notebook.community

Edit and run

aaa

Based on https://github.com/damienpontifex/BlogCodeSamples/tree/master/DataToTfRecords





In [24]:



    


#! /usr/env/bin python3

"""Convert MNIST Dataset to local TFRecords"""

import argparse
import os
import sys
import numpy as np
import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data

def _data_path(data_directory:str, name:str) -> str:
    """Construct a full path to a TFRecord file to be stored in the 
    data_directory. Will also ensure the data directory exists
    
    Args:
        data_directory: The directory where the records will be stored
        name:           The name of the TFRecord
    
    Returns:
        The full path to the TFRecord file
    """
    if not os.path.isdir(data_directory):
        os.makedirs(data_directory)

    return os.path.join(data_directory, f'{name}.tfrecords')

def _int64_feature(value:int) -> tf.train.Features.FeatureEntry:
    """Create a Int64List Feature
    
    Args:
        value: The value to store in the feature
    
    Returns:
        The FeatureEntry
    """
    return tf.train.Feature(int64_list=tf.train.Int64List(value=[value]))

def _bytes_feature(value:str) -> tf.train.Features.FeatureEntry:
    """Create a BytesList Feature
    
    Args:
        value: The value to store in the feature
    
    Returns:
        The FeatureEntry
    """
    return tf.train.Feature(bytes_list=tf.train.BytesList(value=[value]))

def convert_to_tf_record(images, labels, name:str, data_directory:str, num_shards:int=1):
    """Convert the dataset into TFRecords on disk
    
    Args:
        images:         The MNIST images
        labels:         The MNIST labels
        name:           The name of the data set
        data_directory: The directory where records will be stored
        num_shards:     The number of files on disk to separate records into
    """
    print(f'\nProcessing {name} data')

    
    num_examples, rows, cols, depth = images.shape

    def _process_examples(start_idx:int, end_index:int, filename:str):
        with tf.python_io.TFRecordWriter(filename) as writer:
            for index in range(start_idx, end_index):
                sys.stdout.write(f"\rProcessing sample {index+1} of {num_examples}")
                sys.stdout.flush()

                image_raw = images[index].tostring()
                example = tf.train.Example(features=tf.train.Features(feature={
                    'height': _int64_feature(rows),
                    'width': _int64_feature(cols),
                    'depth': _int64_feature(depth),
                    'label': _int64_feature(int(labels[index])),
                    'image_raw': _bytes_feature(image_raw)
                }))
                writer.write(example.SerializeToString())
    
    if num_shards == 1:
        _process_examples(0, num_examples, _data_path(data_directory, name))
    else:
        total_examples = num_examples
        samples_per_shard = total_examples // num_shards

        for shard in range(num_shards):
            start_index = shard * samples_per_shard
            end_index = start_index + samples_per_shard
            _process_examples(start_index, end_index, _data_path(data_directory, f'{name}-{shard+1}'))



    











In [26]:



    


# Load MNIST data
(X_train, y_train), (X_tst, y_tst) = tf.keras.datasets.mnist.load_data(path='mnist.npz')
X_train = np.reshape(X_train, X_train.shape + (1,))
X_tst  = np.reshape(X_tst,  X_tst.shape + (1,))
X_train.shape


# Separate train and valid datasets
from sklearn.model_selection import train_test_split
X_trn, X_val, y_trn, y_val = train_test_split(X_train, y_train, train_size=0.9, random_state=42)



    


















    






/Users/jorge/anaconda3/envs/tm/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2026: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.
  FutureWarning)

















In [27]:



    


data_directory ='/tmp/mnist/data'
    
convert_to_tf_record(X_trn, y_trn, 'train', data_directory, num_shards=10)
convert_to_tf_record(X_val, y_val, 'validation', data_directory)
convert_to_tf_record(X_tst,y_tst, 'test', data_directory)



    


















    






Processing train data
Processing sample 54000 of 54000Processing validation data
Processing sample 6000 of 6000Processing test data
Processing sample 10000 of 10000

