Python data_utils.generate_feed_dict() Examples

The following are 24 code examples of data_utils.generate_feed_dict(). You can vote up the ones you like or vote down the ones you don't like, and go to the original project or source file by following the links above each example. You may also want to check out all available functions/classes of the module data_utils , or try the search function .
Example #1
Source File: neural_programmer.py    From object_detection_kitti with Apache License 2.0 5 votes vote down vote up 
def evaluate(sess, data, batch_size, graph, i):
  #computes accuracy
  num_examples = 0.0
  gc = 0.0
  for j in range(0, len(data) - batch_size + 1, batch_size):
    [ct] = sess.run([graph.final_correct],
                    feed_dict=data_utils.generate_feed_dict(data, j, batch_size,
                                                            graph))
    gc += ct * batch_size
    num_examples += batch_size
  print "dev set accuracy   after ", i, " : ", gc / num_examples
  print num_examples, len(data)
  print "--------"
Example #2
Source File: neural_programmer.py    From multilabel-image-classification-tensorflow with MIT License 5 votes vote down vote up 
def Train(graph, utility, batch_size, train_data, sess, model_dir,
          saver):
  #performs training
  curr = 0
  train_set_loss = 0.0
  utility.random.shuffle(train_data)
  start = time.time()
  for i in range(utility.FLAGS.train_steps):
    curr_step = i
    if (i > 0 and i % FLAGS.write_every == 0):
      model_file = model_dir + "/model_" + str(i)
      saver.save(sess, model_file)
    if curr + batch_size >= len(train_data):
      curr = 0
      utility.random.shuffle(train_data)
    step, cost_value = sess.run(
        [graph.step, graph.total_cost],
        feed_dict=data_utils.generate_feed_dict(
            train_data, curr, batch_size, graph, train=True, utility=utility))
    curr = curr + batch_size
    train_set_loss += cost_value
    if (i > 0 and i % FLAGS.eval_cycle == 0):
      end = time.time()
      time_taken = end - start
      print("step ", i, " ", time_taken, " seconds ")
      start = end
      print(" printing train set loss: ", train_set_loss / utility.FLAGS.eval_cycle)
      train_set_loss = 0.0
Example #3
Source File: neural_programmer.py    From multilabel-image-classification-tensorflow with MIT License 5 votes vote down vote up 
def evaluate(sess, data, batch_size, graph, i):
  #computes accuracy
  num_examples = 0.0
  gc = 0.0
  for j in range(0, len(data) - batch_size + 1, batch_size):
    [ct] = sess.run([graph.final_correct],
                    feed_dict=data_utils.generate_feed_dict(data, j, batch_size,
                                                            graph))
    gc += ct * batch_size
    num_examples += batch_size
  print("dev set accuracy   after ", i, " : ", gc / num_examples)
  print(num_examples, len(data))
  print("--------")
Example #4
Source File: neural_programmer.py    From models with Apache License 2.0 5 votes vote down vote up 
def Train(graph, utility, batch_size, train_data, sess, model_dir,
          saver):
  #performs training
  curr = 0
  train_set_loss = 0.0
  utility.random.shuffle(train_data)
  start = time.time()
  for i in range(utility.FLAGS.train_steps):
    curr_step = i
    if (i > 0 and i % FLAGS.write_every == 0):
      model_file = model_dir + "/model_" + str(i)
      saver.save(sess, model_file)
    if curr + batch_size >= len(train_data):
      curr = 0
      utility.random.shuffle(train_data)
    step, cost_value = sess.run(
        [graph.step, graph.total_cost],
        feed_dict=data_utils.generate_feed_dict(
            train_data, curr, batch_size, graph, train=True, utility=utility))
    curr = curr + batch_size
    train_set_loss += cost_value
    if (i > 0 and i % FLAGS.eval_cycle == 0):
      end = time.time()
      time_taken = end - start
      print("step ", i, " ", time_taken, " seconds ")
      start = end
      print(" printing train set loss: ", train_set_loss / utility.FLAGS.eval_cycle)
      train_set_loss = 0.0
Example #5
Source File: neural_programmer.py    From models with Apache License 2.0 5 votes vote down vote up 
def evaluate(sess, data, batch_size, graph, i):
  #computes accuracy
  num_examples = 0.0
  gc = 0.0
  for j in range(0, len(data) - batch_size + 1, batch_size):
    [ct] = sess.run([graph.final_correct],
                    feed_dict=data_utils.generate_feed_dict(data, j, batch_size,
