Python data_utils.get_batch方法代碼示例

# 需要導入模塊: import data_utils [as 別名]
# 或者: from data_utils import get_batch [as 別名]
def interactive():
  """Interactively probe an existing model."""
  with tf.Session() as sess:
    model, _, _, _, _, _ = initialize(sess)
    sys.stdout.write("Input to Neural GPU, e.g., 0 1. Use -1 for PAD.\n")
    sys.stdout.write("> ")
    sys.stdout.flush()
    inpt = sys.stdin.readline()
    while inpt:
      ids = [data.to_id(s) for s in inpt.strip().split()]
      inpt, target = data.get_batch(len(ids), 1, False, "",
                                    preset=(ids, [0 for _ in ids]))
      _, res, _, _ = model.step(sess, inpt, target, False)
      res = [np.argmax(o, axis=1) for o in res]
      res = [o for o in res[:len(ids)] if o > 0]
      print "  " + " ".join([data.to_symbol(output[0]) for output in res])
      sys.stdout.write("> ")
      sys.stdout.flush()
      inpt = sys.stdin.readline() 

# 需要導入模塊: import data_utils [as 別名]
# 或者: from data_utils import get_batch [as 別名]
def single_test(bin_id, model, sess, nprint, batch_size, dev, p, print_out=True,
                offset=None, beam_model=None):
  """Test model on test data of length l using the given session."""
  if not dev[p][bin_id]:
    data.print_out("  bin %d (%d)\t%s\tppl NA errors NA seq-errors NA"
                   % (bin_id, data.bins[bin_id], p))
    return 1.0, 1.0, 0.0
  inpt, target = data.get_batch(
      bin_id, batch_size, dev[p], FLAGS.height, offset)
  if FLAGS.beam_size > 1 and beam_model:
    loss, res, new_tgt, scores = m_step(
        model, beam_model, sess, batch_size, inpt, target, bin_id,
        FLAGS.eval_beam_steps, p)
    score_avgs = [sum(s) / float(len(s)) for s in scores]
    score_maxs = [max(s) for s in scores]
    score_str = ["(%.2f, %.2f)" % (score_avgs[i], score_maxs[i])
                 for i in xrange(FLAGS.eval_beam_steps)]
    data.print_out("  == scores (avg, max): %s" % "; ".join(score_str))
    errors, total, seq_err = data.accuracy(inpt, res, target, batch_size,
                                           nprint, new_tgt, scores[-1])
  else:
    loss, res, _, _ = model.step(sess, inpt, target, False)
    errors, total, seq_err = data.accuracy(inpt, res, target, batch_size,
                                           nprint)
  seq_err = float(seq_err) / batch_size
  if total > 0:
    errors = float(errors) / total
  if print_out:
    data.print_out("  bin %d (%d)\t%s\tppl %.2f errors %.2f seq-errors %.2f"
                   % (bin_id, data.bins[bin_id], p, data.safe_exp(loss),
                      100 * errors, 100 * seq_err))
  return (errors, seq_err, loss) 

# 需要導入模塊: import data_utils [as 別名]
# 或者: from data_utils import get_batch [as 別名]
def single_test(l, model, sess, task, nprint, batch_size, print_out=True,
                offset=None, ensemble=None, get_steps=False):
  """Test model on test data of length l using the given session."""
  inpt, target = data.get_batch(l, batch_size, False, task, offset)
  _, res, _, steps = model.step(sess, inpt, target, False, get_steps=get_steps)
  errors, total, seq_err = data.accuracy(inpt, res, target, batch_size, nprint)
  seq_err = float(seq_err) / batch_size
  if total > 0:
    errors = float(errors) / total
  if print_out:
    data.print_out("  %s len %d errors %.2f sequence-errors %.2f"
                   % (task, l, 100*errors, 100*seq_err))
  # Ensemble eval.
  if ensemble:
    results = []
    for m in ensemble:
      model.saver.restore(sess, m)
      _, result, _, _ = model.step(sess, inpt, target, False)
      m_errors, m_total, m_seq_err = data.accuracy(inpt, result, target,
                                                   batch_size, nprint)
      m_seq_err = float(m_seq_err) / batch_size
      if total > 0:
        m_errors = float(m_errors) / m_total
      data.print_out("     %s len %d m-errors %.2f m-sequence-errors %.2f"
                     % (task, l, 100*m_errors, 100*m_seq_err))
      results.append(result)
    ens = [sum(o) for o in zip(*results)]
    errors, total, seq_err = data.accuracy(inpt, ens, target,
                                           batch_size, nprint)
    seq_err = float(seq_err) / batch_size
    if total > 0:
      errors = float(errors) / total
    if print_out:
      data.print_out("  %s len %d ens-errors %.2f ens-sequence-errors %.2f"
                     % (task, l, 100*errors, 100*seq_err))
  return errors, seq_err, (steps, inpt, [np.argmax(o, axis=1) for o in res]) 

