Skip to content

Commit

Permalink
1.0.0rc7
Browse files Browse the repository at this point in the history
  • Loading branch information
Yang Heng committed Aug 11, 2021
1 parent b98c463 commit 269c40c
Show file tree
Hide file tree
Showing 201 changed files with 3,129 additions and 3,881 deletions.
3 changes: 2 additions & 1 deletion .gitignore
View file
Original file line number Diff line number Diff line change
Expand Up @@ -6,6 +6,7 @@ datasets/
*.tokenizer
*.model
*.config
*.args
*.zip
*.gz
*.bin
Expand All @@ -22,7 +23,7 @@ glove.840B.300d.txt
glove.42B.300d.txt
glove.twitter.27B.txt

# project config files
# project main files
checkpoints.json
release_note.json

Expand Down
215 changes: 50 additions & 165 deletions README.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -68,50 +68,50 @@ python sentiment_inference_chinese.py
# Model Support

Except for the following models, we provide a template model involving LCF vec, you can develop your model based on
the [LCF-APC](pyabsa/tasks/apc/models/lcf_template_apc.py) model template
or [LCF-ATEPC](pyabsa/tasks/atepc/models/lcf_template_atepc.py) model template.
the [LCF-APC](pyabsa/core/apc/models/lcf_template_apc.py) model template
or [LCF-ATEPC](pyabsa/core/atepc/models/lcf_template_atepc.py) model template.

## ATEPC

1. [LCF-ATEPC](pyabsa/tasks/atepc/models/lcf_atepc.py)
2. [LCF-ATEPC-LARGE](pyabsa/tasks/atepc/models/lcf_atepc_large.py) (Dual BERT)
2. [FAST-LCF-ATEPC](pyabsa/tasks/atepc/models/fast_lcf_atepc.py)
3. [LCFS-ATEPC](pyabsa/tasks/atepc/models/lcfs_atepc.py)
4. [LCFS-ATEPC-LARGE](pyabsa/tasks/atepc/models/lcfs_atepc_large.py) (Dual BERT)
5. [FAST-LCFS-ATEPC](pyabsa/tasks/atepc/models/fast_lcfs_atepc.py)
6. [BERT-BASE](pyabsa/tasks/atepc/models/bert_base_atepc.py)
1. [LCF-ATEPC](pyabsa/core/atepc/models/lcf_atepc.py)
2. [LCF-ATEPC-LARGE](pyabsa/core/atepc/models/lcf_atepc_large.py) (Dual BERT)
2. [FAST-LCF-ATEPC](pyabsa/core/atepc/models/fast_lcf_atepc.py)
3. [LCFS-ATEPC](pyabsa/core/atepc/models/lcfs_atepc.py)
4. [LCFS-ATEPC-LARGE](pyabsa/core/atepc/models/lcfs_atepc_large.py) (Dual BERT)
5. [FAST-LCFS-ATEPC](pyabsa/core/atepc/models/fast_lcfs_atepc.py)
6. [BERT-BASE](pyabsa/core/atepc/models/bert_base_atepc.py)