                                                            graph))
    gc += ct * batch_size
    num_examples += batch_size
  print("dev set accuracy   after ", i, " : ", gc / num_examples)
  print(num_examples, len(data))
  print("--------")
Example #6
Source File: neural_programmer.py    From g-tensorflow-models with Apache License 2.0 5 votes vote down vote up 
def Train(graph, utility, batch_size, train_data, sess, model_dir,
          saver):
  #performs training
  curr = 0
  train_set_loss = 0.0
  utility.random.shuffle(train_data)
  start = time.time()
  for i in range(utility.FLAGS.train_steps):
    curr_step = i
    if (i > 0 and i % FLAGS.write_every == 0):
      model_file = model_dir + "/model_" + str(i)
      saver.save(sess, model_file)
    if curr + batch_size >= len(train_data):
      curr = 0
      utility.random.shuffle(train_data)
    step, cost_value = sess.run(
        [graph.step, graph.total_cost],
        feed_dict=data_utils.generate_feed_dict(
            train_data, curr, batch_size, graph, train=True, utility=utility))
    curr = curr + batch_size
    train_set_loss += cost_value
    if (i > 0 and i % FLAGS.eval_cycle == 0):
      end = time.time()
      time_taken = end - start
      print("step ", i, " ", time_taken, " seconds ")
      start = end
      print(" printing train set loss: ", train_set_loss / utility.FLAGS.eval_cycle)
      train_set_loss = 0.0
Example #7
Source File: neural_programmer.py    From g-tensorflow-models with Apache License 2.0 5 votes vote down vote up 
def evaluate(sess, data, batch_size, graph, i):
  #computes accuracy
  num_examples = 0.0
  gc = 0.0
  for j in range(0, len(data) - batch_size + 1, batch_size):
    [ct] = sess.run([graph.final_correct],
                    feed_dict=data_utils.generate_feed_dict(data, j, batch_size,
                                                            graph))
    gc += ct * batch_size
    num_examples += batch_size
  print("dev set accuracy   after ", i, " : ", gc / num_examples)
  print(num_examples, len(data))
  print("--------")
Example #8
Source File: neural_programmer.py    From HumanRecognition with MIT License 5 votes vote down vote up 
def Train(graph, utility, batch_size, train_data, sess, model_dir,
          saver):
  #performs training
  curr = 0
  train_set_loss = 0.0
  utility.random.shuffle(train_data)
  start = time.time()
  for i in range(utility.FLAGS.train_steps):
    curr_step = i
    if (i > 0 and i % FLAGS.write_every == 0):
      model_file = model_dir + "/model_" + str(i)
      saver.save(sess, model_file)
    if curr + batch_size >= len(train_data):
      curr = 0
      utility.random.shuffle(train_data)
    step, cost_value = sess.run(
        [graph.step, graph.total_cost],
        feed_dict=data_utils.generate_feed_dict(
            train_data, curr, batch_size, graph, train=True, utility=utility))
    curr = curr + batch_size
    train_set_loss += cost_value
    if (i > 0 and i % FLAGS.eval_cycle == 0):
      end = time.time()
      time_taken = end - start
      print "step ", i, " ", time_taken, " seconds "
      start = end
      print " printing train set loss: ", train_set_loss / utility.FLAGS.eval_cycle
      train_set_loss = 0.0
Example #9
Source File: neural_programmer.py    From HumanRecognition with MIT License 5 votes vote down vote up 
def evaluate(sess, data, batch_size, graph, i):
  #computes accuracy
  num_examples = 0.0
  gc = 0.0
  for j in range(0, len(data) - batch_size + 1, batch_size):
    [ct] = sess.run([graph.final_correct],
                    feed_dict=data_utils.generate_feed_dict(data, j, batch_size,
                                                            graph))
    gc += ct * batch_size
    num_examples += batch_size
  print "dev set accuracy   after ", i, " : ", gc / num_examples
  print num_examples, len(data)
  print "--------"
Example #10
Source File: neural_programmer.py    From object_detection_with_tensorflow with MIT License 5 votes vote down vote up 
def Train(graph, utility, batch_size, train_data, sess, model_dir,
          saver):
  #performs training
  curr = 0
  train_set_loss = 0.0
  utility.random.shuffle(train_data)
  start = time.time()
  for i in range(utility.FLAGS.train_steps):
    curr_step = i
    if (i > 0 and i % FLAGS.write_every == 0):
      model_file = model_dir + "/model_" + str(i)
      saver.save(sess, model_file)
