!10534 add export to textrcnn, remove useless code in eval

From: @chenmai1102
Reviewed-by: @oacjiewen,@c_34
Signed-off-by: @c_34

model_zoo/research/nlp/textrcnn/data_helpers.py

@@ -23,7 +23,6 @@ parser.add_argument('--data_dir', type=str, help='the source dataset directory.'
 parser.add_argument('--out_dir', type=str, help='the target dataset directory.', default='./data')
 
 args = parser.parse_args()
-np.random.seed(2)
 
 
 def dataset_split(label):
@@ -34,6 +33,7 @@ def dataset_split(label):
     pfhand = open(pos_file, encoding='utf-8')
     pos_samples += pfhand.readlines()
     pfhand.close()
+    np.random.seed(0)
     perm = np.random.permutation(len(pos_samples))
     perm_train = perm[0:int(len(pos_samples) * 0.9)]
     perm_test = perm[int(len(pos_samples) * 0.9):]

model_zoo/research/nlp/textrcnn/eval.py

@@ -48,13 +48,12 @@ if __name__ == '__main__':
     network = textrcnn(weight=Tensor(embedding_table), vocab_size=embedding_table.shape[0],
                        cell=cfg.cell, batch_size=cfg.batch_size)
     loss = nn.SoftmaxCrossEntropyWithLogits(sparse=True)
-    opt = nn.Momentum(network.trainable_params(), cfg.lr, cfg.momentum)
     loss_cb = LossMonitor()
     print("============== Starting Testing ==============")
-    ds_eval = create_dataset(cfg.preprocess_path, cfg.batch_size, 1, False)
+    ds_eval = create_dataset(cfg.preprocess_path, cfg.batch_size, False)
     param_dict = load_checkpoint(args.ckpt_path)
     load_param_into_net(network, param_dict)
     network.set_train(False)
-    model = Model(network, loss, opt, metrics={'acc': Accuracy()}, amp_level='O3')
+    model = Model(network, loss, metrics={'acc': Accuracy()}, amp_level='O3')
     acc = model.eval(ds_eval, dataset_sink_mode=False)
     print("============== Accuracy:{} ==============".format(acc))

model_zoo/research/nlp/textrcnn/export.py

@@ -0,0 +1,49 @@
+# 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.
+# ============================================================================
+"""textrcnn export ckpt file to mindir/air"""
+import os
+import argparse
+import numpy as np
+from mindspore import Tensor, context, load_checkpoint, load_param_into_net, export
+
+from src.textrcnn import textrcnn
+from src.config import textrcnn_cfg as config
+
+parser = argparse.ArgumentParser(description="textrcnn")
+parser.add_argument("--device_id", type=int, default=0, help="Device id")
+parser.add_argument("--ckpt_file", type=str, required=True, help="textrcnn ckpt file.")
+parser.add_argument("--file_name", type=str, default="textrcnn", help="textrcnn output file name.")
+parser.add_argument("--file_format", type=str, choices=["AIR", "MINDIR"],
+                    default="MINDIR", help="file format")
+parser.add_argument("--device_target", type=str, choices=["Ascend"], default="Ascend",
+                    help="device target")
+args = parser.parse_args()
+
+context.set_context(mode=context.GRAPH_MODE, device_target=args.device_target, device_id=args.device_id)
+
+if __name__ == "__main__":
+    # define net
+    embedding_table = np.loadtxt(os.path.join(config.preprocess_path, "weight.txt")).astype(np.float32)
+
+    net = textrcnn(weight=Tensor(embedding_table), vocab_size=embedding_table.shape[0],
+                   cell=config.cell, batch_size=config.batch_size)
+
+    # load checkpoint
+    param_dict = load_checkpoint(args.ckpt_file)
+    load_param_into_net(net, param_dict)
+    net.set_train(False)
+
+    image = Tensor(np.ones([config.batch_size, 50], np.int32))
+    export(net, image, file_name=args.file_name, file_format=args.file_format)

