From 7bd6edbdc478ea1ce3e1d3ce17973b584eb50085 Mon Sep 17 00:00:00 2001 From: zhangyi Date: Fri, 19 Jun 2020 15:52:13 +0800 Subject: [PATCH] add content that supporting CPU environment --- .../source_en/advanced_use/nlp_application.md | 40 +++++++++++------- .../advanced_use/nlp_application.md | 42 ++++++++++++------- 2 files changed, 51 insertions(+), 31 deletions(-) diff --git a/tutorials/source_en/advanced_use/nlp_application.md b/tutorials/source_en/advanced_use/nlp_application.md index becfac943f..e4aa6bda58 100644 --- a/tutorials/source_en/advanced_use/nlp_application.md +++ b/tutorials/source_en/advanced_use/nlp_application.md @@ -79,9 +79,9 @@ In the IMDb dataset, the number of positive and negative samples does not vary g ### Determining the Network and Process -Currently, MindSpore GPU supports the long short-term memory (LSTM) network for NLP. +Currently, MindSpore GPU and CPU supports SentimentNet network based on the long short-term memory (LSTM) network for NLP. 1. Load the dataset in use and process data. -2. Use the LSTM network training data to generate a model. +2. Use the SentimentNet network based on LSTM training data to generate a model. Long short-term memory (LSTM) is an artificial recurrent neural network (RNN) architecture used for processing and predicting an important event with a long interval and delay in a time sequence. For details, refer to online documentation. 3. After the model is obtained, use the validation dataset to check the accuracy of model. @@ -102,8 +102,6 @@ import json from itertools import chain import numpy as np from config import lstm_cfg as cfg -# Install gensim with 'pip install gensim' -import gensim import mindspore.nn as nn import mindspore.context as context @@ -116,7 +114,9 @@ from mindspore.mindrecord import FileWriter from mindspore.train import Model from mindspore.nn.metrics import Accuracy from mindspore.train.serialization import load_checkpoint, load_param_into_net -from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, LossMonitor +from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, LossMonitor, TimeMonitor +# Install gensim with 'pip install gensim' +import gensim ``` ### Configuring Environment Information @@ -125,22 +125,25 @@ from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, LossMoni ```python parser = argparse.ArgumentParser(description='MindSpore LSTM Example') parser.add_argument('--preprocess', type=str, default='false', choices=['true', 'false'], - help='Whether to perform data preprocessing') + help='whether to preprocess data.') parser.add_argument('--mode', type=str, default="train", choices=['train', 'test'], help='implement phase, set to train or test') # Download dataset from 'http://ai.stanford.edu/~amaas/data/sentiment/aclImdb_v1.tar.gz' and extract to 'aclimdb_path' parser.add_argument('--aclimdb_path', type=str, default="./aclImdb", - help='path where the dataset is store') + help='path where the dataset is stored') # Download GloVe from 'http://nlp.stanford.edu/data/glove.6B.zip' and extract to 'glove_path' # Add a new line '400000 300' at the beginning of 'glove.6B.300d.txt' with '40000' for total words and '300' for vector length parser.add_argument('--glove_path', type=str, default="./glove", - help='path where the GloVe is store') + help='path where the GloVe is stored') # Specify the path to save preprocessed data parser.add_argument('--preprocess_path', type=str, default="./preprocess", - help='path where the pre-process data is store') + help='path where the pre-process data is stored') # Specify the path to save the CheckPoint file - parser.add_argument('--ckpt_path', type=str, default="./ckpt", help='if mode is test, must provide\ - path where the trained ckpt