TextCNN is an algorithm that uses convolutional neural networks to classify text. It was proposed by Yoon Kim in the article "Convolutional Neural Networks for Sentence Classification" in 2014. It is widely used in various tasks of text classification (such as sentiment analysis). It has become the standard benchmark for the new text classification framework. Each module of TextCNN can complete text classification tasks independently, and it is convenient for distributed configuration and parallel execution. TextCNN is very suitable for the semantic analysis of short texts such as Weibo/News/E-commerce reviews and video bullet screens.
Paper: Kim Y. Convolutional neural networks for sentence classification[J]. arXiv preprint arXiv:1408.5882, 2014.
The basic network structure design of TextCNN can refer to the paper "Convolutional Neural Networks for Sentence Classification". The specific implementation takes reading a sentence "I like this movie very much!" as an example. First, the word segmentation algorithm is used to divide the words into 7 words, and then the words in each part are expanded into a five-dimensional vector through the embedding method. Then use different convolution kernels ([3,4,5]*5) to perform convolution operations on them to obtain feature maps. The default number of convolution kernels is 2. Then use the maxpool operation to pool all the feature maps, and finally merge the pooling result into a one-dimensional feature vector through the connection operation. At last, it can be divided into 2 categories with softmax, and the positive/negative emotions are obtained.
Note that you can run the scripts based on the dataset mentioned in original paper or widely used in relevant domain/network architecture. In the following sections, we will introduce how to run the scripts using the related dataset below.
Dataset used:
data
directory.After installing MindSpore via the official website, you can start training and evaluation as follows:
running on Ascend/CPU
# run training example
# need set config_path in config.py file and set data_path in yaml file
python train.py --config_path [CONFIG_PATH] \
--device_target [TARGET] \
--data_path [DATA_PATH]> train.log 2>&1 &
OR
# Ascend
bash scripts/run_train.sh [DATASET]
# CPU
bash scripts/run_train_cpu.sh
# run evaluation example
# need set config_path in config.py file and set data_path, checkpoint_file_path in yaml file
python eval.py --config_path [CONFIG_PATH] \
--device_target [TARGET] \
--checkpoint_file_path [CKPT_FILE] \
--data_path [DATA_PATH] > eval.log 2>&1 &
OR
# Ascend
bash scripts/run_eval.sh [CKPT_FILE] [DATASET]
# CPU
bash scripts/run_eval_cpu.sh [CKPT_FILE]
running on GPU
# run training example
# need set config_path in config.py file and set data_path in yaml file
python train.py --config_path [CONFIG_PATH] \
--device_target GPU \
--data_path [DATA_PATH]> train.log 2>&1 &
OR
sh scripts/run_train_gpu.sh [DATASET] [DATA_PATH]
# run evaluation example
# need set config_path in config.py file and set data_path, checkpoint_file_path in yaml file
python eval.py --config_path [CONFIG_PATH] \
--device_target GPU \
--checkpoint_file_path [CKPT_FILE] \
--data_path [DATA_PATH] > eval.log 2>&1 &
OR
sh scripts/run_eval_gpu.sh [CKPT_FILE] [DATASET] [DATA_PATH]
If you want to run in modelarts, please check the official documentation of modelarts, and you can start training and evaluation as follows:
# run distributed training on modelarts example
# (1) First, Perform a or b.
# a. Set "enable_modelarts=True" on yaml file.
# Set other parameters on yaml file you need.
# b. Add "enable_modelarts=True" on the website UI interface.
# Add other parameters on the website UI interface.
# (2) Set the config_path="/path/yaml" on the website UI interface.
# (3) Set the code directory to "/path/textcnn" on the website UI interface.
# (4) Set the startup file to "train.py" on the website UI interface.
# (5) Set the "Dataset path" and "Output file path" and "Job log path" to your path on the website UI interface.
# (6) Create your job.
# run evaluation on modelarts example
# (1) Copy or upload your trained model to S3 bucket.
# (2) Perform a or b.
# a.Set "enable_modelarts=True" on yaml file.
# Set "checkpoint_file_path='/cache/checkpoint_path/model.ckpt'" on yaml file.
# Set "checkpoint_url=/The path of checkpoint in S3/" on yaml file.
# b. Add "enable_modelarts=True" on the website UI interface.
# Add "checkpoint_file_path='/cache/checkpoint_path/model.ckpt'" on the website UI interface.
# Add "checkpoint_url=/The path of checkpoint in S3/" on the website UI interface.
# (3) Set the config_path="/path/yaml" on the website UI interface
# (4) Set the code directory to "/path/textcnn" on the website UI interface.
# (5) Set the startup file to "eval.py" on the website UI interface.
# (6) Set the "Dataset path" and "Output file path" and "Job log path" to your path on the website UI interface.
# (7) Create your job.
├── model_zoo
├── README.md // descriptions about all the models
├── textcnn
├── README.md // descriptions about textcnn
├──scripts
│ ├── run_train.sh // shell script for distributed on Ascend
│ ├── run_eval.sh // shell script for evaluation on Ascend
│ ├── run_train_cpu.sh // shell script for training on CPU
│ ├── run_eval_cpu.sh // shell script for evaluation on CPU
│ ├── run_train_gpu.sh // shell script for training on GPU
│ ├── run_eval_gpu.sh // shell script for evaluation on GPU
│ ├── run_eval_onnx_gpu.sh // shell script for onnx evaluation on GPU
├── src
│ ├── dataset.py // Processing dataset
│ ├── textcnn.py // textcnn architecture
├── model_utils
│ ├──device_adapter.py // device adapter
│ ├──local_adapter.py // local adapter
│ ├──moxing_adapter.py // moxing adapter
│ ├──config.py // parameter analysis
├── mr_config.yaml // parameter configuration
├── mr_config_cpu.yaml // parameter configuration
├── sst2_config.yaml // parameter configuration
├── subj_config.yaml // parameter configuration
├── train.py // training script
├── eval.py // evaluation script
├── eval_onnx.py // onnx evaluation script
├── export.py // export checkpoint to other format file
Parameters for both training and evaluation can be set in config.py
config for movie review dataset
'pre_trained': 'False' # whether training based on the pre-trained model
'nump_classes': 2 # the number of classes in the dataset
'batch_size': 64 # training batch size
'epoch_size': 4 # total training epochs
'weight_decay': 3e-5 # weight decay value
'data_path': './data/' # absolute full path to the train and evaluation datasets
'device_target': 'Ascend' # device running the program
'device_id': 0 # device ID used to train or evaluate the dataset. Ignore it when you use run_train.sh for distributed training
'keep_checkpoint_max': 1 # only keep the last keep_checkpoint_max checkpoint
'checkpoint_path': './train_textcnn.ckpt' # the absolute full path to save the checkpoint file
'onnx_file': './sst2_textcnn.onnx' # the absolute full path to the exported onnx file
'word_len': 51 # The length of the word
'vec_length': 40 # The length of the vector
'base_lr': 1e-3 # The base learning rate
For more configuration details, please refer the script *.yaml
.
running on Ascend/CPU
# `CONFIG_PATH` `DATA_PATH` `DATASET` `DEVICE_TARGET` parameters need to be passed externally or modified yaml file
# `DATASET` must choose from ['MR', 'SUBJ', 'SST2']"
python train.py --config_path [CONFIG_PATH] \
--device_target [DEVICE_TARGET] \
--data_path [DATA_PATH]> train.log 2>&1 &
OR
# Ascend
bash scripts/run_train.sh [DATASET]
# CPU
bash scripts/run_train_cpu.sh
running on GPU
# `CONFIG_PATH` `DATA_PATH` `DATASET` parameters need to be passed externally or modified yaml file
# `DATASET` must choose from ['MR', 'SUBJ', 'SST2']"
python train.py --config_path [CONFIG_PATH] \
--device_target GPU \
--data_path [DATA_PATH]> train.log 2>&1 &
OR
bash scripts/run_train_gpu.sh [DATASET] [DATA_PATH]
The python command above will run in the background, you can view the results through the file train.log
.