## APC

1. [SLIDE-LCF-BERT *](pyabsa/tasks/apc/models/slide_lcf_bert.py) (Faster & Performs Better than LCF/LCFS-BERT)
2. [SLIDE-LCFS-BERT *](pyabsa/tasks/apc/models/slide_lcfs_bert.py) (Faster & Performs Better than LCF/LCFS-BERT)
3. [LCF-BERT](pyabsa/tasks/apc/models/lcf_bert.py) (Reimplemented & Enhanced)
4. [LCFS-BERT](pyabsa/tasks/apc/models/lcfs_bert.py) (Reimplemented & Enhanced)
5. [FAST-LCF-BERT](pyabsa/tasks/apc/models/fast_lcf_bert.py) (Faster with slightly performance loss)
6. [FAST_LCFS-BERT](pyabsa/tasks/apc/models/fast_lcfs_bert.py) (Faster with slightly performance loss)
7. [LCF-DUAL-BERT](pyabsa/tasks/apc/models/lcf_dual_bert.py) (Dual BERT)
8. [LCFS-DUAL-BERT](pyabsa/tasks/apc/models/lcfs_dual_bert.py) (Dual BERT)
9. [BERT-BASE](pyabsa/tasks/apc/models/bert_base.py)
10. [BERT-SPC](pyabsa/tasks/apc/models/bert_spc.py)
11. [LCA-Net](pyabsa/tasks/apc/models/lca_bert.py)
12. [DLCF-DCA-BERT](pyabsa/tasks/apc/models/dlcf_dca_bert.py)
1. [SLIDE-LCF-BERT *](pyabsa/core/apc/models/slide_lcf_bert.py) (Faster & Performs Better than LCF/LCFS-BERT)
2. [SLIDE-LCFS-BERT *](pyabsa/core/apc/models/slide_lcfs_bert.py) (Faster & Performs Better than LCF/LCFS-BERT)
3. [LCF-BERT](pyabsa/core/apc/models/lcf_bert.py) (Reimplemented & Enhanced)
4. [LCFS-BERT](pyabsa/core/apc/models/lcfs_bert.py) (Reimplemented & Enhanced)
5. [FAST-LCF-BERT](pyabsa/core/apc/models/fast_lcf_bert.py) (Faster with slightly performance loss)
6. [FAST_LCFS-BERT](pyabsa/core/apc/models/fast_lcfs_bert.py) (Faster with slightly performance loss)
7. [LCF-DUAL-BERT](pyabsa/core/apc/models/lcf_dual_bert.py) (Dual BERT)
8. [LCFS-DUAL-BERT](pyabsa/core/apc/models/lcfs_dual_bert.py) (Dual BERT)
9. [BERT-BASE](pyabsa/core/apc/models/bert_base.py)
10. [BERT-SPC](pyabsa/core/apc/models/bert_spc.py)
11. [LCA-Net](pyabsa/core/apc/models/lca_bert.py)
12. [DLCF-DCA-BERT](pyabsa/core/apc/models/dlcf_dca_bert.py)

'*' Copyrights Reserved, please wait for the publishing of our paper to get the introduction of them in detail.

### GloVe-based APC models

13. [AOA](pyabsa/tasks/apc/__glove__/models/aoa.py) (AOA_BERT)
14. [ASGCN](pyabsa/tasks/apc/__glove__/models/asgcn.py) (ASGCN_BERT)
15. [ATAE-LSTM](pyabsa/tasks/apc/__glove__/models/atae_lstm.py) (ATAE_LSTM_BERT)
16. [Cabasc](pyabsa/tasks/apc/__glove__/models/cabasc.py) (Cabasc_BERT)
17. [IAN](pyabsa/tasks/apc/__glove__/models/ian.py) (IAN_BERT)
18. [LSTM](pyabsa/tasks/apc/__glove__/models/lstm.py) (LSTM_BERT)
19. [MemNet](pyabsa/tasks/apc/__glove__/models/memnet.py) (MemNet_BERT)
20. [MGAN](pyabsa/tasks/apc/__glove__/models/mgan.py) (MGAN_BERT)
21. [RAM](pyabsa/tasks/apc/__glove__/models/ram.py) (RAM_BERT)
22. [TD-LSTM](pyabsa/tasks/apc/__glove__/models/td_lstm.py) (TD_LSTM_BERT)
23. [TD-LSTM](pyabsa/tasks/apc/__glove__/models/tc_lstm.py) (TC_LSTM_BERT)
24. [TNet_LF](pyabsa/tasks/apc/__glove__/models/tnet_lf.py) (TNet_LF_BERT)
13. [AOA](pyabsa/core/apc/__glove__/models/aoa.py) (AOA_BERT)
14. [ASGCN](pyabsa/core/apc/__glove__/models/asgcn.py) (ASGCN_BERT)
15. [ATAE-LSTM](pyabsa/core/apc/__glove__/models/atae_lstm.py) (ATAE_LSTM_BERT)
16. [Cabasc](pyabsa/core/apc/__glove__/models/cabasc.py) (Cabasc_BERT)
17. [IAN](pyabsa/core/apc/__glove__/models/ian.py) (IAN_BERT)
18. [LSTM](pyabsa/core/apc/__glove__/models/lstm.py) (LSTM_BERT)
19. [MemNet](pyabsa/core/apc/__glove__/models/memnet.py) (MemNet_BERT)
20. [MGAN](pyabsa/core/apc/__glove__/models/mgan.py) (MGAN_BERT)
21. [RAM](pyabsa/core/apc/__glove__/models/ram.py) (RAM_BERT)
22. [TD-LSTM](pyabsa/core/apc/__glove__/models/td_lstm.py) (TD_LSTM_BERT)
23. [TD-LSTM](pyabsa/core/apc/__glove__/models/tc_lstm.py) (TC_LSTM_BERT)
24. [TNet_LF](pyabsa/core/apc/__glove__/models/tnet_lf.py) (TNet_LF_BERT)