    if curr + batch_size >= len(train_data):
      curr = 0
      utility.random.shuffle(train_data)
    step, cost_value = sess.run(
        [graph.step, graph.total_cost],
        feed_dict=data_utils.generate_feed_dict(
            train_data, curr, batch_size, graph, train=True, utility=utility))
    curr = curr + batch_size
    train_set_loss += cost_value
    if (i > 0 and i % FLAGS.eval_cycle == 0):
      end = time.time()
      time_taken = end - start
      print "step ", i, " ", time_taken, " seconds "
      start = end
      print " printing train set loss: ", train_set_loss / utility.FLAGS.eval_cycle
      train_set_loss = 0.0
Example #11
Source File: neural_programmer.py    From object_detection_with_tensorflow with MIT License 5 votes vote down vote up 
def evaluate(sess, data, batch_size, graph, i):
  #computes accuracy
  num_examples = 0.0
  gc = 0.0
  for j in range(0, len(data) - batch_size + 1, batch_size):
    [ct] = sess.run([graph.final_correct],
                    feed_dict=data_utils.generate_feed_dict(data, j, batch_size,
                                                            graph))
    gc += ct * batch_size
    num_examples += batch_size
  print "dev set accuracy   after ", i, " : ", gc / num_examples
  print num_examples, len(data)
  print "--------"
Example #12
Source File: neural_programmer.py    From object_detection_kitti with Apache License 2.0 5 votes vote down vote up 
def Train(graph, utility, batch_size, train_data, sess, model_dir,
          saver):
  #performs training
  curr = 0
  train_set_loss = 0.0
  utility.random.shuffle(train_data)
  start = time.time()
  for i in range(utility.FLAGS.train_steps):
    curr_step = i
    if (i > 0 and i % FLAGS.write_every == 0):
      model_file = model_dir + "/model_" + str(i)
      saver.save(sess, model_file)
    if curr + batch_size >= len(train_data):
      curr = 0
      utility.random.shuffle(train_data)
    step, cost_value = sess.run(
        [graph.step, graph.total_cost],
        feed_dict=data_utils.generate_feed_dict(
            train_data, curr, batch_size, graph, train=True, utility=utility))
    curr = curr + batch_size
    train_set_loss += cost_value
    if (i > 0 and i % FLAGS.eval_cycle == 0):
      end = time.time()
      time_taken = end - start
      print "step ", i, " ", time_taken, " seconds "
      start = end
      print " printing train set loss: ", train_set_loss / utility.FLAGS.eval_cycle
      train_set_loss = 0.0
Example #13
Source File: neural_programmer.py    From DOTA_models with Apache License 2.0 5 votes vote down vote up 
def evaluate(sess, data, batch_size, graph, i):
  #computes accuracy
  num_examples = 0.0
  gc = 0.0
  for j in range(0, len(data) - batch_size + 1, batch_size):
    [ct] = sess.run([graph.final_correct],
                    feed_dict=data_utils.generate_feed_dict(data, j, batch_size,
                                                            graph))
    gc += ct * batch_size
    num_examples += batch_size
  print "dev set accuracy   after ", i, " : ", gc / num_examples
  print num_examples, len(data)
  print "--------"
Example #14
Source File: neural_programmer.py    From hands-detection with MIT License 5 votes vote down vote up 
def Train(graph, utility, batch_size, train_data, sess, model_dir,
          saver):
  #performs training
  curr = 0
  train_set_loss = 0.0
  utility.random.shuffle(train_data)
  start = time.time()
  for i in range(utility.FLAGS.train_steps):
    curr_step = i
    if (i > 0 and i % FLAGS.write_every == 0):
      model_file = model_dir + "/model_" + str(i)
      saver.save(sess, model_file)
    if curr + batch_size >= len(train_data):
      curr = 0
      utility.random.shuffle(train_data)
    step, cost_value = sess.run(
        [graph.step, graph.total_cost],
        feed_dict=data_utils.generate_feed_dict(
            train_data, curr, batch_size, graph, train=True, utility=utility))
    curr = curr + batch_size
    train_set_loss += cost_value
    if (i > 0 and i % FLAGS.eval_cycle == 0):
      end = time.time()
      time_taken = end - start
      print "step ", i, " ", time_taken, " seconds "
      start = end
      print " printing train set loss: ", train_set_loss / utility.FLAGS.eval_cycle
      train_set_loss = 0.0
Example #15
Source File: neural_programmer.py    From hands-detection with MIT License 5 votes vote down vote up 
def evaluate(sess, data, batch_size, graph, i):
  #computes accuracy