After training, you'll get some checkpoint files in ckpt
. The loss value will be achieved as follows:
# grep "loss is " train.log
epoch: 1 step 149, loss is 0.6194226145744324
epoch: 2 step 149, loss is 0.38729554414749146
...
The model checkpoint will be saved in the ckpt
directory.
evaluation on movie review dataset when running on Ascend/CPU
Before running the command below, please check the checkpoint path used for evaluation. Please set the checkpoint path to be the absolute full path, e.g., "username/textcnn/ckpt/train_textcnn.ckpt".
# `CONFIG_PATH` `DEVICE_TARGET` `CKPT_FILE` `DATA_PATH` `DATASET` parameters need to be passed externally or modified yaml file
# `DATASET` must choose from ['MR', 'SUBJ', 'SST2']"
python eval.py --config_path [CONFIG_PATH] \
--device_target [DEVICE_TARGET] \
--checkpoint_file_path [CKPT_FILE] \
--data_path [DATA_PATH] > eval.log 2>&1 &
OR
# Ascend
bash scripts/run_eval.sh [CKPT_FILE] [DATASET]
# CPU
bash scripts/run_eval_cpu.sh [CKPT_FILE]
evaluation on movie review dataset when running on GPU
# `CONFIG_PATH` `CKPT_FILE` `DATA_PATH` `DATASET` parameters need to be passed externally or modified yaml file
# `DATASET` must choose from ['MR', 'SUBJ', 'SST2']"
python eval.py --config_path [CONFIG_PATH] \
--device_target GPU \
--checkpoint_file_path [CKPT_FILE] \
--data_path [DATA_PATH] > eval.log 2>&1 &
OR
bash scripts/run_eval_gpu.sh [CKPT_FILE] [DATASET] [DATA_PATH]
The above python command will run in the background. You can view the results through the file "eval.log". The accuracy of the test dataset will be as follows:
# grep "accuracy: " eval.log
accuracy: {'acc': 0.7971428571428572}
Export on local
python export.py --checkpoint_file_path [CKPT_PATH] --file_name [FILE_NAME] --file_format [FILE_FORMAT] --config_path [CONFIG_FILE]
The checkpoint_file_path parameter is required,
EXPORT_FORMAT
should be in ["AIR", "MINDIR"]
Export on ModelArts (If you want to run in modelarts, please check the official documentation of modelarts, and you can start as follows)