### You can use $glove model name$_BERT to use the BERT version

Expand Down Expand Up @@ -189,49 +189,49 @@ Check the detailed usages in [ATE examples](examples/aspect_term_extraction) dir
### 1. Import necessary entries

```
from pyabsa import train_atepc, atepc_config_handler
from pyabsa import ABSADatasetList
from pyabsa import ATEPCModelList
from pyabsa.functional import ATEPCModelList
from pyabsa.functional import Trainer, ATEPCTrainer
from pyabsa.functional import ABSADatasetList
from pyabsa.functional import ATEPCConfigManager
```

### 2. Choose a base param_dict

```
param_dict = atepc_config_handler.get_apc_param_dict_chinese()
config = ATEPCConfigManager.get_atepc_config_english()
```

### 3. Specify an ATEPC model and alter some hyper-parameters (if necessary)

```
atepc_param_dict_chinese['model'] = ATEPCModelList.LCF_ATEPC
atepc_param_dict_chinese['log_step'] = 20
atepc_param_dict_chinese['evaluate_begin'] = 5
atepc_config_english = ATEPCConfigManager.get_atepc_config_english()
atepc_config_english.num_epoch = 10
atepc_config_english.evaluate_begin = 4
atepc_config_english.lot_step = 100
atepc_config_english.model = ATEPCModelList.LCF_ATEPC
```

### 4. Configure runtime setting and running training

```
save_path = 'state_dict'
chinese_sets = ABSADatasetList.Chinese
sent_classifier = train_apc(parameter_dict=param_dict, # set param_dict=None to use default model
dataset_path=chinese_sets, # train set and test set will be automatically detected
model_path_to_save=save_path, # set model_path_to_save=None to avoid save model
auto_evaluate=True, # evaluate model while training_tutorials if test set is available
auto_device=True # automatic choose CUDA or CPU
)
laptop14 = ABSADatasetList.Laptop14
aspect_extractor = ATEPCTrainer(config=atepc_config_english, # set config=None to use default model
dataset=laptop14, # file or dir, dataset_utils(s) will be automatically detected
save_checkpoint=True, # set model_path_to_save=None to avoid save model
auto_device=True # Auto choose CUDA or CPU
)
```

### 5. Aspect term extraction & sentiment inference

```
from pyabsa import load_aspect_extractor
from pyabsa import ATEPCCheckpointManager
examples = ['相比较原系列锐度高了不少这一点好与不好大家有争议',
'这款手机的大小真的很薄，但是颜色不太好看， 总体上我很满意啦。'
]
model_path = ATEPCCheckpointManager.get_checkpoint(checkpoint_name='Chinese')
aspect_extractor = ATEPCCheckpointManager.get_aspect_extractor(checkpoint_name='Chinese')
sentiment_map = {0: 'Bad', 1: 'Good', -999: ''}
aspect_extractor = load_aspect_extractor(trained_model_path=model_path,
Expand All @@ -243,121 +243,6 @@ atepc_result = aspect_extractor.extract_aspect(examples=examples, # list-supp
print_result=True, # print the result
pred_sentiment=True, # Predict the sentiment of extracted aspect terms
)
```

# Aspect Polarity Classification (APC)

Check the detailed usages in [APC examples](examples/aspect_polarity_classification) directory.