  num_examples = 0.0
  gc = 0.0
  for j in range(0, len(data) - batch_size + 1, batch_size):
    [ct] = sess.run([graph.final_correct],
                    feed_dict=data_utils.generate_feed_dict(data, j, batch_size,
                                                            graph))
    gc += ct * batch_size
    num_examples += batch_size
  print "dev set accuracy   after ", i, " : ", gc / num_examples
  print num_examples, len(data)
  print "--------"
Example #16
Source File: neural_programmer.py    From ECO-pytorch with BSD 2-Clause "Simplified" License 5 votes vote down vote up 
def Train(graph, utility, batch_size, train_data, sess, model_dir,
          saver):
  #performs training
  curr = 0
  train_set_loss = 0.0
  utility.random.shuffle(train_data)
  start = time.time()
  for i in range(utility.FLAGS.train_steps):
    curr_step = i
    if (i > 0 and i % FLAGS.write_every == 0):
      model_file = model_dir + "/model_" + str(i)
      saver.save(sess, model_file)
    if curr + batch_size >= len(train_data):
      curr = 0
      utility.random.shuffle(train_data)
    step, cost_value = sess.run(
        [graph.step, graph.total_cost],
        feed_dict=data_utils.generate_feed_dict(
            train_data, curr, batch_size, graph, train=True, utility=utility))
    curr = curr + batch_size
    train_set_loss += cost_value
    if (i > 0 and i % FLAGS.eval_cycle == 0):
      end = time.time()
      time_taken = end - start
      print "step ", i, " ", time_taken, " seconds "
      start = end
      print " printing train set loss: ", train_set_loss / utility.FLAGS.eval_cycle
      train_set_loss = 0.0
Example #17
Source File: neural_programmer.py    From ECO-pytorch with BSD 2-Clause "Simplified" License 5 votes vote down vote up 
def evaluate(sess, data, batch_size, graph, i):
  #computes accuracy
  num_examples = 0.0
  gc = 0.0
  for j in range(0, len(data) - batch_size + 1, batch_size):
    [ct] = sess.run([graph.final_correct],
                    feed_dict=data_utils.generate_feed_dict(data, j, batch_size,
                                                            graph))
    gc += ct * batch_size
    num_examples += batch_size
  print "dev set accuracy   after ", i, " : ", gc / num_examples
  print num_examples, len(data)
  print "--------"
Example #18
Source File: neural_programmer.py    From Action_Recognition_Zoo with MIT License 5 votes vote down vote up 
def Train(graph, utility, batch_size, train_data, sess, model_dir,
          saver):
  #performs training
  curr = 0
  train_set_loss = 0.0
  utility.random.shuffle(train_data)
  start = time.time()
  for i in range(utility.FLAGS.train_steps):
    curr_step = i
    if (i > 0 and i % FLAGS.write_every == 0):
      model_file = model_dir + "/model_" + str(i)
      saver.save(sess, model_file)
    if curr + batch_size >= len(train_data):
      curr = 0
      utility.random.shuffle(train_data)
    step, cost_value = sess.run(
        [graph.step, graph.total_cost],
        feed_dict=data_utils.generate_feed_dict(
            train_data, curr, batch_size, graph, train=True, utility=utility))
    curr = curr + batch_size
    train_set_loss += cost_value
    if (i > 0 and i % FLAGS.eval_cycle == 0):
      end = time.time()
      time_taken = end - start
      print "step ", i, " ", time_taken, " seconds "
      start = end
      print " printing train set loss: ", train_set_loss / utility.FLAGS.eval_cycle
      train_set_loss = 0.0
Example #19
Source File: neural_programmer.py    From Action_Recognition_Zoo with MIT License 5 votes vote down vote up 
def evaluate(sess, data, batch_size, graph, i):
  #computes accuracy
  num_examples = 0.0
  gc = 0.0
  for j in range(0, len(data) - batch_size + 1, batch_size):
    [ct] = sess.run([graph.final_correct],
                    feed_dict=data_utils.generate_feed_dict(data, j, batch_size,
                                                            graph))
    gc += ct * batch_size
    num_examples += batch_size
  print "dev set accuracy   after ", i, " : ", gc / num_examples
  print num_examples, len(data)
  print "--------"
Example #20
Source File: neural_programmer.py    From Gun-Detector with Apache License 2.0 5 votes vote down vote up 
def Train(graph, utility, batch_size, train_data, sess, model_dir,
          saver):
  #performs training
  curr = 0
  train_set_loss = 0.0
  utility.random.shuffle(train_data)
  start = time.time()