# Export on ModelArts
# (1) Perform a or b.
# a. Set "enable_modelarts=True" on default_config.yaml file.
# Set "data_path='/cache/data/' " on default_config.yaml file.
# Set "checkpoint_file_path='/cache/checkpoint_path/model.ckpt'" on default_config.yaml file.
# Set "checkpoint_url='s3://dir_to_trained_ckpt/'" on default_config.yaml file.
# Set "file_name='./textcnn'" on default_config.yaml file.
# Set "file_format='MINDIR'" on default_config.yaml file.
# Set other parameters on default_config.yaml file you need.
# b. Add "enable_modelarts=True" on the website UI interface.
# Add "data_path='/cache/data/' " on default_config.yaml file.
# Add "checkpoint_file_path='/cache/checkpoint_path/model.ckpt'" on the website UI interface.
# Add "checkpoint_url='s3://dir_to_trained_ckpt/'" on the website UI interface.
# Add "file_name='./textcnn'" on the website UI interface.
# Add "file_format='MINDIR'" on the website UI interface.
# Add other parameters on the website UI interface.
# (2) Set the config_path="/path/yaml file" on the website UI interface.
# (3) Set the code directory to "/path/textcnn" on the website UI interface.
# (4) Set the startup file to "export.py" on the website UI interface.
# (5) Set the "Output file path" and "Job log path" to your path on the website UI interface.
# (6) Create your job.
python export.py --checkpoint_file_path [CKPT_PATH] --file_name [FILE_NAME] --file_format "ONNX" --config_path [CONFIG_FILE] --data_path [DATA_PATH] --device_target "GPU"
# example:python export.py --checkpoint_file_path train_textcnn-4_1052.ckpt --file_name 'sst2_textcnn' --file_format "ONNX" --config_path sst2_config.yaml --data_path data/SST-2/ --device_target "GPU"
Running scripts for onnx evaluation of GRU. The command as below.
# `CONFIG_PATH` `ONNX_FILE` `DATA_PATH` `DATASET` parameters need to be passed externally or modified yaml file
# `DATASET` must choose from ['MR', 'SUBJ', 'SST2']"
python eval_onnx.py --config_path [CONFIG_PATH] \
--device_target GPU \
--onnx_file [ONNX_FILE] \
--data_path [DATA_PATH] > eval_onnx.log 2>&1 &
OR
bash scripts/run_eval_onnx_gpu.sh [ONNX_FILE] [DATASET] [DATA_PATH]
The above python command will run in the background. You can view the results through the file "eval_onnx.log". The accuracy of the test dataset will be as follows:
acc: 0.8329326923076923
Before inference, please refer to MindSpore Inference with C++ Deployment Guide to set environment variables.
Before performing inference, the mindir file must be exported by export.py. Input files must be in bin format.
# Ascend310 inference
bash run_infer_310.sh [MINDIR_PATH] [DATASET_NAME] [NEED_PREPROCESS] [DEVICE_ID]
DATASET_NAME
must choose from ['MR', 'SUBJ', 'SST2']
NEED_PREPROCESS
means weather need preprocess or not, it's value is 'y' or 'n'."
DEVICE_ID
is optional, it can be set by environment variable DEVICE_ID, default value is 0.
Inference result is saved in current path, you can find result in acc.log file.
# grep "accuracy: " acc.log
accuracy: 0.7971428571428572
Parameters | Ascend | GPU |
---|---|---|
Model Version | TextCNN | TextCNN |
Resource | Ascend 910; CPU 2.60GHz, 192cores; Memory 755G; OS Euler2.8 | Nvidia V100 SXM2; CPU 1.526GHz; 72cores; Memory 42G; OS Ubuntu16 |
uploaded Date | 11/10/2020 (month/day/year) | 12/16/2021 (month/day/year) |
MindSpore Version | 1.0.1 | 1.5.0 |
Dataset | Movie Review Data | Movie Review Data |
