modify

model_zoo/deeplabv3/evaluation.py

@@ -1,4 +1,3 @@
-#!/bin/bash
 # Copyright 2020 Huawei Technologies Co., Ltd
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
@@ -14,18 +13,10 @@
 # limitations under the License.
 # ============================================================================
 """evaluation."""
-import os, time
 import argparse
 from mindspore import context
-from mindspore import log as logger
-from mindspore.communication.management import init
-import mindspore.nn as nn
-from mindspore.nn.optim.momentum import Momentum
-from mindspore.train.loss_scale_manager import FixedLossScaleManager
-from mindspore import Model, ParallelMode
-import argparse
+from mindspore import Model
 from mindspore.train.serialization import load_checkpoint, load_param_into_net
-from mindspore.train.callback import Callback,CheckpointConfig, ModelCheckpoint, TimeMonitor
 from src.md_dataset import create_dataset
 from src.losses import OhemLoss
 from src.miou_precision import MiouPrecision
@@ -56,3 +47,4 @@ if __name__ == "__main__":
     loss = OhemLoss(config.seg_num_classes, config.ignore_label)
     model = Model(net, loss, metrics=metrics)
     model.eval(eval_dataset)
+    

model_zoo/deeplabv3/train.py

@@ -1,4 +1,3 @@
-#!/bin/bash
 # Copyright 2020 Huawei Technologies Co., Ltd
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
@@ -14,16 +13,11 @@
 # limitations under the License.
 # ============================================================================
 """train."""
-import os, time
 import argparse
 from mindspore import context
-from mindspore import log as logger
 from mindspore.communication.management import init
-import mindspore.nn as nn
 from mindspore.nn.optim.momentum import Momentum
-from mindspore.train.loss_scale_manager import FixedLossScaleManager
 from mindspore import Model, ParallelMode
-import argparse
 from mindspore.train.serialization import load_checkpoint, load_param_into_net
 from mindspore.train.callback import Callback, CheckpointConfig, ModelCheckpoint, TimeMonitor
 from src.md_dataset import create_dataset
@@ -40,8 +34,7 @@ parser.add_argument("--device_id", type=int, default=0, help="Device id, default
 parser.add_argument('--checkpoint_url', default=None, help='Checkpoint path')
 parser.add_argument("--enable_save_ckpt", type=str, default="true", help="Enable save checkpoint, default is true.")
 parser.add_argument('--max_checkpoint_num', type=int, default=5, help='Max checkpoint number.')
-parser.add_argument("--save_checkpoint_steps", type=int, default=1000, help="Save checkpoint steps, "
-                                                                                "default is 1000.")
+parser.add_argument("--save_checkpoint_steps", type=int, default=1000, help="Save checkpoint steps, default is 1000.")
 parser.add_argument("--save_checkpoint_num", type=int, default=1, help="Save checkpoint numbers, default is 1.")
 args_opt = parser.parse_args()
 print(args_opt)
@@ -63,13 +56,13 @@ class LossCallBack(Callback):
         cb_params = run_context.original_args()
         print("epoch: {}, step: {}, outputs are {}".format(cb_params.cur_epoch_num, cb_params.cur_step_num,
                                                            str(cb_params.net_outputs)))
-def model_fine_tune(flags, net, fix_weight_layer):
+def model_fine_tune(flags, train_net, fix_weight_layer):
     checkpoint_path = flags.checkpoint_url
     if checkpoint_path is None:
         return
     param_dict = load_checkpoint(checkpoint_path)
-    load_param_into_net(net, param_dict)
-    for para in net.trainable_params():
+    load_param_into_net(train_net, param_dict)
+    for para in train_net.trainable_params():
         if fix_weight_layer in para.name:
             para.requires_grad = False
 if __name__ == "__main__":
@@ -97,3 +90,4 @@ if __name__ == "__main__":
     opt = Momentum(filter(lambda x: 'beta' not in x.name and 'gamma' not in x.name and 'depth' not in x.name and 'bias' not in x.name, net.trainable_params()), learning_rate=config.learning_rate, momentum=config.momentum, weight_decay=config.weight_decay)
     model = Model(net, loss, opt)
     model.train(args_opt.epoch_size, train_dataset, callback)
+