# 需要導入模塊: import data_utils [as 別名]
# 或者: from data_utils import get_batch [as 別名]
def train_epoch(epoch, samples, labels, sess, Z, X, CG, CD, CS, D_loss, G_loss, D_solver, G_solver, 
                batch_size, use_time, D_rounds, G_rounds, seq_length, 
                latent_dim, num_generated_features, cond_dim, max_val, WGAN_clip, one_hot):
    """
    Train generator and discriminator for one epoch.
    """
    for batch_idx in range(0, int(len(samples) / batch_size) - (D_rounds + (cond_dim > 0)*G_rounds), D_rounds + (cond_dim > 0)*G_rounds):
        # update the discriminator
        for d in range(D_rounds):
            X_mb, Y_mb = data_utils.get_batch(samples, batch_size, batch_idx + d, labels)
            Z_mb = sample_Z(batch_size, seq_length, latent_dim, use_time)
            if cond_dim > 0:
                # CGAN
                Y_mb = Y_mb.reshape(-1, cond_dim)
                if one_hot:
                    # change all of the labels to a different one
                    offsets = np.random.choice(cond_dim-1, batch_size) + 1
                    new_labels = (np.argmax(Y_mb, axis=1) + offsets) % cond_dim
                    Y_wrong = np.zeros_like(Y_mb)
                    Y_wrong[np.arange(batch_size), new_labels] = 1
                else:
                    # flip all of the bits (assuming binary...)
                    Y_wrong = 1 - Y_mb
                _ = sess.run(D_solver, feed_dict={X: X_mb, Z: Z_mb, CD: Y_mb, CS: Y_wrong, CG: Y_mb})
            else:
                _ = sess.run(D_solver, feed_dict={X: X_mb, Z: Z_mb})
            if WGAN_clip:
                # clip the weights
                _ = sess.run([clip_disc_weights])
        # update the generator
        for g in range(G_rounds):
            if cond_dim > 0:
                # note we are essentially throwing these X_mb away...
                X_mb, Y_mb = data_utils.get_batch(samples, batch_size, batch_idx + D_rounds + g, labels)
                _ = sess.run(G_solver,
                        feed_dict={Z: sample_Z(batch_size, seq_length, latent_dim, use_time=use_time), CG: Y_mb})
            else:
                _ = sess.run(G_solver,
                        feed_dict={Z: sample_Z(batch_size, seq_length, latent_dim, use_time=use_time)})
    # at the end, get the loss
    if cond_dim > 0:
        D_loss_curr, G_loss_curr = sess.run([D_loss, G_loss], feed_dict={X: X_mb, Z: sample_Z(batch_size, seq_length, latent_dim, use_time=use_time), CG: Y_mb, CD: Y_mb})
        D_loss_curr = np.mean(D_loss_curr)
        G_loss_curr = np.mean(G_loss_curr)
    else:
        D_loss_curr, G_loss_curr = sess.run([D_loss, G_loss], feed_dict={X: X_mb, Z: sample_Z(batch_size, seq_length, latent_dim, use_time=use_time)})
        D_loss_curr = np.mean(D_loss_curr)
        G_loss_curr = np.mean(G_loss_curr)
    return D_loss_curr, G_loss_curr