  for i in range(utility.FLAGS.train_steps):
    curr_step = i
    if (i > 0 and i % FLAGS.write_every == 0):
      model_file = model_dir + "/model_" + str(i)
      saver.save(sess, model_file)
    if curr + batch_size >= len(train_data):
      curr = 0
      utility.random.shuffle(train_data)
    step, cost_value = sess.run(
        [graph.step, graph.total_cost],
        feed_dict=data_utils.generate_feed_dict(
            train_data, curr, batch_size, graph, train=True, utility=utility))
    curr = curr + batch_size
    train_set_loss += cost_value
    if (i > 0 and i % FLAGS.eval_cycle == 0):
      end = time.time()
      time_taken = end - start
      print("step ", i, " ", time_taken, " seconds ")
      start = end
      print(" printing train set loss: ", train_set_loss / utility.FLAGS.eval_cycle)
      train_set_loss = 0.0
Example #21
Source File: neural_programmer.py    From Gun-Detector with Apache License 2.0 5 votes vote down vote up 
def evaluate(sess, data, batch_size, graph, i):
  #computes accuracy
  num_examples = 0.0
  gc = 0.0
  for j in range(0, len(data) - batch_size + 1, batch_size):
    [ct] = sess.run([graph.final_correct],
                    feed_dict=data_utils.generate_feed_dict(data, j, batch_size,
                                                            graph))
    gc += ct * batch_size
    num_examples += batch_size
  print("dev set accuracy   after ", i, " : ", gc / num_examples)
  print(num_examples, len(data))
  print("--------")
Example #22
Source File: neural_programmer.py    From yolo_v2 with Apache License 2.0 5 votes vote down vote up 
def Train(graph, utility, batch_size, train_data, sess, model_dir,
          saver):
  #performs training
  curr = 0
  train_set_loss = 0.0
  utility.random.shuffle(train_data)
  start = time.time()
  for i in range(utility.FLAGS.train_steps):
    curr_step = i
    if (i > 0 and i % FLAGS.write_every == 0):
      model_file = model_dir + "/model_" + str(i)
      saver.save(sess, model_file)
    if curr + batch_size >= len(train_data):
      curr = 0
      utility.random.shuffle(train_data)
    step, cost_value = sess.run(
        [graph.step, graph.total_cost],
        feed_dict=data_utils.generate_feed_dict(
            train_data, curr, batch_size, graph, train=True, utility=utility))
    curr = curr + batch_size
    train_set_loss += cost_value
    if (i > 0 and i % FLAGS.eval_cycle == 0):
      end = time.time()
      time_taken = end - start
      print "step ", i, " ", time_taken, " seconds "
      start = end
      print " printing train set loss: ", train_set_loss / utility.FLAGS.eval_cycle
      train_set_loss = 0.0
Example #23
Source File: neural_programmer.py    From yolo_v2 with Apache License 2.0 5 votes vote down vote up 
def evaluate(sess, data, batch_size, graph, i):
  #computes accuracy
  num_examples = 0.0
  gc = 0.0
  for j in range(0, len(data) - batch_size + 1, batch_size):
    [ct] = sess.run([graph.final_correct],
                    feed_dict=data_utils.generate_feed_dict(data, j, batch_size,
                                                            graph))
    gc += ct * batch_size
    num_examples += batch_size
  print "dev set accuracy   after ", i, " : ", gc / num_examples
  print num_examples, len(data)
  print "--------"
Example #24
Source File: neural_programmer.py    From DOTA_models with Apache License 2.0 5 votes vote down vote up 
def Train(graph, utility, batch_size, train_data, sess, model_dir,
          saver):
  #performs training
  curr = 0
  train_set_loss = 0.0
  utility.random.shuffle(train_data)
  start = time.time()
  for i in range(utility.FLAGS.train_steps):
    curr_step = i
    if (i > 0 and i % FLAGS.write_every == 0):
      model_file = model_dir + "/model_" + str(i)
      saver.save(sess, model_file)
    if curr + batch_size >= len(train_data):
      curr = 0
      utility.random.shuffle(train_data)
    step, cost_value = sess.run(
        [graph.step, graph.total_cost],
        feed_dict=data_utils.generate_feed_dict(
            train_data, curr, batch_size, graph, train=True, utility=utility))
    curr = curr + batch_size
    train_set_loss += cost_value
    if (i > 0 and i % FLAGS.eval_cycle == 0):
      end = time.time()
      time_taken = end - start
      print "step ", i, " ", time_taken, " seconds "
      start = end
      print " printing train set loss: ", train_set_loss / utility.FLAGS.eval_cycle
      train_set_loss = 0.0