Code designed to run training and evaluation on a DNNClassifier

text_classifier_train_eval.py

import tensorflow as tf
import tensorflow_hub as hub
import pandas as pd
import numpy as np
import shutil

import seaborn as sns
import matplotlib.pyplot as plt

MODULE_SPEC_50 = 'https://tfhub.dev/google/nnlm-en-dim50/1'
MODULE_SPEC_128='https://tfhub.dev/google/nnlm-en-dim128/1'

print(tf.__version__)

# Input function
CSV_COLUMNS = ['sentiment', 'cleanReviewText']
LABEL_COLUMN = 'sentiment'
DEFAULTS = [[0], [' ']]

def read_dataset(filename, mode, batch_size = 512):
  def _input_fn():
    def decode_csv(value_column):
      columns = tf.decode_csv(value_column, record_defaults = DEFAULTS)
      features = dict(zip(CSV_COLUMNS, columns))
      label = features.pop(LABEL_COLUMN)
      return features, label

    # Create list of files that match pattern
    file_list = tf.gfile.Glob(filename)

    # Create dataset from file list
    dataset = tf.data.TextLineDataset(file_list).map(decode_csv)
    if mode == tf.estimator.ModeKeys.TRAIN:
        num_epochs = None # indefinitely
        dataset = dataset.shuffle(buffer_size = 10 * batch_size)
    else:
        num_epochs = 1 # end-of-input after this

    dataset = dataset.repeat(num_epochs).batch(batch_size)
    return dataset.make_one_shot_iterator().get_next()
  return _input_fn

# Embedding Feature
embedding_feat_col = hub.text_embedding_column(key='cleanReviewText', module_spec=MODULE_SPEC_50)

# Serving Input Function
def serving_input_fn():
  feature_placeholders = {
    'cleanReviewText' : tf.placeholder(tf.string, [None])
  }
  features = {
      'cleanReviewText': hub.text_embedding_column(key='cleanReviewText', module_spec=MODULE_SPEC_50)
  }
  return tf.estimator.export.ServingInputReceiver(features, feature_placeholders)

# Train and Evaluate
def train_and_evaluate(output_dir, num_train_steps):
  estimator = tf.estimator.DNNClassifier(
                       model_dir = output_dir,  hidden_units=[512, 1024, 512, 256], n_classes=2,
                       feature_columns = [embedding_feat_col])
  train_spec=tf.estimator.TrainSpec(
                       input_fn = read_dataset('./csv_headless/amazon_instant_video.csv', mode = tf.estimator.ModeKeys.TRAIN),
                       max_steps = num_train_steps)
  exporter = tf.estimator.LatestExporter('exporter', serving_input_fn)
  eval_spec=tf.estimator.EvalSpec(
                       input_fn = read_dataset('./csv_headless/automotive.csv', mode = tf.estimator.ModeKeys.EVAL),
                       steps = None,
                       start_delay_secs = 1, # start evaluating after N seconds
                       throttle_secs = 10,  # evaluate every N seconds
                       exporters = exporter)
  tf.estimator.train_and_evaluate(estimator, train_spec, eval_spec)
  
# Run training
OUTDIR = 'sentiment_trained'
shutil.rmtree(OUTDIR, ignore_errors = True) 
train_and_evaluate(OUTDIR, num_train_steps = 100)