def backward():
class_num = config.voc_class_num
yolo = YOLO(class_num,
config.voc_anchors,
width=config.width,
height=config.height)
data = Data(config.voc_root_dir,
config.voc_dir_ls,
"./data/voc.names",
class_num,
config.batch_size,
config.voc_anchors,
config.data_augment,
config.width,
config.height,
data_debug=config.data_debug)
inputs = tf.compat.v1.placeholder(dtype=tf.float32,
shape=[config.batch_size, None, None, 3])
y1_true = tf.compat.v1.placeholder(
dtype=tf.float32,
shape=[config.batch_size, None, None, 3, 4 + 1 + class_num])
y2_true = tf.compat.v1.placeholder(
dtype=tf.float32,
shape=[config.batch_size, None, None, 3, 4 + 1 + class_num])
y3_true = tf.compat.v1.placeholder(
dtype=tf.float32,
shape=[config.batch_size, None, None, 3, 4 + 1 + class_num])
feature_y1, feature_y2, feature_y3 = yolo.forward(
inputs, weight_decay=config.weight_decay, isTrain=True)
global_step = tf.Variable(0, trainable=False)
# 损失 yolov4
loss = yolo.get_loss_v4(feature_y1, feature_y2, feature_y3, y1_true,
y2_true, y3_true, config.cls_normalizer,
config.ignore_thresh, config.prob_thresh,
config.score_thresh)
l2_loss = tf.compat.v1.losses.get_regularization_loss()
epoch = compute_curr_epoch(global_step, config.batch_size, data.num_imgs)
lr = config_lr(config.lr_type, config.lr_init, epoch)
optimizer = config_optimizer(config.optimizer_type, lr, config.momentum)
update_ops = tf.compat.v1.get_collection(tf.compat.v1.GraphKeys.UPDATE_OPS)
with tf.control_dependencies(update_ops):
gvs = optimizer.compute_gradients(loss + l2_loss)
clip_grad_var = [
gv if gv[0] is None else [tf.clip_by_norm(gv[0], 100.), gv[1]]
for gv in gvs
]
train_step = optimizer.apply_gradients(clip_grad_var,
global_step=global_step)
# 初始化
init = tf.compat.v1.global_variables_initializer()
saver = tf.compat.v1.train.Saver()
with tf.compat.v1.Session() as sess:
sess.run(init)
step = 0
ckpt = tf.compat.v1.train.get_checkpoint_state(config.voc_model_path)
if ckpt and ckpt.model_checkpoint_path:
saver.restore(sess, ckpt.model_checkpoint_path)
step = ckpt.model_checkpoint_path.split('/')[-1].split('-')[-1]
step = eval(step)
Log.add_log("message:存在 ckpt 模型, global_step=" + str(step))
else:
Log.add_log("message:不存在 ckpt 模型")
# 一共迭代这么多次
total_steps = np.ceil(config.total_epoch * data.num_imgs /
config.batch_size)
while step < total_steps:
start = time.perf_counter()
batch_img, y1, y2, y3 = next(data)
_, loss_, step, lr_ = sess.run([train_step, loss, global_step, lr],
feed_dict={
inputs: batch_img,
y1_true: y1,
y2_true: y2,
y3_true: y3
})
end = time.perf_counter()
print(
"step: %6d, loss: %.5g\t, w: %3d, h: %3d, lr:%.5g\t, time: %5f s"
% (step, loss_, data.width, data.height, lr_, end - start))
if (loss_ > 1e3) and (step > 1e3):
Log.add_log("error:loss exception, loss_value = " + str(loss_))
''' break the process or lower learning rate '''
raise ValueError("error:loss exception, loss_value = " +
str(loss_) +
", please lower your learning rate")
# lr = tf.math.maximum(tf.math.divide(lr, 10), config.lr_lower)
if step % 5 == 2:
Log.add_loss(str(step) + "\t" + str(loss_))
if (step + 1) % config.save_step == 0:
if config.save_ckpt_model:
# save ckpt model
Log.add_log("message: save ckpt model, step=" + str(step) +
", lr=" + str(lr_))
saver.save(sess,
path.join(config.voc_model_path,
config.voc_model_name),
global_step=step)
if config.save_pb_model:
Log.add_log("message: save pb model, step=" + str(step) +
", lr=" + str(lr_))
pb_model_name = path.join(
config.voc_model_path,
config.voc_model_name) + '-' + str(step) + ".pb"
constant_graph = graph_util.convert_variables_to_constants(
sess, sess.graph_def, [
'yolo/Conv_1/BiasAdd', 'yolo/Conv_9/BiasAdd',