## Aspect-Polarity Classification Output Format (方面极性分类输出示例如下):

```
love selena gomez !!!! she rock !!!!!!!!!!!!!!!! and she 's cool she 's my idol
selena gomez --> Positive Real: Positive (Correct)
thehils Heard great things about the ipad for speech/communication . Educational discounts are problem best bet . Maybe Thanksgiving ?
ipad --> Neutral Real: Neutral (Correct)
Jamie fox , Eddie Murphy , and barack obama because they all are exciting , cute , and inspirational to lots of people including me !!!
barack obama --> Positive Real: Neutral (Wrong)
```

## Quick Start

### 1. Import necessary entries

```
from pyabsa import train_apc, apc_config_handler
from pyabsa import APCModelList
from pyabsa import ABSADatasetList
```

### 2. Choose a base param_dict

```
param_dict = apc_config_handler.get_apc_param_dict_english()
```

### 3. Specify an APC model and alter some hyper-parameters (if necessary)

```
apc_param_dict_english['model'] = APCModelList.SLIDE_LCF_BERT
apc_param_dict_english['evaluate_begin'] = 2 # to reduce evaluation times and save resources
apc_param_dict_english['similarity_threshold'] = 1
apc_param_dict_english['max_seq_len'] = 80
apc_param_dict_english['dropout'] = 0.5
apc_param_dict_english['log_step'] = 5
apc_param_dict_english['l2reg'] = 0.0001
apc_param_dict_english['dynamic_truncate'] = True
apc_param_dict_english['srd_alignment'] = True
```

check [parameter introduction](examples/common_usages/param_dict_introduction.py) and learn how to set them

### 4. Configure runtime setting and running training

```
laptop14 = ABSADatasetList.Laptop14 # Here I use the integrated dataset, you can use your dataset instead
sent_classifier = train_apc(parameter_dict=apc_param_dict_english, # ignore this parameter will use defualt setting
dataset_path=laptop14, # datasets will be recurrsively detected in this path
model_path_to_save=save_path, # ignore this parameter to avoid saving model
auto_evaluate=True, # evaluate model if testset is available
auto_device=True # automatic choose CUDA if any, False means always use CPU
)
```

### 5. Sentiment inference

```
from pyabsa import load_sentiment_classifier
from pyabsa import ABSADatasetList
from pyabsa.models import APCCheckpointManager
# 如果有需要，使用以下方法自定义情感索引到情感标签的词典， 其中-999为必需的填充， e.g.,
sentiment_map = {0: 'Negative', 1: 'Neutral', 2: 'Positive', -999: ''}
# Here I provided some pre-trained models in case of having no resource to train a model,
# you can train a model and specify the model path to infer instead
model_path = APCCheckpointManager.get_checkpoint(checkpoint_name='English')
sent_classifier = load_sentiment_classifier(trained_model_path=model_path,
auto_device=True, # Use CUDA if available
sentiment_map=sentiment_map # define polarity2name map
)
text = 'everything is always cooked to perfection , the [ASP]service[ASP] is excellent , the [ASP]decor[ASP] cool and understated . !sent! 1 1'
# Note reference sentiment like '!sent! 1 1' are not mandatory
sent_classifier.infer(text, print_result=True)
# batch inferring_tutorials returns the results, save the result if necessary using save_result=True
inference_sets = ABSADatasetList.SemEval
results = sent_classifier.batch_infer(target_file=inference_sets,
print_result=True,
save_result=True,
ignore_error=True, # some data are broken so ignore them
)
```

### Searching optimal hyper-parameter in the alternative parameter set.

You use this function to search the optimal setting of some params, e.g., learning_rate.

```
from pyabsa.research.parameter_search.search_param_for_apc import apc_param_search
from pyabsa import ABSADatasetList
from pyabsa.config.apc_config import apc_config_handler
apc_param_dict_english = apc_config_handler.get_apc_param_dict_english()
apc_param_dict_english['log_step'] = 10
apc_param_dict_english['evaluate_begin'] = 2
param_to_search = ['l2reg', [1e-5, 5e-5, 1e-4, 5e-4, 1e-3]]
apc_param_search(parameter_dict=apc_param_dict_english,
dataset_path=ABSADatasetList.Laptop14,
search_param=param_to_search,
auto_evaluate=True,
auto_device=True)
```

