{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Tensorflow for poets\n",
"\n",
"This notebook steps through the [Tensorflow for poets](https://codelabs.developers.google.com/codelabs/tensorflow-for-poets/) tutorial.\n",
"\n",
"First clone the code repository."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"!git clone https://github.com/googlecodelabs/tensorflow-for-poets-2"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Move into the new directory."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"cd tensorflow-for-poets-2"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Download the flowers dataset."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"!curl http://download.tensorflow.org/example_images/flower_photos.tgz | tar xz -C tf_files"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ls tf_files/flower_photos"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Run this in a terminal, Jupyter doesn't allow background processes...\n",
"\n",
"I'm assuming this won't be possible on Binder?\n",
"\n",
"```\n",
"tensorboard --logdir tf_files/training_summaries &\n",
"```"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train the model"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%bash\n",
"IMAGE_SIZE=224\n",
"ARCHITECTURE=\"mobilenet_0.50_${IMAGE_SIZE}\"\n",
"\n",
"python -m scripts.retrain \\\n",
" --bottleneck_dir=tf_files/bottlenecks \\\n",
" --how_many_training_steps=500 \\\n",
" --model_dir=tf_files/models/ \\\n",
" --summaries_dir=tf_files/training_summaries/\"${ARCHITECTURE}\" \\\n",
" --output_graph=tf_files/retrained_graph.pb \\\n",
" --output_labels=tf_files/retrained_labels.txt \\\n",
" --architecture=\"${ARCHITECTURE}\" \\\n",
" --image_dir=tf_files/flower_photos"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Test the trained model"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%bash\n",
"python -m scripts.label_image \\\n",
" --graph=tf_files/retrained_graph.pb \\\n",
" --image=tf_files/flower_photos/daisy/21652746_cc379e0eea_m.jpg"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## More testing\n",
"\n",
"This bit isn't in the tutorial. I just thought it would be good to do some random testing..."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Make a list of all the flower images\n",
"import os\n",
"import random\n",
"from IPython.display import display, HTML\n",
"flowers = []\n",
"flower_dir = 'tf_files/flower_photos/'\n",
"for img_dir in [d for d in os.listdir(flower_dir) if os.path.isdir(os.path.join(flower_dir, d))]:\n",
" for img in [i for i in os.listdir(os.path.join(flower_dir, img_dir)) if i[-4:] == '.jpg']:\n",
" flowers.append(os.path.join(flower_dir, img_dir, img)) "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Choose one flower at random\n",
"flower = random.sample(flowers, 1)[0]\n",
"display(HTML('
{0}'.format(flower)))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"!python -m scripts.label_image --graph=tf_files/retrained_graph.pb --image=$flower"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"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": "3.6.5"
}
},
"nbformat": 4,
"nbformat_minor": 2
}