Training Parameters | epoch=4, steps=149, batch_size=64 | epoch=4, steps=149, batch_size=64 |
Optimizer | Adam | Adam |
Loss Function | Softmax Cross Entropy | Softmax Cross Entropy |
outputs | probability | probability |
Loss | 0.1724 | 0.1072 |
Speed | 1pc: 12.069 ms/step | 1pc: 4.226 ms/step |
Total time | 1pc: 28.074s | 1pc: 13.274s |
Scripts | textcnn script |
Parameters | Ascend | GPU |
---|---|---|
Resource | Ascend 910; CPU 2.60GHz, 192cores; Memory 755G; OS Euler2.8 | Nvidia V100 SXM2; CPU 1.526GHz; 72cores; Memory 42G; OS Ubuntu16 |
Uploaded Date | 12/16/2021 (month/day/year) | 12/16/2021 (month/day/year) |
MindSpore Version | 1.5.0 | 1.5.0 |
Dataset | Movie Review Data | Movie Review Data |
batch_size | 64 | 64 |
outputs | accuracy | accuracy |
Accuracy | 0.7705 | 0.7764 |
Model for inference | 9.9M (.ckpt file) | 9.9M (.ckpt file) |
Parameters | Ascend | GPU |
---|---|---|
Model Version | TextCNN | TextCNN |
Resource | Ascend 910; CPU 2.60GHz, 192cores; Memory 755G; OS Euler2.8 | Nvidia V100 SXM2; CPU 1.526GHz; 72cores; Memory 42G; OS Ubuntu16 |
uploaded Date | 12/16/2020 (month/day/year) | 12/16/2021 (month/day/year) |
MindSpore Version | 1.5.0 | 1.5.0 |
Dataset | SST2 | SST2 |
Training Parameters | epoch=4, steps=149, batch_size=64 | epoch=4, steps=149, batch_size=64 |
Optimizer | Adam | Adam |
Loss Function | Softmax Cross Entropy | Softmax Cross Entropy |
outputs | probability | probability |
Loss | 0.1665 | 0.1582 |
Speed | 1pc: 6.398 ms/step | 1pc: 3.945 ms/step |
Total time | 1pc: 58.455s | 1pc: 29.257s |
Scripts | textcnn script |
Parameters | Ascend | GPU |
---|---|---|
Resource | Ascend 910; CPU 2.60GHz, 192cores; Memory 755G; OS Euler2.8 | Nvidia V100 SXM2; CPU 1.526GHz; 72cores; Memory 42G; OS Ubuntu16 |
Uploaded Date | 12/16/2021 (month/day/year) | 12/16/2021 (month/day/year) |
MindSpore Version | 1.5.0 | 1.5.0 |
Dataset | SST2 | SST2 |
batch_size | 64 | 64 |
outputs | accuracy | accuracy |
Accuracy | 0.8437 | 0.8353 |
Model for inference | 7.7M (.ckpt file) | 7.7M (.ckpt file) |
Parameters | Ascend | GPU |
---|---|---|
Model Version | TextCNN | TextCNN |
Resource | Ascend 910; CPU 2.60GHz, 192cores; Memory 755G; OS Euler2.8 | Nvidia V100 SXM2; CPU 1.526GHz; 72cores; Memory 42G; OS Ubuntu16 |
uploaded Date | 12/16/2020 (month/day/year) | 12/16/2021 (month/day/year) |
MindSpore Version | 1.5.0 | 1.5.0 |
Dataset | SUBJ | SUBJ |
Training Parameters | epoch=4, steps=149, batch_size=64 | epoch=4, steps=149, batch_size=64 |
Optimizer | Adam | Adam |
Loss Function | Softmax Cross Entropy | Softmax Cross Entropy |
outputs | probability | probability |
Loss | 0.0350 | 0.0930 |
Speed | 1pc: 6.565ms/step | 1pc: 4.345 ms/step |
Total time | 1pc: 33.817s | 1pc: 13.507s |
Scripts | textcnn script |
Parameters | Ascend | GPU |
---|---|---|
Resource | Ascend 910; CPU 2.60GHz, 192cores; Memory 755G; OS Euler2.8 | Nvidia V100 SXM2; CPU 1.526GHz; 72cores; Memory 42G; OS Ubuntu16 |
Uploaded Date | 12/16/2021 (month/day/year) | 12/16/2021 (month/day/year) |
MindSpore Version | 1.5.0 | 1.5.0 |
Dataset | SUBJ | SUBJ |
batch_size | 64 | 64 |
outputs | accuracy | accuracy |
Accuracy | 0.9052 | 0.8927 |
Model for inference | 11M (.ckpt file) | 11M (.ckpt file) |
Please check the official homepage.
此处可能存在不合适展示的内容,页面不予展示。您可通过相关编辑功能自查并修改。
如您确认内容无涉及 不当用语 / 纯广告导流 / 暴力 / 低俗色情 / 侵权 / 盗版 / 虚假 / 无价值内容或违法国家有关法律法规的内容,可点击提交进行申诉,我们将尽快为您处理。