Model





In [ ]:



    






    











In [ ]:



    






    











In [ ]:



    






    











In [ ]:



    






    











In [ ]:



    






    











In [ ]:



    






    











In [3]:



    


#!/usr/bin/env python3

#from argparse import ArgumentParser
import os
import glob
import tensorflow as tf

def mnist_model(features, mode, params):
    is_training = mode == tf.estimator.ModeKeys.TRAIN
    
    with tf.name_scope('Input'):
        # Input Layer
        input_layer = tf.reshape(features, [-1, 28, 28, 1], name='input_reshape')
        tf.summary.image('input', input_layer)

    with tf.name_scope('Conv_1'):
        # Convolutional Layer #1
        conv1 = tf.layers.conv2d(
          inputs=input_layer,
          filters=32,
          kernel_size=(5, 5),
          padding='same',
          activation=tf.nn.relu,
          trainable=is_training)

        # Pooling Layer #1
        pool1 = tf.layers.max_pooling2d(inputs=conv1, pool_size=(2, 2), strides=2, padding='same')

    with tf.name_scope('Conv_2'):
        # Convolutional Layer #2 and Pooling Layer #2
        conv2 = tf.layers.conv2d(
            inputs=pool1,
            filters=64,
            kernel_size=(5, 5),
            padding='same',
            activation=tf.nn.relu,
            trainable=is_training)
        
        pool2 = tf.layers.max_pooling2d(inputs=conv2, pool_size=(2, 2), strides=2, padding='same')

    with tf.name_scope('Dense_Dropout'):
        # Dense Layer
        # pool2_flat = tf.reshape(pool2, [-1, 7 * 7 * 64])
        pool2_flat = tf.contrib.layers.flatten(pool2)
        dense = tf.layers.dense(inputs=pool2_flat, units=1_024, activation=tf.nn.relu, trainable=is_training)
        dropout = tf.layers.dropout(inputs=dense, rate=params.dropout_rate, training=is_training)

    with tf.name_scope('Predictions'):
        # Logits Layer
        logits = tf.layers.dense(inputs=dropout, units=10, trainable=is_training)

        return logits

def cnn_model_fn(features, labels, mode, params):
    """Model function for CNN."""

    logits = mnist_model(features, mode, params)
    predicted_logit = tf.argmax(input=logits, axis=1, output_type=tf.int32)
    scores = tf.nn.softmax(logits, name='softmax_tensor')
    # Generate Predictions
    predictions = {
      'classes': predicted_logit,
      'probabilities': scores
    }

    export_outputs = {
        'prediction': tf.estimator.export.ClassificationOutput(
            scores=scores,
            classes=tf.cast(predicted_logit, tf.string))
    }

    # PREDICT
    if mode == tf.estimator.ModeKeys.PREDICT:
        return tf.estimator.EstimatorSpec(mode=mode, predictions=predictions, export_outputs=export_outputs)

    # TRAIN and EVAL
    loss = tf.losses.softmax_cross_entropy(onehot_labels=labels, logits=logits)

    accuracy = tf.metrics.accuracy(tf.argmax(labels, axis=1), predicted_logit)
    eval_metric = { 'accuracy': accuracy }

    # Configure the Training Op (for TRAIN mode)
    if mode == tf.estimator.ModeKeys.TRAIN:
        tf.summary.scalar('accuracy', accuracy[0])
        train_op = tf.contrib.layers.optimize_loss(
            loss=loss,
            global_step=tf.contrib.framework.get_global_step(),
            learning_rate=params.learning_rate,
            optimizer='Adam')
    else:
        train_op = None

    return tf.estimator.EstimatorSpec(
        mode=mode,
        loss=loss,
        train_op=train_op,
        eval_metric_ops=eval_metric,
        predictions=predictions,
        export_outputs=export_outputs)

def data_input_fn(filenames, batch_size=1000, shuffle=False):
    