'yolo/Conv_17/BiasAdd'
])
# save PB model
with tf.gfile.FastGFile(pb_model_name, mode='wb') as f:
f.write(constant_graph.SerializeToString())
# save ckpt model
if config.save_ckpt_model:
Log.add_log("message: save final ckpt model, step=" + str(step))
saver.save(sess,
path.join(config.voc_model_path, config.voc_model_name),
global_step=step)
# save pb model
if config.save_pb_model:
Log.add_log("message: save final pb model, step=" + str(step))
pb_model_name = path.join(
config.voc_model_path,
config.voc_model_name) + '-' + str(step) + ".pb"
constant_graph = graph_util.convert_variables_to_constants(
sess, sess.graph_def, [
'yolo/Conv_1/BiasAdd', 'yolo/Conv_9/BiasAdd',
'yolo/Conv_17/BiasAdd'
])
# save PB model
with tf.gfile.FastGFile(pb_model_name, mode='wb') as f:
f.write(constant_graph.SerializeToString())
return 0
def backward():
yolo = YOLO(config.class_num,
config.anchors,
width=config.width,
height=config.height)
data = Data(config.train_file,
config.class_num,
config.batch_size,
config.anchors,
config.multi_scale_img,
width=config.width,
height=config.height)
inputs = tf.compat.v1.placeholder(dtype=tf.float32,
shape=[config.batch_size, None, None, 3])
y1_true = tf.compat.v1.placeholder(
dtype=tf.float32,
shape=[config.batch_size, None, None, 3, 4 + 1 + config.class_num])
y2_true = tf.compat.v1.placeholder(
dtype=tf.float32,
shape=[config.batch_size, None, None, 3, 4 + 1 + config.class_num])
y3_true = tf.compat.v1.placeholder(
dtype=tf.float32,
shape=[config.batch_size, None, None, 3, 4 + 1 + config.class_num])
feature_y1, feature_y2, feature_y3 = yolo.forward(
inputs, weight_decay=config.weight_decay, isTrain=True)
global_step = tf.Variable(0, trainable=False)
# 损失 yolov4
loss = yolo.get_loss_v4(feature_y1, feature_y2, feature_y3, y1_true,
y2_true, y3_true, config.cls_normalizer,
config.ignore_thresh, config.prob_thresh,
config.score_thresh)
l2_loss = tf.compat.v1.losses.get_regularization_loss()
epoch = compute_curr_epoch(global_step, config.batch_size,
len(data.imgs_path))
lr = config_lr(config.lr_type, config.lr_init, epoch)
optimizer = config_optimizer(config.optimizer_type, lr, config.momentum)
update_ops = tf.compat.v1.get_collection(tf.compat.v1.GraphKeys.UPDATE_OPS)
with tf.control_dependencies(update_ops):
gvs = optimizer.compute_gradients(loss + l2_loss)
clip_grad_var = [
gv if gv[0] is None else [tf.clip_by_norm(gv[0], 100.), gv[1]]
for gv in gvs
]
train_step = optimizer.apply_gradients(clip_grad_var,
global_step=global_step)
# 初始化
init = tf.compat.v1.global_variables_initializer()
saver = tf.compat.v1.train.Saver()
with tf.compat.v1.Session() as sess:
sess.run(init)
step = 0
ckpt = tf.compat.v1.train.get_checkpoint_state(config.model_path)
if ckpt and ckpt.model_checkpoint_path:
saver.restore(sess, ckpt.model_checkpoint_path)
step = ckpt.model_checkpoint_path.split('/')[-1].split('-')[-1]
step = eval(step)
Log.add_log("message:存在 ckpt 模型, global_step=" + str(step))
else:
Log.add_log("message:不存在 ckpt 模型")
# 一共迭代这么多次
total_steps = np.ceil(config.total_epoch * len(data.imgs_path) /
config.batch_size)
while step < total_steps:
start = time.perf_counter()
batch_img, y1, y2, y3 = next(data)
_, loss_, step, lr_ = sess.run([train_step, loss, global_step, lr],
feed_dict={
inputs: batch_img,
y1_true: y1,
y2_true: y2,
y3_true: y3
})
end = time.perf_counter()
print(
"step: %6d, loss: %.5g\t, w: %3d, h: %3d, lr:%.5g\t, time: %5f s"
% (step, loss_, data.width, data.height, lr_, end - start))
if step % 5 == 2:
Log.add_loss(str(step) + "\t" + str(loss_))
if (step + 1) % config.save_step == 0:
Log.add_log("message:当前运行模型保存, step=" + str(step) + ", lr=" +
str(lr_))
saver.save(sess,