model_zoo/research/nlp/textrcnn/readme.md

@@ -100,6 +100,7 @@ bash scripts/run_eval.sh
         │   ├──textrcnn.py          // textrcnn architecture
         │   ├──config.py            // parameter configuration
         ├── train.py               // training script
+        ├── export.py             // export script
         ├── eval.py               //  evaluation script
         ├── data_helpers.py               //  dataset split script
         ├── sample.txt               //  the shell to train and eval the model without scripts
@@ -129,8 +130,7 @@ Parameters for both training and evaluation can be set in config.py
   'emb_path': './word2vec', # the directory to save the embedding file
   'embed_size': 300, # the dimension of the word embedding
   'save_checkpoint_steps': 149, # per step to save the checkpoint
-  'keep_checkpoint_max': 10, # max checkpoints to save
-  'momentum': 0.9 # the momentum rate
+  'keep_checkpoint_max': 10 # max checkpoints to save
   ```
 
 ### Performance

model_zoo/research/nlp/textrcnn/src/config.py

@@ -39,5 +39,4 @@ textrcnn_cfg = edict({
     'embed_size': 300,
     'save_checkpoint_steps': 149,
     'keep_checkpoint_max': 10,
-    'momentum': 0.9
 })

model_zoo/research/nlp/textrcnn/src/dataset.py

@@ -76,9 +76,7 @@ def tokenizer(text):
 def collect_weight(glove_path, vocab, word_to_idx, embed_size):
     """ collect weight """
     vocab_size = len(vocab)
-    # wvmodel = gensim.models.KeyedVectors.load_word2vec_format(os.path.join(glove_path, 'glove.6B.300d.txt'),
-    #                                                           binary=False, encoding='utf-8')
-    wvmodel = gensim.models.KeyedVectors.load_word2vec_format(os.path.join(glove_path, \
+    wvmodel = gensim.models.KeyedVectors.load_word2vec_format(os.path.join(glove_path,
                                                                            'GoogleNews-vectors-negative300.bin'),
                                                               binary=True)
     weight_np = np.zeros((vocab_size + 1, embed_size)).astype(np.float32)
@@ -164,7 +162,7 @@ def convert_to_mindrecord(embed_size, data_path, proprocess_path, glove_path):
     writer.commit()
 
 
-def create_dataset(base_path, batch_size, num_epochs, is_train):
+def create_dataset(base_path, batch_size, is_train):
     """Create dataset for training."""
     columns_list = ["feature", "label"]
     num_consumer = 4
@@ -175,7 +173,7 @@ def create_dataset(base_path, batch_size, num_epochs, is_train):
         path = os.path.join(base_path, 'aclImdb_test.mindrecord0')
 
     data_set = ds.MindDataset(path, columns_list, num_consumer)
-    ds.config.set_seed(1)
+    ds.config.set_seed(0)
     data_set = data_set.shuffle(buffer_size=data_set.get_dataset_size())
     data_set = data_set.batch(batch_size, drop_remainder=True)
     return data_set

model_zoo/research/nlp/textrcnn/src/textrcnn.py

@@ -47,16 +47,16 @@ class textrcnn(nn.Cell):
             self.lstm = P.DynamicRNN(forget_bias=0.0)
             self.w1_fw = Parameter(
                 np.random.uniform(-k, k, (self.embed_size + self.num_hiddens, 4 * self.num_hiddens)).astype(
-                    np.float32), name="w1_fw")
-            self.b1_fw = Parameter(np.random.uniform(-k, k, (4 * self.num_hiddens)).astype(np.float32),
+                    np.float16), name="w1_fw")
+            self.b1_fw = Parameter(np.random.uniform(-k, k, (4 * self.num_hiddens)).astype(np.float16),
                                    name="b1_fw")
             self.w1_bw = Parameter(
                 np.random.uniform(-k, k, (self.embed_size + self.num_hiddens, 4 * self.num_hiddens)).astype(
-                    np.float32), name="w1_bw")