    def _parser(record):
        features={
            'label': tf.FixedLenFeature([], tf.int64),
            'image_raw': tf.FixedLenFeature([], tf.string)
        }
        parsed_record = tf.parse_single_example(record, features)
        image = tf.decode_raw(parsed_record['image_raw'], tf.float32)

        label = tf.cast(parsed_record['label'], tf.int32)

        return image, tf.one_hot(label, depth=10)
        
    def _input_fn():
        dataset = (tf.data.TFRecordDataset(filenames)
            .map(_parser))
        if shuffle:
            dataset = dataset.shuffle(buffer_size=10_000)

        dataset = dataset.repeat(None) # Infinite iterations: let experiment determine num_epochs
        dataset = dataset.batch(batch_size)
        
        iterator = dataset.make_one_shot_iterator()
        features, labels = iterator.get_next()
        
        return features, labels
    return _input_fn



    











In [4]:



    


model_directory = '/tmp/mnist/models'
data_directory = '/tmp/mnist/data'
if 1:



    tf.logging.set_verbosity(tf.logging.INFO)

    run_config = tf.contrib.learn.RunConfig(
        model_dir=model_directory, 
        save_checkpoints_steps=20, 
        save_summary_steps=20)

    hparams = tf.contrib.training.HParams(
        learning_rate=1e-3, 
        dropout_rate=0.4,
        data_directory=os.path.expanduser(data_directory))

    mnist_classifier = tf.estimator.Estimator(
        model_fn=cnn_model_fn, 
        config=run_config,
        params=hparams
    )

    train_batch_size = 128
    train_steps = 55000 // train_batch_size # len dataset // batch size
    train_input_fn = data_input_fn(glob.glob(os.path.join(hparams.data_directory, 'train-*.tfrecords')), batch_size=train_batch_size)
    eval_input_fn = data_input_fn(os.path.join(hparams.data_directory, 'validation.tfrecords'), batch_size=100)
    
    experiment = tf.contrib.learn.Experiment(
        mnist_classifier,
        train_input_fn=train_input_fn,
        eval_input_fn=eval_input_fn,
        train_steps=train_steps
    )

    experiment.train_and_evaluate()



    


















    