path.join(config.model_path, config.model_name),
global_step=0)
Log.add_log("message:训练完成保存模型, step=" + str(step))
saver.save(sess,
path.join(config.model_path, config.model_name),
global_step=step)
return 0
def backward():
yolo = YOLO(config.class_num,
config.anchors,
width=config.width,
height=config.height)
data = Data(config.train_file,
config.class_num,
config.batch_size,
config.anchors,
config.data_augment,
width=config.width,
height=config.height,
data_debug=config.data_debug)
inputs = tf.compat.v1.placeholder(dtype=tf.float32,
shape=[config.batch_size, None, None, 3])
y1_true = tf.compat.v1.placeholder(
dtype=tf.float32,
shape=[config.batch_size, None, None, 3, 4 + 1 + config.class_num])
y2_true = tf.compat.v1.placeholder(
dtype=tf.float32,
shape=[config.batch_size, None, None, 3, 4 + 1 + config.class_num])
y3_true = tf.compat.v1.placeholder(
dtype=tf.float32,
shape=[config.batch_size, None, None, 3, 4 + 1 + config.class_num])
feature_y1, feature_y2, feature_y3 = yolo.forward(
inputs, weight_decay=config.weight_decay, isTrain=True)
global_step = tf.Variable(0, trainable=False)
# loss value of yolov4
loss = yolo.get_loss_v4(feature_y1, feature_y2, feature_y3, y1_true,
y2_true, y3_true, config.cls_normalizer,
config.ignore_thresh, config.prob_thresh,
config.score_thresh)
l2_loss = tf.compat.v1.losses.get_regularization_loss()
epoch = compute_curr_epoch(global_step, config.batch_size,
len(data.imgs_path))
lr = config_lr(config.lr_type, config.lr_init, epoch)
optimizer = config_optimizer(config.optimizer_type, lr, config.momentum)
update_ops = tf.compat.v1.get_collection(tf.compat.v1.GraphKeys.UPDATE_OPS)
with tf.control_dependencies(update_ops):
gvs = optimizer.compute_gradients(loss + l2_loss)
clip_grad_var = [
gv if gv[0] is None else [tf.clip_by_norm(gv[0], 100.), gv[1]]
for gv in gvs
]
train_step = optimizer.apply_gradients(clip_grad_var,
global_step=global_step)
# initialize
init = tf.compat.v1.global_variables_initializer()
saver = tf.compat.v1.train.Saver()
with tf.compat.v1.Session() as sess:
sess.run(init)
step = 0
ckpt = tf.compat.v1.train.get_checkpoint_state(config.model_path)
if ckpt and ckpt.model_checkpoint_path:
saver.restore(sess, ckpt.model_checkpoint_path)
step = ckpt.model_checkpoint_path.split('/')[-1].split('-')[-1]
step = eval(step)
Log.add_log("message: load ckpt model, global_step=" + str(step))
else:
Log.add_log("message: can not find ckpt model")
total_steps = np.ceil(config.total_epoch * len(data.imgs_path) /
config.batch_size)
while step < total_steps:
start = time.perf_counter()
batch_img, y1, y2, y3 = next(data)
_, loss_, step, lr_ = sess.run([train_step, loss, global_step, lr],
feed_dict={
inputs: batch_img,
y1_true: y1,
y2_true: y2,
y3_true: y3
})
end = time.perf_counter()
print(
"step: %6d, loss: %.5g\t, w: %3d, h: %3d, lr:%.5g\t, time: %5f s"
% (step, loss_, data.width, data.height, lr_, end - start))
if (loss_ > 1e3) and (step > 1e3):
Log.add_log("error:loss exception, loss_value = " + str(loss_))
''' break the process or lower learning rate '''
raise ValueError("error:loss exception, loss_value = " +
str(loss_) +
", please lower your learning rate")
# lr = tf.math.maximum(tf.math.divide(lr, 10), config.lr_lower)
if step % 5 == 2:
Log.add_loss(str(step) + "\t" + str(loss_))
if (step + 1) % config.save_step == 0:
# save ckpt model
Log.add_log("message: save ckpt model, step=" + str(step) +
", lr=" + str(lr_))
saver.save(sess,
path.join(config.model_path, config.model_name),
global_step=step)
# save ckpt model
Log.add_log("message:save final ckpt model, step=" + str(step))
saver.save(sess,
path.join(config.model_path, config.model_name),
global_step=step)
return 0