# [Datasets](https://github.com/yangheng95/ABSADatasetList)
Expand Down
66 changes: 33 additions & 33 deletions README_CN.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -54,50 +54,50 @@ python sentiment_inference_chinese.py

# 模型支持

除了下面的模型之外，我们提供了一个包含LCF vec的模板模型， 您可以基于 [LCF-APC](pyabsa/tasks/apc/models/lcf_template_apc.py) 模型模板
或 [LCF-ATEPC](pyabsa/tasks/atepc/models/lcf_template_atepc.py) 模型模版来开始你的创新.
除了下面的模型之外，我们提供了一个包含LCF vec的模板模型， 您可以基于 [LCF-APC](pyabsa/core/apc/models/lcf_template_apc.py) 模型模板
或 [LCF-ATEPC](pyabsa/core/atepc/models/lcf_template_atepc.py) 模型模版来开始你的创新.

## ATEPC

1. [LCF-ATEPC](pyabsa/tasks/atepc/models/lcf_atepc.py)
2. [LCF-ATEPC-LARGE](pyabsa/tasks/atepc/models/lcf_atepc_large.py)
2. [FAST-LCF-ATEPC](pyabsa/tasks/atepc/models/fast_lcf_atepc.py)
3. [LCFS-ATEPC](pyabsa/tasks/atepc/models/lcfs_atepc.py)
4. [LCFS-ATEPC-LARGE](pyabsa/tasks/atepc/models/lcfs_atepc_large.py)
5. [FAST-LCFS-ATEPC](pyabsa/tasks/atepc/models/fast_lcfs_atepc.py)
6. [BERT-BASE](pyabsa/tasks/atepc/models/bert_base_atepc.py)
1. [LCF-ATEPC](pyabsa/core/atepc/models/lcf_atepc.py)
2. [LCF-ATEPC-LARGE](pyabsa/core/atepc/models/lcf_atepc_large.py)
2. [FAST-LCF-ATEPC](pyabsa/core/atepc/models/fast_lcf_atepc.py)
3. [LCFS-ATEPC](pyabsa/core/atepc/models/lcfs_atepc.py)
4. [LCFS-ATEPC-LARGE](pyabsa/core/atepc/models/lcfs_atepc_large.py)
5. [FAST-LCFS-ATEPC](pyabsa/core/atepc/models/fast_lcfs_atepc.py)
6. [BERT-BASE](pyabsa/core/atepc/models/bert_base_atepc.py)

## APC

1. [SLIDE-LCF-BERT *](pyabsa/tasks/apc/models/slide_lcf_bert.py) (Faster & Performs Better than LCF/LCFS-BERT)
2. [SLIDE-LCFS-BERT *](pyabsa/tasks/apc/models/slide_lcfs_bert.py) (Faster & Performs Better than LCF/LCFS-BERT)
3. [LCF-BERT](pyabsa/tasks/apc/models/lcf_bert.py) (Reimplemented & Enhanced)
4. [LCFS-BERT](pyabsa/tasks/apc/models/lcfs_bert.py) (Reimplemented & Enhanced)
5. [FAST-LCF-BERT](pyabsa/tasks/apc/models/fast_lcf_bert.py) (Faster with slightly performance loss)
6. [FAST_LCFS-BERT](pyabsa/tasks/apc/models/fast_lcfs_bert.py) (Faster with slightly performance loss)
7. [LCF-DUAL-BERT](pyabsa/tasks/apc/models/lcf_dual_bert.py) (Dual BERT)
8. [LCFS-DUAL-BERT](pyabsa/tasks/apc/models/lcfs_dual_bert.py) (Dual BERT)
9. [BERT-BASE](pyabsa/tasks/apc/models/bert_base.py)
10. [BERT-SPC](pyabsa/tasks/apc/models/bert_spc.py)
11. [LCA-Net](pyabsa/tasks/apc/models/lca_bert.py)
12. [DLCF-DCA-BERT](pyabsa/tasks/apc/models/dlcf_dca_bert.py)
1. [SLIDE-LCF-BERT *](pyabsa/core/apc/models/slide_lcf_bert.py) (Faster & Performs Better than LCF/LCFS-BERT)
2. [SLIDE-LCFS-BERT *](pyabsa/core/apc/models/slide_lcfs_bert.py) (Faster & Performs Better than LCF/LCFS-BERT)
3. [LCF-BERT](pyabsa/core/apc/models/lcf_bert.py) (Reimplemented & Enhanced)
4. [LCFS-BERT](pyabsa/core/apc/models/lcfs_bert.py) (Reimplemented & Enhanced)
5. [FAST-LCF-BERT](pyabsa/core/apc/models/fast_lcf_bert.py) (Faster with slightly performance loss)
6. [FAST_LCFS-BERT](pyabsa/core/apc/models/fast_lcfs_bert.py) (Faster with slightly performance loss)
7. [LCF-DUAL-BERT](pyabsa/core/apc/models/lcf_dual_bert.py) (Dual BERT)
8. [LCFS-DUAL-BERT](pyabsa/core/apc/models/lcfs_dual_bert.py) (Dual BERT)
9. [BERT-BASE](pyabsa/core/apc/models/bert_base.py)
10. [BERT-SPC](pyabsa/core/apc/models/bert_spc.py)
11. [LCA-Net](pyabsa/core/apc/models/lca_bert.py)
12. [DLCF-DCA-BERT](pyabsa/core/apc/models/dlcf_dca_bert.py)