WARNING:tensorflow:From <ipython-input-4-78698ff130ce>:12: RunConfig.__init__ (from tensorflow.contrib.learn.python.learn.estimators.run_config) is deprecated and will be removed in a future version.
Instructions for updating:
When switching to tf.estimator.Estimator, use tf.estimator.RunConfig instead.
INFO:tensorflow:Using config: {'_task_type': None, '_task_id': 0, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x1319de0b8>, '_master': '', '_num_ps_replicas': 0, '_num_worker_replicas': 0, '_environment': 'local', '_is_chief': True, '_evaluation_master': '', '_tf_config': gpu_options {
  per_process_gpu_memory_fraction: 1.0
}
, '_tf_random_seed': None, '_save_summary_steps': 20, '_save_checkpoints_secs': None, '_log_step_count_steps': 100, '_session_config': None, '_save_checkpoints_steps': 20, '_keep_checkpoint_max': 5, '_keep_checkpoint_every_n_hours': 10000, '_model_dir': '/tmp/mnist/models'}
WARNING:tensorflow:From <ipython-input-4-78698ff130ce>:34: Experiment.__init__ (from tensorflow.contrib.learn.python.learn.experiment) is deprecated and will be removed in a future version.
Instructions for updating:
Please switch to tf.estimator.train_and_evaluate. You will also have to convert to a tf.estimator.Estimator.
WARNING:tensorflow:From /Users/jorge/anaconda3/envs/tm/lib/python3.6/site-packages/tensorflow/contrib/learn/python/learn/monitors.py:279: BaseMonitor.__init__ (from tensorflow.contrib.learn.python.learn.monitors) is deprecated and will be removed after 2016-12-05.
Instructions for updating:
Monitors are deprecated. Please use tf.train.SessionRunHook.
INFO:tensorflow:Calling model_fn.
WARNING:tensorflow:From <ipython-input-3-a3f528fa6b75>:87: get_global_step (from tensorflow.contrib.framework.python.ops.variables) is deprecated and will be removed in a future version.
Instructions for updating:
Please switch to tf.train.get_global_step
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Create CheckpointSaverHook.
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Saving checkpoints for 1 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:Calling model_fn.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Starting evaluation at 2018-05-11-09:41:14
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/mnist/models/model.ckpt-1
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Evaluation [10/100]
INFO:tensorflow:Evaluation [20/100]
INFO:tensorflow:Evaluation [30/100]
INFO:tensorflow:Evaluation [40/100]
INFO:tensorflow:Evaluation [50/100]
INFO:tensorflow:Evaluation [60/100]
INFO:tensorflow:Evaluation [70/100]
INFO:tensorflow:Evaluation [80/100]
INFO:tensorflow:Evaluation [90/100]
INFO:tensorflow:Evaluation [100/100]
INFO:tensorflow:Finished evaluation at 2018-05-11-09:41:20
INFO:tensorflow:Saving dict for global step 1: accuracy = 0.3244, global_step = 1, loss = 2.191008
INFO:tensorflow:Validation (step 1): accuracy = 0.3244, loss = 2.191008, global_step = 1
INFO:tensorflow:loss = 2.2995539, step = 1
INFO:tensorflow:Saving checkpoints for 21 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:Calling model_fn.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Starting evaluation at 2018-05-11-09:41:25
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/mnist/models/model.ckpt-21
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Evaluation [10/100]
INFO:tensorflow:Evaluation [20/100]
INFO:tensorflow:Evaluation [30/100]
INFO:tensorflow:Evaluation [40/100]
INFO:tensorflow:Evaluation [50/100]
INFO:tensorflow:Evaluation [60/100]
INFO:tensorflow:Evaluation [70/100]
INFO:tensorflow:Evaluation [80/100]
INFO:tensorflow:Evaluation [90/100]
INFO:tensorflow:Evaluation [100/100]
INFO:tensorflow:Finished evaluation at 2018-05-11-09:41:31
INFO:tensorflow:Saving dict for global step 21: accuracy = 0.859, global_step = 21, loss = 0.44589746
INFO:tensorflow:Validation (step 28): accuracy = 0.859, loss = 0.44589746, global_step = 21
INFO:tensorflow:Saving checkpoints for 41 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:Calling model_fn.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Starting evaluation at 2018-05-11-09:41:36