file') + parser.add_argument('--ckpt_path', type=str, default="./", + help='if mode is test, must provide path where the trained ckpt file.') + # Specify the target device to run + parser.add_argument('--device_target', type=str, default="GPU", choices=['GPU', 'CPU'], + help='the target device to run, support "GPU", "CPU". Default: "GPU".') args = parser.parse_args() ``` @@ -150,7 +153,7 @@ from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, LossMoni context.set_context( mode=context.GRAPH_MODE, save_graphs=False, - device_target="GPU") + device_target=args.device_target) ``` For details about the API configuration, see the `context.set_context`. @@ -458,7 +461,11 @@ ds_train = create_dataset(args.preprocess_path, cfg.batch_size, cfg.num_epochs, config_ck = CheckpointConfig(save_checkpoint_steps=cfg.save_checkpoint_steps, keep_checkpoint_max=cfg.keep_checkpoint_max) ckpoint_cb = ModelCheckpoint(prefix="lstm", directory=args.ckpt_path, config=config_ck) -model.train(cfg.num_epochs, ds_train, callbacks=[ckpoint_cb, loss_cb]) +time_cb = TimeMonitor(data_size=ds_train.get_dataset_size()) +if args.device_target == "CPU": + model.train(cfg.num_epochs, ds_train, callbacks=[time_cb, ckpoint_cb, loss_cb], dataset_sink_mode=False) +else: + model.train(cfg.num_epochs, ds_train, callbacks=[time_cb, ckpoint_cb, loss_cb]) ``` ### Validating the Model @@ -470,7 +477,10 @@ print("============== Starting Testing ==============") ds_eval = create_dataset(args.preprocess_path, cfg.batch_size, 1, False) param_dict = load_checkpoint(args.ckpt_path) load_param_into_net(network, param_dict) -acc = model.eval(ds_eval) +if args.device_target == "CPU": + acc = model.eval(ds_eval, dataset_sink_mode=False) +else: + acc = model.eval(ds_eval) print("============== Accuracy:{} ==============".format(acc)) ``` @@ -480,7 +490,7 @@ After 10 epochs, the accuracy on the training set converges to about 85%, and th **Training Execution** 1. Run the training code and view the running result. ```shell - $ python main.py --preprocess=true --mode=train --ckpt_path=./ckpt + $ python main.py --preprocess=true --mode=train --ckpt_path=./ckpt --device_target=GPU ``` As shown in the following output, the loss value decreases gradually with the training process and reaches about 0.249. That is, after 10 epochs of training, the accuracy of the current text analysis result is about 85%. diff --git a/tutorials/source_zh_cn/advanced_use/nlp_application.md b/tutorials/source_zh_cn/advanced_use/nlp_application.md index d505a65e60..84265b3d9f 100644 --- a/tutorials/source_zh_cn/advanced_use/nlp_application.md +++ b/tutorials/source_zh_cn/advanced_use/nlp_application.md @@ -79,13 +79,13 @@ $F1分数 = (2 * Precision * Recall) / (Precision + Recall)$ ### 确定网络及流程 -我们使用LSTM网络进行自然语言处理。 +我们使用基于LSTM构建的SentimentNet网络进行自然语言处理。 1. 加载使用的数据集，并进行必要的数据处理。 -2. 使用LSTM网络训练数据，生成模型。 +2. 使用基于LSTM构建的SentimentNet网络训练数据，生成模型。 > LSTM（Long short-term memory，长短期记忆）网络是一种时间循环神经网络，适合于处理和预测时间序列中间隔和延迟非常长的重要事件。具体介绍可参考网上资料，在此不再赘述。 3. 得到模型之后，使用验证数据集，查看模型精度情况。 -> 本例面向GPU硬件平台，你可以在这里下载完整的样例代码： +> 本例面向GPU或CPU硬件平台，你可以在这里下载完整的样例代码： > - `main.py`：代码文件，包括数据预处理、网络定义、模型训练等代码。 > - `config.py`：网络中的一些配置，包括`batch size`、进行几次epoch训练等。 @@ -101,8 +101,6 @@ import json from itertools import chain import numpy as np from config import lstm_cfg as cfg -# Install gensim with 'pip install gensim' -import gensim import mindspore.nn as nn import mindspore.context as context @@ -115,7 +113,9 @@ from mindspore.mindrecord import FileWriter from mindspore.train import Model from mindspore.nn.metrics import Accuracy from mindspore.train.serialization import load_checkpoint, load_param_into_net -from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, LossMonitor +from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, LossMonitor, TimeMonitor +# Install gensim with 'pip install gensim' +import gensim ``` ### 配置环境信息 @@ -124,22 +124,25 @@ from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, LossMoni ```python parser = argparse.ArgumentParser(description='MindSpore LSTM Example') parser.add_argument('--preprocess', type=str, default='false', choices=['true', 'false'], - help='Whether to perform data preprocessing') + help='whether to preprocess data.') parser.add_argument('--mode', type=str, default="train", choices=['train', 'test'], help='implement phase, set to train or test') # Download dataset from 'http://ai.stanford.edu/~amaas/data/sentiment/aclImdb_v1.tar.gz' and extract to 'aclimdb_path' parser.add_argument('--aclimdb_path', type=str, default="./aclImdb", - help='path where the dataset is store') + help='path where the dataset is stored') # Download GloVe from 'http://nlp.stanford.edu/data/glove.6B.zip' and extract to 'glove_path' # Add a new line '400000 300' at the beginning of 'glove.6B.300d.txt' with '40000' for total words and '300' for vector length parser.add_argument('--glove_path', type=str, default="./glove", - help='path where the GloVe is store') + help='path where the GloVe is stored') # Specify the path to save preprocessed data parser.add_argument('--preprocess_path', type=str, default="./preprocess", - help='path where the pre-process data is store') + help='path where the pre-process data is stored') # Specify the path to save the CheckPoint file - parser.add_argument('--ckpt_path', type=str, default="./ckpt", help='if mode is test, must provide\ - path where the trained ckpt file') + parser.add_argument('--ckpt_path', type=str, default="./", + help='if mode is test, must provide path where the trained ckpt file.') + # Specify the target device to run + parser.add_argument('--device_target', type=str, default="GPU", choices=['GPU', 'CPU'], + help='the target device to run, support "GPU", "CPU". Default: "GPU".') args = parser.parse_args() ``` @@ -149,7 +152,7 @@ from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, LossMoni context.set_context( mode=context.GRAPH_MODE, save_graphs=False, - device_target="GPU") + device_target=args.device_target) ``` 详细的接口配置信息，请参见`context.set_context`接口说明。 @@ -457,7 +460,11 @@ ds_train = create_dataset(args.preprocess_path, cfg.batch_size, cfg.num_epochs, config_ck = CheckpointConfig(save_checkpoint_steps=cfg.save_checkpoint_steps, keep_checkpoint_max=cfg.keep_checkpoint_max) ckpoint_cb = ModelCheckpoint(prefix="lstm", directory=args.ckpt_path, config=config_ck) -model.train(cfg.num_epochs, ds_train, callbacks=[ckpoint_cb, loss_cb]) +time_cb = TimeMonitor(data_size=ds_train.get_dataset_size()) +if args.device_target == "CPU": + model.train(cfg.num_epochs, ds_train, callbacks=[time_cb, ckpoint_cb, loss_cb], dataset_sink_mode=False) +else: + model.train(cfg.num_epochs, ds_train, callbacks=[time_cb, ckpoint_cb, loss_cb]) ``` ### 模型验证 @@ -469,7 +476,10 @@ print("============== Starting Testing ==============") ds_eval = create_dataset(args.preprocess_path, cfg.batch_size, 1, False) param_dict = load_checkpoint(args.ckpt_path) load_param_into_net(network, param_dict) -acc = model.eval(ds_eval) +if args.device_target == "CPU": + acc = model.eval(ds_eval, dataset_sink_mode=False) +else: + acc = model.eval(ds_eval) print("============== Accuracy:{} ==============".format(acc)) ``` @@ -479,7 +489,7 @@ print("============== Accuracy:{} ==============".format(acc)) **执行训练** 1. 运行训练代码，查看运行结果。 ```shell - $ python main.py --preprocess=true --mode=train --ckpt_path=./ckpt + $ python main.py --preprocess=true --mode=train --ckpt_path=./ckpt --device_target=GPU ``` 输出如下，可以看到loss值随着训练逐步降低，最后达到0.249左右，即经过10个epoch的训练，对当前文本分析的结果正确率在85%左右： -- Gitee