'*' * 版权所有，请等待我们的论文发表以获取详细的介绍。

### GloVe-based APC models

13. [AOA](pyabsa/tasks/apc/__glove__/models/aoa.py)
14. [ASGCN](pyabsa/tasks/apc/__glove__/models/asgcn.py)
15. [ATAE-LSTM](pyabsa/tasks/apc/__glove__/models/atae_lstm.py)
16. [Cabasc](pyabsa/tasks/apc/__glove__/models/cabasc.py)
17. [IAN](pyabsa/tasks/apc/__glove__/models/ian.py)
18. [LSTM](pyabsa/tasks/apc/__glove__/models/lstm.py)
19. [MemNet](pyabsa/tasks/apc/__glove__/models/memnet.py)
20. [MGAN](pyabsa/tasks/apc/__glove__/models/mgan.py)
21. [RAM](pyabsa/tasks/apc/__glove__/models/ram.py)
22. [TD-LSTM](pyabsa/tasks/apc/__glove__/models/td_lstm.py)
23. [TD-LSTM](pyabsa/tasks/apc/__glove__/models/tc_lstm.py)
24. [TNet_LF](pyabsa/tasks/apc/__glove__/models/tnet_lf.py)
13. [AOA](pyabsa/core/apc/__glove__/models/aoa.py)
14. [ASGCN](pyabsa/core/apc/__glove__/models/asgcn.py)
15. [ATAE-LSTM](pyabsa/core/apc/__glove__/models/atae_lstm.py)
16. [Cabasc](pyabsa/core/apc/__glove__/models/cabasc.py)
17. [IAN](pyabsa/core/apc/__glove__/models/ian.py)
18. [LSTM](pyabsa/core/apc/__glove__/models/lstm.py)
19. [MemNet](pyabsa/core/apc/__glove__/models/memnet.py)
20. [MGAN](pyabsa/core/apc/__glove__/models/mgan.py)
21. [RAM](pyabsa/core/apc/__glove__/models/ram.py)
22. [TD-LSTM](pyabsa/core/apc/__glove__/models/td_lstm.py)
23. [TD-LSTM](pyabsa/core/apc/__glove__/models/tc_lstm.py)
24. [TNet_LF](pyabsa/core/apc/__glove__/models/tnet_lf.py)

## 模型性能表现

Expand Down
3 changes: 2 additions & 1 deletion Release Note.json
View file
Original file line number Diff line number Diff line change
@@ -1,4 +1,5 @@
{ "0.9.2.1": {
{
"0.9.2.1": {
"1": "fix path of #49"
},
"0.9.2.0": {
Expand Down
Loading

0 comments on commit 269c40c

Please sign in to comment.