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/mnist/models/model.ckpt-41
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Evaluation [10/100]
INFO:tensorflow:Evaluation [20/100]
INFO:tensorflow:Evaluation [30/100]
INFO:tensorflow:Evaluation [40/100]
INFO:tensorflow:Evaluation [50/100]
INFO:tensorflow:Evaluation [60/100]
INFO:tensorflow:Evaluation [70/100]
INFO:tensorflow:Evaluation [80/100]
INFO:tensorflow:Evaluation [90/100]
INFO:tensorflow:Evaluation [100/100]
INFO:tensorflow:Finished evaluation at 2018-05-11-09:41:42
INFO:tensorflow:Saving dict for global step 41: accuracy = 0.9114, global_step = 41, loss = 0.30477622
INFO:tensorflow:Validation (step 55): accuracy = 0.9114, loss = 0.30477622, global_step = 41
INFO:tensorflow:Saving checkpoints for 61 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:Saving checkpoints for 81 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:Calling model_fn.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Starting evaluation at 2018-05-11-09:41:47
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/mnist/models/model.ckpt-81
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Evaluation [10/100]
INFO:tensorflow:Evaluation [20/100]
INFO:tensorflow:Evaluation [30/100]
INFO:tensorflow:Evaluation [40/100]
INFO:tensorflow:Evaluation [50/100]
INFO:tensorflow:Evaluation [60/100]
INFO:tensorflow:Evaluation [70/100]
INFO:tensorflow:Evaluation [80/100]
INFO:tensorflow:Evaluation [90/100]
INFO:tensorflow:Evaluation [100/100]
INFO:tensorflow:Finished evaluation at 2018-05-11-09:41:52
INFO:tensorflow:Saving dict for global step 81: accuracy = 0.9516, global_step = 81, loss = 0.15887938
INFO:tensorflow:Validation (step 81): accuracy = 0.9516, loss = 0.15887938, global_step = 81
INFO:tensorflow:Saving checkpoints for 101 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:global_step/sec: 2.35954
INFO:tensorflow:loss = 0.25827867, step = 101 (36.479 sec)
INFO:tensorflow:Calling model_fn.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Starting evaluation at 2018-05-11-09:41:58
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/mnist/models/model.ckpt-101
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Evaluation [10/100]
INFO:tensorflow:Evaluation [20/100]
INFO:tensorflow:Evaluation [30/100]
INFO:tensorflow:Evaluation [40/100]
INFO:tensorflow:Evaluation [50/100]
INFO:tensorflow:Evaluation [60/100]
INFO:tensorflow:Evaluation [70/100]
INFO:tensorflow:Evaluation [80/100]
INFO:tensorflow:Evaluation [90/100]
INFO:tensorflow:Evaluation [100/100]
INFO:tensorflow:Finished evaluation at 2018-05-11-09:42:03
INFO:tensorflow:Saving dict for global step 101: accuracy = 0.959, global_step = 101, loss = 0.14267102
INFO:tensorflow:Validation (step 108): accuracy = 0.959, loss = 0.14267102, global_step = 101
INFO:tensorflow:Saving checkpoints for 121 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:Calling model_fn.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Starting evaluation at 2018-05-11-09:42:08
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/mnist/models/model.ckpt-121
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Evaluation [10/100]
INFO:tensorflow:Evaluation [20/100]
INFO:tensorflow:Evaluation [30/100]
INFO:tensorflow:Evaluation [40/100]
INFO:tensorflow:Evaluation [50/100]
INFO:tensorflow:Evaluation [60/100]
INFO:tensorflow:Evaluation [70/100]
INFO:tensorflow:Evaluation [80/100]
INFO:tensorflow:Evaluation [90/100]
INFO:tensorflow:Evaluation [100/100]
INFO:tensorflow:Finished evaluation at 2018-05-11-09:42:13
INFO:tensorflow:Saving dict for global step 121: accuracy = 0.9628, global_step = 121, loss = 0.13172294
INFO:tensorflow:Validation (step 139): accuracy = 0.9628, loss = 0.13172294, global_step = 121
INFO:tensorflow:Saving checkpoints for 141 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:Calling model_fn.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Starting evaluation at 2018-05-11-09:42:13
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/mnist/models/model.ckpt-141
