eileen-code4fun

from tensorflow.keras import layers, models, losses
import argparse

parser = argparse.ArgumentParser()
parser.add_argument('--epochs', dest='epochs', type=int, default=5)
parser.add_argument('--dropout_rate', dest='dropout_rate', type=float, default=0.1)

args = parser.parse_args()

def create_model():

eileen-code4fun

import tensorflow as tf
import hypertune

hpt = hypertune.HyperTune()

class CustomCallback(tf.keras.callbacks.Callback):
  def on_epoch_end(self, epoch, logs=None):
    hpt.report_hyperparameter_tuning_metric(
      hyperparameter_metric_tag='val_accuracy',
      metric_value=logs['val_accuracy'],

eileen-code4fun

studySpec:
  metrics:
  # Correspond to the metrics we use the hypertune library to report.
  - metricId: val_accuracy
    goal: MAXIMIZE
  parameters:
  # Correspond to the command line argument our Python code expects.
  - parameterId: dropout_rate
    doubleValueSpec:
      minValue: 0.01

eileen-code4fun

import os

tensorboard_callback = tf.keras.callbacks.TensorBoard(
  log_dir=os.environ['AIP_TENSORBOARD_LOG_DIR'],
  histogram_freq=1
)

model.fit(train_dataset, epochs=args.epochs, validation_data=val_dataset, callbacks=[tensorboard_callback])

eileen-code4fun

{
  "displayName": "e2e-tutorial-viz",
  "jobSpec": {
    "workerPoolSpecs": [
      {
        "replicaCount": 1,
        "machineSpec": {
          "machineType": "n1-standard-4",
          "acceleratorType": "NVIDIA_TESLA_V100",
          "acceleratorCount": 2

eileen-code4fun

import tensorflow as tf
from tensorflow.keras import datasets
import matplotlib.pyplot as plt

_, (test_images, test_labels) = datasets.cifar10.load_data()

class_names = ['airplane', 'automobile', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'ship', 'truck']
plt.xticks([])
plt.yticks([])
plt.title(class_names[test_labels[0][0]])

eileen-code4fun

import json

data = {
    "instances": [
        (test_images[0]/255.0).tolist()
    ]
}
print(json.dumps(data),  file=open('prediction_request.json', 'w'))

eileen-code4fun

{
  "inputs": {
    "image": {
      "inputTensorName": "conv2d_input",
      "modality": "image"
    }
  },
  "outputs": {
    "explanation": {
      "outputTensorName": "dense_1"

eileen-code4fun

import base64
import io
attr_map = tf.io.decode_jpeg(base64.b64decode(BASE64_JPEG_DATA))

plt.subplot(1, 2, 1)
plt.xticks([])
plt.yticks([])
plt.title(class_names[test_labels[0][0]])
plt.imshow(test_images[0])
plt.xticks([])

eileen-code4fun

import tensorflow as tf
import tensorflow_datasets as tfds

cifar10 = tfds.load('cifar10', as_supervised=True)
class_names = ['airplane', 'automobile', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'ship', 'truck']

def save_images_write_metadata(dataset, purpose, mf):
  i = 1
  for image, label in dataset:
    fn = '{}_{}.jpeg'.format(purpose, i)