Merge branch 'v2' into dev

# Conflicts:
#	README.MD
#	examples-v2/aspect_polarity_classification/ensemble_inference.py
#	examples-v2/aspect_polarity_classification/train_apc.py
#	examples-v2/aspect_term_extraction/train_atepc.py
#	examples-v2/text_classification/train_text_classification.py
#	pyabsa/__init__.py
#	pyabsa/tasks/RNARegression/dataset_utils/__plm__/data_utils_for_inference.py
#	pyabsa/tasks/RNARegression/dataset_utils/__plm__/data_utils_for_training.py
#	pyabsa/tasks/_Archive/RNAClassification/dataset_utils/data_utils_for_training.py

README.MD

@@ -1,45 +1,46 @@
-# [PyABSA - Open Framework for Aspect-based Sentiment Analysis](https://arxiv.org/pdf/2208.01368)
+# PyABSA - Open Framework for Aspect-based Sentiment Analysis ([paper](https://dl.acm.org/doi/abs/10.1145/3583780.3614752))
 
-![PyPI - Python Version](https://img.shields.io/badge/python-3.8-gold.svg)
+![PyPI - Python Version](https://img.shields.io/badge/python-3.6-blue.svg)
 [![PyPI](https://img.shields.io/pypi/v/pyabsa)](https://pypi.org/project/pyabsa/)
 [![Downloads](https://pepy.tech/badge/pyabsa)](https://pepy.tech/project/pyabsa)
 [![Downloads](https://pepy.tech/badge/pyabsa/month)](https://pepy.tech/project/pyabsa)
 ![License](https://img.shields.io/pypi/l/pyabsa?logo=PyABSA)
-[![Documentation Status](https://readthedocs.org/projects/pyabsa/badge/?version=v2)](https://pyabsa.readthedocs.io/en/latest/)
+[![Documentation Status](https://readthedocs.org/projects/pyabsa/badge/?version=v2)](https://pyabsa.readthedocs.io/en/v2/?badge=v2)
 
+[![total views](https://raw.githubusercontent.com/yangheng95/PyABSA/traffic/total_views.svg)](https://github.com/yangheng95/PyABSA/tree/traffic#-total-traffic-data-badge)
+[![total views per week](https://raw.githubusercontent.com/yangheng95/PyABSA/traffic/total_views_per_week.svg)](https://github.com/yangheng95/PyABSA/tree/traffic#-total-traffic-data-badge)
+[![total clones](https://raw.githubusercontent.com/yangheng95/PyABSA/traffic/total_clones.svg)](https://github.com/yangheng95/PyABSA/tree/traffic#-total-traffic-data-badge)
+[![total clones per week](https://raw.githubusercontent.com/yangheng95/PyABSA/traffic/total_clones_per_week.svg)](https://github.com/yangheng95/PyABSA/tree/traffic#-total-traffic-data-badge)
+
+[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/back-to-reality-leveraging-pattern-driven/aspect-based-sentiment-analysis-on-semeval)](https://paperswithcode.com/sota/aspect-based-sentiment-analysis-on-semeval?p=back-to-reality-leveraging-pattern-driven)
+
+**Hi, there!** Please star this repo if it helps you! Each Star helps PyABSA go further, many thanks. PyABSA is a free
+and open-source tool for everyone, but please do not forget to attach the (informal or formal) author
+information and project address in your works, products and publications, etc.
 
 ## Try our demos on Huggingface Space
 
-Apart from the [paper](https://arxiv.org/pdf/2208.01368), there are two new features in PyABSA: Aspect sentiment triplet extraction and Aspect quadruple extraction.
+Apart from the [paper](https://arxiv.org/pdf/2208.01368), there are two new features in PyABSA: Aspect sentiment triplet
+extraction and Aspect quadruple extraction.
 We have deployed the demos on Huggingface Space, you can try them online.
+
 - **[Aspect sentiment quadruple extraction](https://huggingface.co/spaces/yangheng/PyABSA) (English) New feature**
 - **[Aspect sentiment triplet extraction](https://huggingface.co/spaces/yangheng/PyABSA) (English) New feature**
 - [(Gradio) Aspect term extraction & sentiment classification](https://huggingface.co/spaces/Gradio-Blocks/Multilingual-Aspect-Based-Sentiment-Analysis) (
   English, Chinese, Arabic, Dutch, French, Russian, Spanish, Turkish, etc.)
 - [(Prototype) Aspect term extraction & sentiment classification](https://huggingface.co/spaces/yangheng/PyABSA-ATEPC) (
   English,
   Chinese, Arabic, Dutch, French, Russian, Spanish, Turkish, etc.)
+- [方面术语提取和情感分类](https://huggingface.co/spaces/yangheng/PyABSA-ATEPC-Chinese) （中文, etc.）
 - [Aspect-based sentiment classification (Multilingual)](https://huggingface.co/spaces/yangheng/PyABSA-APC) （English,
   Chinese, etc.）
 
 ## Usage Examples
+
 We have prepared many examples for different tasks. Please refer to [Examples](./examples-v2) for more usage examples.
 