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Evaluation [10/100]
INFO:tensorflow:Evaluation [20/100]
INFO:tensorflow:Evaluation [30/100]
INFO:tensorflow:Evaluation [40/100]
INFO:tensorflow:Evaluation [50/100]
INFO:tensorflow:Evaluation [60/100]
INFO:tensorflow:Evaluation [70/100]
INFO:tensorflow:Evaluation [80/100]
INFO:tensorflow:Evaluation [90/100]
INFO:tensorflow:Evaluation [100/100]
INFO:tensorflow:Finished evaluation at 2018-05-11-09:42:18
INFO:tensorflow:Saving dict for global step 141: accuracy = 0.9696, global_step = 141, loss = 0.106897295
INFO:tensorflow:Validation (step 141): accuracy = 0.9696, loss = 0.106897295, global_step = 141
INFO:tensorflow:Saving checkpoints for 161 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:Calling model_fn.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Starting evaluation at 2018-05-11-09:42:24
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/mnist/models/model.ckpt-161
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Evaluation [10/100]
INFO:tensorflow:Evaluation [20/100]
INFO:tensorflow:Evaluation [30/100]
INFO:tensorflow:Evaluation [40/100]
INFO:tensorflow:Evaluation [50/100]
INFO:tensorflow:Evaluation [60/100]
INFO:tensorflow:Evaluation [70/100]
INFO:tensorflow:Evaluation [80/100]
INFO:tensorflow:Evaluation [90/100]
INFO:tensorflow:Evaluation [100/100]
INFO:tensorflow:Finished evaluation at 2018-05-11-09:42:28
INFO:tensorflow:Saving dict for global step 161: accuracy = 0.9748, global_step = 161, loss = 0.08659887
INFO:tensorflow:Validation (step 173): accuracy = 0.9748, loss = 0.08659887, global_step = 161
INFO:tensorflow:Saving checkpoints for 181 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:Saving checkpoints for 201 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:global_step/sec: 2.71351
INFO:tensorflow:loss = 0.07934669, step = 201 (36.853 sec)
INFO:tensorflow:Calling model_fn.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Starting evaluation at 2018-05-11-09:42:34
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/mnist/models/model.ckpt-201
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Evaluation [10/100]
INFO:tensorflow:Evaluation [20/100]
INFO:tensorflow:Evaluation [30/100]
INFO:tensorflow:Evaluation [40/100]
INFO:tensorflow:Evaluation [50/100]
INFO:tensorflow:Evaluation [60/100]
INFO:tensorflow:Evaluation [70/100]
INFO:tensorflow:Evaluation [80/100]
INFO:tensorflow:Evaluation [90/100]
INFO:tensorflow:Evaluation [100/100]
INFO:tensorflow:Finished evaluation at 2018-05-11-09:42:38
INFO:tensorflow:Saving dict for global step 201: accuracy = 0.979, global_step = 201, loss = 0.07541576
INFO:tensorflow:Validation (step 204): accuracy = 0.979, loss = 0.07541576, global_step = 201
INFO:tensorflow:Saving checkpoints for 221 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:Calling model_fn.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Starting evaluation at 2018-05-11-09:42:44
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/mnist/models/model.ckpt-221
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Evaluation [10/100]
INFO:tensorflow:Evaluation [20/100]
INFO:tensorflow:Evaluation [30/100]
INFO:tensorflow:Evaluation [40/100]
INFO:tensorflow:Evaluation [50/100]
INFO:tensorflow:Evaluation [60/100]
INFO:tensorflow:Evaluation [70/100]
INFO:tensorflow:Evaluation [80/100]
INFO:tensorflow:Evaluation [90/100]
INFO:tensorflow:Evaluation [100/100]
INFO:tensorflow:Finished evaluation at 2018-05-11-09:42:48
INFO:tensorflow:Saving dict for global step 221: accuracy = 0.9792, global_step = 221, loss = 0.07064063
INFO:tensorflow:Validation (step 235): accuracy = 0.9792, loss = 0.07064063, global_step = 221
INFO:tensorflow:Saving checkpoints for 241 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:Saving checkpoints for 261 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:Calling model_fn.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Starting evaluation at 2018-05-11-09:42:54
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/mnist/models/model.ckpt-261
