!10134 improve ncf accuracy.

From: @linqingke
Reviewed-by: @wuxuejian,@liangchenghui
Signed-off-by: @liangchenghui

model_zoo/official/cv/psenet/src/config.py

@@ -37,7 +37,6 @@ config = ed({
 
     # dataset for train
     "TRAIN_ROOT_DIR": "psenet/ic15/",
-    "TRAIN_IS_TRANSFORM": True,
     "TRAIN_LONG_SIZE": 640,
     "TRAIN_MIN_SCALE": 0.4,
     "TRAIN_BATCH_SIZE": 4,

model_zoo/official/cv/psenet/src/dataset.py

@@ -160,7 +160,7 @@ def shrink(bboxes, rate, max_shr=20):
 
 class TrainDataset:
     def __init__(self):
-        self.is_transform = config.TRAIN_IS_TRANSFORM
+        self.is_transform = True
         self.img_size = config.TRAIN_LONG_SIZE
         self.kernel_num = config.KERNEL_NUM
         self.min_scale = config.TRAIN_MIN_SCALE

model_zoo/official/recommend/ncf/src/lr_schedule.py

@@ -0,0 +1,37 @@
+# Copyright 2020 Huawei Technologies Co., Ltd
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ============================================================================
+"""lr generator for ncf"""
+import math
+
+def _linear_warmup_learning_rate(current_step, warmup_steps, base_lr, init_lr):
+    lr_inc = (float(base_lr) - float(init_lr)) / float(warmup_steps)
+    learning_rate = float(init_lr) + lr_inc * current_step
+    return learning_rate
+
+def _cosine_learning_rate(current_step, base_lr, warmup_steps, decay_steps):
+    base = float(current_step - warmup_steps) / float(decay_steps)
+    learning_rate = (1 + math.cos(base * math.pi)) / 2 * base_lr
+    return learning_rate
+
+def dynamic_lr(base_lr, total_steps, warmup_steps):
+    """dynamic learning rate generator"""
+    lr = []
+    for i in range(total_steps):
+        if i < warmup_steps:
+            lr.append(_linear_warmup_learning_rate(i, warmup_steps, base_lr, base_lr * 0.01))
+        else:
+            lr.append(_cosine_learning_rate(i, base_lr, warmup_steps, total_steps))
+
+    return lr

model_zoo/official/recommend/ncf/src/ncf.py

@@ -26,6 +26,8 @@ from mindspore.parallel._utils import _get_device_num, _get_parallel_mode, _get_
 from mindspore.context import ParallelMode
 from mindspore.nn.wrap.grad_reducer import DistributedGradReducer
 
+from src.lr_schedule import dynamic_lr
+
 class DenseLayer(nn.Cell):
     """
     Dense layer definition
@@ -223,14 +225,16 @@ class TrainStepWrap(nn.Cell):
     """
     TrainStepWrap definition
     """
-    def __init__(self, network, sens=16384.0):
+    def __init__(self, network, total_steps=1, sens=16384.0):
         super(TrainStepWrap, self).__init__(auto_prefix=False)
         self.network = network
         self.network.set_train()
         self.network.add_flags(defer_inline=True)
         self.weights = ParameterTuple(network.trainable_params())
+
+        lr = dynamic_lr(0.01, total_steps, 5000)
         self.optimizer = nn.Adam(self.weights,
-                                 learning_rate=0.00382059,
+                                 learning_rate=lr,
                                  beta1=0.9,
                                  beta2=0.999,
                                  eps=1e-8,

model_zoo/official/recommend/ncf/train.py

@@ -22,12 +22,15 @@ from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, LossMoni
 from mindspore import context, Model
 from mindspore.context import ParallelMode
 from mindspore.communication.management import get_rank, get_group_size, init
+from mindspore.common import set_seed
 
 from src.dataset import create_dataset
 from src.ncf import NCFModel, NetWithLossClass, TrainStepWrap
 
 from config import cfg
 
+set_seed(1)
+
 logging.set_verbosity(logging.INFO)
 
 parser = argparse.ArgumentParser(description='NCF')
@@ -86,7 +89,7 @@ def test_train():
                        mlp_reg_layers=[0.0, 0.0, 0.0, 0.0],
                        mf_dim=16)
     loss_net = NetWithLossClass(ncf_net)
-    train_net = TrainStepWrap(loss_net)
+    train_net = TrainStepWrap(loss_net, ds_train.get_dataset_size() * (epochs + 1))
 
     train_net.set_train()