\n",
- " \u003ca target=\"_blank\" href=\"https://colab.research.google.com/github/tensorflow/tensorflow/blob/master/tensorflow/lite/g3doc/performance/post_training_quant.ipynb\"\u003e\u003cimg src=\"https://www.tensorflow.org/images/colab_logo_32px.png\" /\u003eRun in Google Colab\u003c/a\u003e\n",
- " \u003c/td\u003e\n",
- " \u003ctd\u003e\n",
- " \u003ca target=\"_blank\" href=\"https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/g3doc/performance/post_training_quant.ipynb\"\u003e\u003cimg src=\"https://www.tensorflow.org/images/GitHub-Mark-32px.png\" /\u003eView source on GitHub\u003c/a\u003e\n",
- " \u003c/td\u003e\n",
- " \u003ctd\u003e\n",
- " \u003ca href=\"https://storage.googleapis.com/tensorflow_docs/tensorflow/lite/g3doc/performance/post_training_quant.ipynb\"\u003e\u003cimg src=\"https://www.tensorflow.org/images/download_logo_32px.png\" /\u003eDownload notebook\u003c/a\u003e\n",
- " \u003c/td\u003e\n",
- "\u003c/table\u003e"
+ "<table class=\"tfo-notebook-buttons\" align=\"left\">\n",
+ " <td>\n",
+ " <a target=\"_blank\" href=\"https://www.tensorflow.org/lite/performance/post_training_quant\"><img src=\"https://www.tensorflow.org/images/tf_logo_32px.png\" />View on TensorFlow.org</a>\n",
+ " </td>\n",
+ " <td>\n",
+ " <a target=\"_blank\" href=\"https://colab.research.google.com/github/tensorflow/tensorflow/blob/master/tensorflow/lite/g3doc/performance/post_training_quant.ipynb\"><img src=\"https://www.tensorflow.org/images/colab_logo_32px.png\" />Run in Google Colab</a>\n",
+ " </td>\n",
+ " <td>\n",
+ " <a target=\"_blank\" href=\"https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/g3doc/performance/post_training_quant.ipynb\"><img src=\"https://www.tensorflow.org/images/GitHub-Mark-32px.png\" />View source on GitHub</a>\n",
+ " </td>\n",
+ " <td>\n",
+ " <a href=\"https://storage.googleapis.com/tensorflow_docs/tensorflow/lite/g3doc/performance/post_training_quant.ipynb\"><img src=\"https://www.tensorflow.org/images/download_logo_32px.png\" />Download notebook</a>\n",
+ " </td>\n",
+ "</table>"
]
},
{
@@ -118,7 +118,7 @@
},
{
"cell_type": "code",
- "execution_count": 2,
+ "execution_count": null,
"metadata": {
"id": "gyqAw1M9lyab"
},
@@ -144,31 +144,11 @@
},
{
"cell_type": "code",
- "execution_count": 3,
+ "execution_count": null,
"metadata": {
"id": "hWSAjQWagIHl"
},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "1875/1875 [==============================] - 10s 5ms/step - loss: 0.2787 - accuracy: 0.9203 - val_loss: 0.1323 - val_accuracy: 0.9624\n"
- ]
- },
- {
- "data": {
- "text/plain": [
- "\u003ctensorflow.python.keras.callbacks.History at 0x7f6443480e80\u003e"
- ]
- },
- "execution_count": 3,
- "metadata": {
- "tags": []
- },
- "output_type": "execute_result"
- }
- ],
+ "outputs": [],
"source": [
"# Load MNIST dataset\n",
"mnist = keras.datasets.mnist\n",
@@ -224,7 +204,7 @@
},
{
"cell_type": "code",
- "execution_count": 4,
+ "execution_count": null,
"metadata": {
"id": "_i8B2nDZmAgQ"
},
@@ -245,7 +225,7 @@
},
{
"cell_type": "code",
- "execution_count": 5,
+ "execution_count": null,
"metadata": {
"id": "vptWZq2xnclo"
},
@@ -257,24 +237,11 @@
},
{
"cell_type": "code",
- "execution_count": 6,
+ "execution_count": null,
"metadata": {
"id": "Ie9pQaQrn5ue"
},
- "outputs": [
- {
- "data": {
- "text/plain": [
- "84452"
- ]
- },
- "execution_count": 6,
- "metadata": {
- "tags": []
- },
- "output_type": "execute_result"
- }
- ],
+ "outputs": [],
"source": [
"tflite_model_file = tflite_models_dir/\"mnist_model.tflite\"\n",
"tflite_model_file.write_bytes(tflite_model)"
@@ -291,24 +258,11 @@
},
{
"cell_type": "code",
- "execution_count": 7,
+ "execution_count": null,
"metadata": {
"id": "g8PUvLWDlmmz"
},
- "outputs": [
- {
- "data": {
- "text/plain": [
- "23840"
- ]
- },
- "execution_count": 7,
- "metadata": {
- "tags": []
- },
- "output_type": "execute_result"
- }
- ],
+ "outputs": [],
"source": [
"converter.optimizations = [tf.lite.Optimize.DEFAULT]\n",
"tflite_quant_model = converter.convert()\n",
@@ -327,24 +281,11 @@
},
{
"cell_type": "code",
- "execution_count": 8,
+ "execution_count": null,
"metadata": {
"id": "JExfcfLDscu4"