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Evaluation [10/100]
INFO:tensorflow:Evaluation [20/100]
INFO:tensorflow:Evaluation [30/100]
INFO:tensorflow:Evaluation [40/100]
INFO:tensorflow:Evaluation [50/100]
INFO:tensorflow:Evaluation [60/100]
INFO:tensorflow:Evaluation [70/100]
INFO:tensorflow:Evaluation [80/100]
INFO:tensorflow:Evaluation [90/100]
INFO:tensorflow:Evaluation [100/100]
INFO:tensorflow:Finished evaluation at 2018-05-11-09:42:59
INFO:tensorflow:Saving dict for global step 261: accuracy = 0.9802, global_step = 261, loss = 0.069362834
INFO:tensorflow:Validation (step 266): accuracy = 0.9802, loss = 0.069362834, global_step = 261
INFO:tensorflow:Saving checkpoints for 281 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:Calling model_fn.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Starting evaluation at 2018-05-11-09:43:04
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/mnist/models/model.ckpt-281
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Evaluation [10/100]
INFO:tensorflow:Evaluation [20/100]
INFO:tensorflow:Evaluation [30/100]
INFO:tensorflow:Evaluation [40/100]
INFO:tensorflow:Evaluation [50/100]
INFO:tensorflow:Evaluation [60/100]
INFO:tensorflow:Evaluation [70/100]
INFO:tensorflow:Evaluation [80/100]
INFO:tensorflow:Evaluation [90/100]
INFO:tensorflow:Evaluation [100/100]
INFO:tensorflow:Finished evaluation at 2018-05-11-09:43:09
INFO:tensorflow:Saving dict for global step 281: accuracy = 0.981, global_step = 281, loss = 0.0665182
INFO:tensorflow:Validation (step 298): accuracy = 0.981, loss = 0.0665182, global_step = 281
INFO:tensorflow:Saving checkpoints for 301 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:global_step/sec: 2.73436
INFO:tensorflow:loss = 0.10867074, step = 301 (36.571 sec)
INFO:tensorflow:Saving checkpoints for 321 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:Calling model_fn.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Starting evaluation at 2018-05-11-09:43:14
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/mnist/models/model.ckpt-321
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Evaluation [10/100]
INFO:tensorflow:Evaluation [20/100]
INFO:tensorflow:Evaluation [30/100]
INFO:tensorflow:Evaluation [40/100]
INFO:tensorflow:Evaluation [50/100]
INFO:tensorflow:Evaluation [60/100]
INFO:tensorflow:Evaluation [70/100]
INFO:tensorflow:Evaluation [80/100]
INFO:tensorflow:Evaluation [90/100]
INFO:tensorflow:Evaluation [100/100]
INFO:tensorflow:Finished evaluation at 2018-05-11-09:43:19
INFO:tensorflow:Saving dict for global step 321: accuracy = 0.9818, global_step = 321, loss = 0.05992224
INFO:tensorflow:Validation (step 328): accuracy = 0.9818, loss = 0.05992224, global_step = 321
INFO:tensorflow:Saving checkpoints for 341 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:Calling model_fn.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Starting evaluation at 2018-05-11-09:43:25
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/mnist/models/model.ckpt-341
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Evaluation [10/100]
INFO:tensorflow:Evaluation [20/100]
INFO:tensorflow:Evaluation [30/100]
INFO:tensorflow:Evaluation [40/100]
INFO:tensorflow:Evaluation [50/100]
INFO:tensorflow:Evaluation [60/100]
INFO:tensorflow:Evaluation [70/100]
INFO:tensorflow:Evaluation [80/100]
INFO:tensorflow:Evaluation [90/100]
INFO:tensorflow:Evaluation [100/100]
INFO:tensorflow:Finished evaluation at 2018-05-11-09:43:29
INFO:tensorflow:Saving dict for global step 341: accuracy = 0.9836, global_step = 341, loss = 0.05459064
INFO:tensorflow:Validation (step 360): accuracy = 0.9836, loss = 0.05459064, global_step = 341
INFO:tensorflow:Saving checkpoints for 361 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:Calling model_fn.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Starting evaluation at 2018-05-11-09:43:30
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/mnist/models/model.ckpt-361
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Evaluation [10/100]
INFO:tensorflow:Evaluation [20/100]
INFO:tensorflow:Evaluation [30/100]
