本文整理汇总了Python中stats.Stats.record_training_cost方法的典型用法代码示例。如果您正苦于以下问题：Python Stats.record_training_cost方法的具体用法？Python Stats.record_training_cost怎么用？Python Stats.record_training_cost使用的例子？那么恭喜您, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在类stats.Stats的用法示例。
在下文中一共展示了Stats.record_training_cost方法的2个代码示例，这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞，您的评价将有助于系统推荐出更棒的Python代码示例。
示例1: str
# 需要导入模块: from stats import Stats [as 别名]
# 或者: from stats.Stats import record_training_cost [as 别名]
ckpt_file = str(int(time.time()))
stats.last_ckpt = ckpt_file
full_ckpt_path = os.path.join(opts.ckpt_dir, RUN_ID, ckpt_file)
saver.save(sess, full_ckpt_path)
log("ckpt saved to %s" % full_ckpt_path)
log("training")
epoch = 0
egs = zip(train_x, train_y)
last_ckpt = ""
while epoch != int(opts.num_epochs):
random.shuffle(egs)
for n, (eg, true_labels) in enumerate(egs):
# train a batch
# TODO: move this into ONE matrix and slice these things out.
# TODO: move away from feeddict and to queueing egs.
batch_cost, _opt = sess.run([cost, optimizer],
feed_dict=eg_and_label_to_feeddict(eg, true_labels))
stats.record_training_cost(batch_cost)
# occasionally check dev set
if stats.n_batches_trained % opts.dev_run_freq == 0:
stats_from_dev_set(stats)
stats.flush_to_stdout(epoch)
# occasionally write a checkpoint
if opts.ckpt_dir and stats.n_batches_trained % opts.ckpt_freq == 0:
save_ckpt()
epoch += 1
if opts.ckpt_dir:
save_ckpt()
示例2: test_fn
# 需要导入模块: from stats import Stats [as 别名]
# 或者: from stats.Stats import record_training_cost [as 别名]
for (s1, s2), y in zip(dev_x, dev_y):
pred_y, cost = test_fn(s1, s2, [y])
actuals.append(y)
predicteds.append(pred_y)
stats.record_dev_cost(cost)
dev_c = confusion_matrix(actuals, predicteds)
dev_accuracy = util.accuracy(dev_c)
stats.set_dev_accuracy(dev_accuracy)
print "dev confusion\n %s (%s)" % (dev_c, dev_accuracy)
log("training")
epoch = 0
training_early_stop_time = opts.max_run_time_sec + time.time()
stats = Stats(os.path.basename(__file__), opts)
egs = zip(train_x, train_y)
while epoch != opts.num_epochs:
random.shuffle(egs)
for (s1, s2), y in egs:
cost, = train_fn(s1, s2, [y])
stats.record_training_cost(cost)
early_stop = False
if opts.max_run_time_sec != -1 and time.time() > training_early_stop_time:
early_stop = True
if stats.n_egs_trained % opts.dev_run_freq == 0 or early_stop:
stats_from_dev_set(stats)
stats.flush_to_stdout(epoch)
if early_stop:
exit(0)
epoch += 1