},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "total 214M\n",
- "-rw-rw-r-- 1 colaboratory-playground 50844828 44K Jun 23 06:04 mnist_model_quant_f16.tflite\n",
- "-rw-rw-r-- 1 colaboratory-playground 50844828 24K Jun 23 06:12 mnist_model_quant.tflite\n",
- "-rw-rw-r-- 1 colaboratory-playground 50844828 83K Jun 23 06:12 mnist_model.tflite\n",
- "-rw-rw-r-- 1 colaboratory-playground 50844828 44M Jun 23 06:10 resnet_v2_101_quantized.tflite\n",
- "-rw-rw-r-- 1 colaboratory-playground 50844828 171M Jun 23 06:09 resnet_v2_101.tflite\n"
- ]
- }
- ],
+ "outputs": [],
"source": [
"!ls -lh {tflite_models_dir}"
]
@@ -372,7 +313,7 @@
},
{
"cell_type": "code",
- "execution_count": 9,
+ "execution_count": null,
"metadata": {
"id": "Jn16Rc23zTss"
},
@@ -384,7 +325,7 @@
},
{
"cell_type": "code",
- "execution_count": 10,
+ "execution_count": null,
"metadata": {
"id": "J8Pztk1mvNVL"
},
@@ -405,7 +346,7 @@
},
{
"cell_type": "code",
- "execution_count": 11,
+ "execution_count": null,
"metadata": {
"id": "AKslvo2kwWac"
},
@@ -423,24 +364,11 @@
},
{
"cell_type": "code",
- "execution_count": 12,
+ "execution_count": null,
"metadata": {
"id": "XZClM2vo3_bm"
},
- "outputs": [
- {
- "data": {
- "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAEICAYAAACQ6CLfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nxfATSYrhJ5LU/wOdAGX9nfSgHgAAAABJRU5ErkJggg==\n",
- "text/plain": [
- "\u003cFigure size 600x400 with 1 Axes\u003e"
- ]
- },
- "metadata": {
- "tags": []
- },
- "output_type": "display_data"
- }
- ],
+ "outputs": [],
"source": [
"import matplotlib.pylab as plt\n",
"\n",
@@ -462,7 +390,7 @@
},
{
"cell_type": "code",
- "execution_count": 13,
+ "execution_count": null,
"metadata": {
"id": "05aeAuWjvjPx"
},
@@ -502,19 +430,11 @@
},
{
"cell_type": "code",
- "execution_count": 14,
+ "execution_count": null,
"metadata": {
"id": "DqXBnDfJ7qxL"
},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "0.9624\n"
- ]
- }
- ],
+ "outputs": [],
"source": [
"print(evaluate_model(interpreter))"
]
@@ -530,19 +450,11 @@
},
{
"cell_type": "code",
- "execution_count": 15,
+ "execution_count": null,
"metadata": {
"id": "-9cnwiPp6EGm"
},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "0.9626\n"
- ]
- }
- ],
+ "outputs": [],
"source": [
"print(evaluate_model(interpreter_quant))"
]
@@ -573,7 +485,7 @@
},
{
"cell_type": "code",
- "execution_count": 16,
+ "execution_count": null,
"metadata": {
"id": "jrXZxSJiJfYN"
},
@@ -591,24 +503,11 @@
},
{
"cell_type": "code",
- "execution_count": 17,
+ "execution_count": null,
"metadata": {
"id": "LwnV4KxwVEoG"
},
- "outputs": [
- {
- "data": {
- "text/plain": [
- "178509092"
- ]
- },
- "execution_count": 17,
- "metadata": {
- "tags": []
- },
- "output_type": "execute_result"
- }
- ],
+ "outputs": [],
"source": [
"# Convert to TF Lite without quantization\n",
"resnet_tflite_file = tflite_models_dir/\"resnet_v2_101.tflite\"\n",
@@ -617,24 +516,11 @@
},
{
"cell_type": "code",
- "execution_count": 18,
+ "execution_count": null,
"metadata": {
"id": "2qkZD0VoVExe"
},
- "outputs": [
- {
- "data": {
- "text/plain": [
- "45182656"
- ]
- },
- "execution_count": 18,
- "metadata": {
- "tags": []
- },
- "output_type": "execute_result"
- }
- ],
+ "outputs": [],
"source": [
"# Convert to TF Lite with quantization\n",
"converter.optimizations = [tf.lite.Optimize.DEFAULT]\n",
@@ -644,23 +530,11 @@
},
{
"cell_type": "code",
- "execution_count": 19,
+ "execution_count": null,
"metadata": {
"id": "vhOjeg1x9Knp"
},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "-rw-rw-r-- 1 colaboratory-playground 50844828 44K Jun 23 06:04 /tmp/mnist_tflite_models/mnist_model_quant_f16.tflite\n",
- "-rw-rw-r-- 1 colaboratory-playground 50844828 24K Jun 23 06:12 /tmp/mnist_tflite_models/mnist_model_quant.tflite\n",
- "-rw-rw-r-- 1 colaboratory-playground 50844828 83K Jun 23 06:12 /tmp/mnist_tflite_models/mnist_model.tflite\n",
- "-rw-rw-r-- 1 colaboratory-playground 50844828 44M Jun 23 06:13 /tmp/mnist_tflite_models/resnet_v2_101_quantized.tflite\n",
- "-rw-rw-r-- 1 colaboratory-playground 50844828 171M Jun 23 06:12 /tmp/mnist_tflite_models/resnet_v2_101.tflite\n"
- ]
- }
- ],
+ "outputs": [],
"source": [
"!ls -lh {tflite_models_dir}/*.tflite"
]