INFO:tensorflow:Evaluation [40/100]
INFO:tensorflow:Evaluation [50/100]
INFO:tensorflow:Evaluation [60/100]
INFO:tensorflow:Evaluation [70/100]
INFO:tensorflow:Evaluation [80/100]
INFO:tensorflow:Evaluation [90/100]
INFO:tensorflow:Evaluation [100/100]
INFO:tensorflow:Finished evaluation at 2018-05-11-09:43:34
INFO:tensorflow:Saving dict for global step 361: accuracy = 0.9844, global_step = 361, loss = 0.050784793
INFO:tensorflow:Validation (step 361): accuracy = 0.9844, loss = 0.050784793, global_step = 361
INFO:tensorflow:Saving checkpoints for 381 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:Calling model_fn.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Starting evaluation at 2018-05-11-09:43:40
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/mnist/models/model.ckpt-381
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Evaluation [10/100]
INFO:tensorflow:Evaluation [20/100]
INFO:tensorflow:Evaluation [30/100]
INFO:tensorflow:Evaluation [40/100]
INFO:tensorflow:Evaluation [50/100]
INFO:tensorflow:Evaluation [60/100]
INFO:tensorflow:Evaluation [70/100]
INFO:tensorflow:Evaluation [80/100]
INFO:tensorflow:Evaluation [90/100]
INFO:tensorflow:Evaluation [100/100]
INFO:tensorflow:Finished evaluation at 2018-05-11-09:43:45
INFO:tensorflow:Saving dict for global step 381: accuracy = 0.9844, global_step = 381, loss = 0.05304583
INFO:tensorflow:Validation (step 393): accuracy = 0.9844, loss = 0.05304583, global_step = 381
INFO:tensorflow:Saving checkpoints for 401 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:global_step/sec: 2.68594
INFO:tensorflow:loss = 0.17122826, step = 401 (37.231 sec)
INFO:tensorflow:Saving checkpoints for 421 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:Calling model_fn.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Starting evaluation at 2018-05-11-09:43:51
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/mnist/models/model.ckpt-421
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Evaluation [10/100]
INFO:tensorflow:Evaluation [20/100]
INFO:tensorflow:Evaluation [30/100]
INFO:tensorflow:Evaluation [40/100]
INFO:tensorflow:Evaluation [50/100]
INFO:tensorflow:Evaluation [60/100]
INFO:tensorflow:Evaluation [70/100]
INFO:tensorflow:Evaluation [80/100]
INFO:tensorflow:Evaluation [90/100]
INFO:tensorflow:Evaluation [100/100]
INFO:tensorflow:Finished evaluation at 2018-05-11-09:43:56
INFO:tensorflow:Saving dict for global step 421: accuracy = 0.9836, global_step = 421, loss = 0.05426397
INFO:tensorflow:Validation (step 421): accuracy = 0.9836, loss = 0.05426397, global_step = 421
INFO:tensorflow:Saving checkpoints for 429 into /tmp/mnist/models/model.ckpt.
INFO:tensorflow:Loss for final step: 0.051717043.
INFO:tensorflow:Calling model_fn.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Starting evaluation at 2018-05-11-09:43:57
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/mnist/models/model.ckpt-429
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Evaluation [10/100]
INFO:tensorflow:Evaluation [20/100]
INFO:tensorflow:Evaluation [30/100]
INFO:tensorflow:Evaluation [40/100]
INFO:tensorflow:Evaluation [50/100]
INFO:tensorflow:Evaluation [60/100]
INFO:tensorflow:Evaluation [70/100]
INFO:tensorflow:Evaluation [80/100]
INFO:tensorflow:Evaluation [90/100]
INFO:tensorflow:Evaluation [100/100]
INFO:tensorflow:Finished evaluation at 2018-05-11-09:44:02
INFO:tensorflow:Saving dict for global step 429: accuracy = 0.9836, global_step = 429, loss = 0.05495295










    






---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
<ipython-input-4-78698ff130ce> in <module>()
     40     mnist_classifier.export_savedmodel(
     41          os.path.join(hparams.data_directory, 'serving'),
---> 42          serving_input_receiver_fn
     43      )

NameError: name 'serving_input_receiver_fn' is not defined
















In [ ]:



    


train_batch_size = 1_000
train_batch_size



    











In [ ]:



    






    











In [ ]:

Content source: sueiras/training

Similar notebooks:

notebook.community | gallery | about