-            self.b1_bw = Parameter(np.random.uniform(-k, k, (4 * self.num_hiddens)).astype(np.float32),
+                    np.float16), name="w1_bw")
+            self.b1_bw = Parameter(np.random.uniform(-k, k, (4 * self.num_hiddens)).astype(np.float16),
                                    name="b1_bw")
-            self.h1 = Tensor(np.zeros(shape=(1, self.batch_size, self.num_hiddens)).astype(np.float32))
-            self.c1 = Tensor(np.zeros(shape=(1, self.batch_size, self.num_hiddens)).astype(np.float32))
+            self.h1 = Tensor(np.zeros(shape=(1, self.batch_size, self.num_hiddens)).astype(np.float16))
+            self.c1 = Tensor(np.zeros(shape=(1, self.batch_size, self.num_hiddens)).astype(np.float16))
 
         if cell == "vanilla":
             self.rnnW_fw = nn.Dense(self.num_hiddens, self.num_hiddens)
@@ -72,6 +72,12 @@ class textrcnn(nn.Cell):
             self.rnnWz_bw = nn.Dense(self.num_hiddens + self.embed_size, self.num_hiddens)
             self.rnnWh_bw = nn.Dense(self.num_hiddens + self.embed_size, self.num_hiddens)
             self.ones = Tensor(np.ones(shape=(self.batch_size, self.num_hiddens)).astype(np.float16))
+            self.rnnWr_fw.to_float(mstype.float16)
+            self.rnnWz_fw.to_float(mstype.float16)
+            self.rnnWh_fw.to_float(mstype.float16)
+            self.rnnWr_bw.to_float(mstype.float16)
+            self.rnnWz_bw.to_float(mstype.float16)
+            self.rnnWh_bw.to_float(mstype.float16)
 
         self.transpose = P.Transpose()
         self.reduce_max = P.ReduceMax()
@@ -91,6 +97,9 @@ class textrcnn(nn.Cell):
         self.tanh = P.Tanh()
         self.sigmoid = P.Sigmoid()
         self.slice = P.Slice()
+        self.text_rep_dense.to_float(mstype.float16)
+        self.mydense.to_float(mstype.float16)
+        self.output_dense.to_float(mstype.float16)
 
     def construct(self, x):
         """class construction"""

model_zoo/research/nlp/textrcnn/train.py

@@ -22,7 +22,7 @@ import mindspore.context as context
 from mindspore import Tensor
 from mindspore.train import Model
 from mindspore.nn.metrics import Accuracy
-from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, LossMonitor
+from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, LossMonitor, TimeMonitor
 from mindspore.common import set_seed
 
 from src.config import textrcnn_cfg as cfg
@@ -31,7 +31,7 @@ from src.dataset import convert_to_mindrecord
 from src.textrcnn import textrcnn
 from src.utils import get_lr
 
-set_seed(2)
+set_seed(0)
 
 if __name__ == '__main__':
 
@@ -58,7 +58,7 @@ if __name__ == '__main__':
     network = textrcnn(weight=Tensor(embedding_table), vocab_size=embedding_table.shape[0],
                        cell=cfg.cell, batch_size=cfg.batch_size)
 
-    ds_train = create_dataset(cfg.preprocess_path, cfg.batch_size, cfg.num_epochs, True)
+    ds_train = create_dataset(cfg.preprocess_path, cfg.batch_size, True)
     step_size = ds_train.get_dataset_size()
 
     loss = nn.SoftmaxCrossEntropyWithLogits(sparse=True)
@@ -70,11 +70,12 @@ if __name__ == '__main__':
     opt = nn.Adam(params=network.trainable_params(), learning_rate=lr)
 
     loss_cb = LossMonitor()
+    time_cb = TimeMonitor()
     model = Model(network, loss, opt, {'acc': Accuracy()}, amp_level="O3")
 
     print("============== Starting Training ==============")
     config_ck = CheckpointConfig(save_checkpoint_steps=cfg.save_checkpoint_steps,
                                  keep_checkpoint_max=cfg.keep_checkpoint_max)
     ckpoint_cb = ModelCheckpoint(prefix=cfg.cell, directory=cfg.ckpt_folder_path, config=config_ck)
-    model.train(num_epochs, ds_train, callbacks=[ckpoint_cb, loss_cb])
+    model.train(num_epochs, ds_train, callbacks=[ckpoint_cb, loss_cb, time_cb])
     print("train success")