 ## Installation
 
-### Install Anaconda (Optional)
-Install [Anaconda](https://www.anaconda.com/products/individual) if you do not have it.
-
-### Create a new environment (Optional)
-```bash
-conda create -n pyabsa python=3.8
-conda activate pyabsa
-```
-
-### Install Pytorch
-Install Pytorch following the [official instructions](https://pytorch.org/get-started/locally/). 
-You need to select your environment, operating system, package, Python version, and CUDA version.
-
 ### install via pip
 
 To use PyABSA, install the latest version from pip or source code:
@@ -89,6 +90,7 @@ print(atepc_result)
 
 
 ```
+
 ### Aspect-based sentiment analysis
 
 ```python3
@@ -119,7 +121,9 @@ apc_result = classifier.batch_predict(target_file=inference_source,  #
 print(apc_result)
 
 ```
+
 ## Dataset Annotation and Model Training
+
 please refer to the documentation: [PyABSA Documentation](https://pyabsa.readthedocs.io/en/v2/).
 If you have any questions about the docs, please feel free to raise an issue. Also, you can join to improve the docs.
 
@@ -157,11 +161,12 @@ If you are looking for the original proposal of local context focus, here are so
 [here](https://github.com/yangheng95/PyABSA/tree/release/demos/documents).
 
 ## Citation
+
 ```bibtex
 @article{YangL22,
   author    = {Heng Yang and
                Ke Li},
-  title     = {A Modularized Framework for Reproducible Aspect-based Sentiment Analysis},
+  title     = {PyABSA: Open Framework for Aspect-based Sentiment Analysis},
   journal   = {CoRR},
   volume    = {abs/2208.01368},
   year      = {2022},
@@ -174,6 +179,7 @@ If you are looking for the original proposal of local context focus, here are so
   bibsource = {dblp computer science bibliography, https://dblp.org}
 }
 ```
+
 ## Contribution
 
 This repository is developed and maintained by HENG YANG ([yangheng95@GitHub](https://github.com/yangheng95)),
@@ -193,8 +199,6 @@ many ways, including:
   hyper-parameters)
 - Star this repository to keep it active
 
-
-
 ## License
 
 PyABSA is released under MIT licence, please cite this repo (or papers) or attach the author information in your work

examples-v2/aspect_polarity_classification/checkpoints-v2.0.json

@@ -1 +0,0 @@
-{"2.0.0": {"APC": {"multilingual": {"id": "", "Training Model": "FAST-LSA-T-V2-Deberta", "Training Dataset": "APCDatasetList.Multilingual", "Language": "Multilingual", "Description": "Trained on RTX3090", "Available Version": "1.10.5+", "Checkpoint File": "fast_lcf_bert_Multilingual_acc_87.18_f1_83.11.zip", "Author": "H, Yang (hy345@exeter.ac.uk)"}, "multilingual2": {"id": "", "Training Model": "FAST-LSA-T-V2-Deberta", "Training Dataset": "APCDatasetList.Multilingual", "Language": "Multilingual", "Description": "Trained on RTX3090", "Available Version": "1.10.5+", "Checkpoint File": "fast_lcf_bert_Multilingual_acc_82.66_f1_82.06.zip", "Author": "H, Yang (hy345@exeter.ac.uk)"}, "english": {"id": "", "Training Model": "FAST-LSA-T-V2-Deberta", "Training Dataset": "APCDatasetList.English", "Language": "English", "Description": "Trained on RTX3090", "Available Version": "1.10.5+", "Checkpoint File": "fast_lsa_t_v2_English_acc_82.21_f1_81.81.zip", "Author": "H, Yang (hy345@exeter.ac.uk)"}, "chinese": {"id": "", "Training Model": "FAST-LSA-T-V2-Deberta", "Training Dataset": "APCDatasetList.Chinese", "Language": "Chinese", "Description": "Trained on RTX3090", "Available Version": "1.10.5+", "Checkpoint File": "fast_lsa_t_v2_Chinese_acc_96.0_f1_95.1.zip", "Author": "H, Yang (hy345@exeter.ac.uk)"}}, "ATEPC": {"multilingual": {"id": "", "Training Model": "FAST-LCF-ATEPC", "Training Dataset": "ABSADatasets.Multilingual", "Language": "Multilingual", "Description": "Trained on RTX3090", "Available Version": "1.16.0+", "Checkpoint File": "fast_lcf_atepc_Multilingual_cdw_apcacc_80.81_apcf1_73.75_atef1_76.01.zip", "Author": "H, Yang (hy345@exeter.ac.uk)"}, "multilingual2": {"id": "", "Training Model": "FAST-LCF-ATEPC", "Training Dataset": "ABSADatasets.Multilingual", "Language": "Multilingual", "Description": "Trained on RTX3090", "Available Version": "1.16.0+", "Checkpoint File": "fast_lcf_atepc_Multilingual_cdw_apcacc_78.08_apcf1_77.81_atef1_75.41.zip", "Author": "H, Yang (hy345@exeter.ac.uk)"}, "english": {"id": "", "Training Model": "FAST-LCF-ATEPC", "Training Dataset": "ATEPCDatasetList.English", "Language": "English", "Description": "Trained on RTX3090", "Available Version": "1.10.5+", "Checkpoint File": "fast_lcf_atepc_English_cdw_apcacc_82.36_apcf1_81.89_atef1_75.43.zip", "Author": "H, Yang (hy345@exeter.ac.uk)"}, "chinese": {"id": "", "Training Model": "FAST-LCF-ATEPC", "Training Dataset": "ATEPCDatasetList.Chinese", "Language": "Chinese", "Description": "Trained on RTX3090", "Available Version": "1.10.5+", "Checkpoint File": "fast_lcf_atepc_Chinese_cdw_apcacc_96.22_apcf1_95.32_atef1_78.73.zip", "Author": "H, Yang (hy345@exeter.ac.uk)"}}, "RNAC": {"degrad_lstm": {"id": "", "Training Model": "LSTM", "Training Dataset": "ABSADatasets.Multilingual", "Language": "RNA", "Description": "Trained on RTX3090", "Available Version": "1.16.0+", "Checkpoint File": "lstm_degrad_acc_85.26_f1_84.62.zip", "Author": "H, Yang (hy345@exeter.ac.uk)"}, "degrad_bert": {"id": "", "Training Model": "MLP", "Training Dataset": "Degrad", "Language": "RNA", "Description": "Trained on RTX3090", "Available Version": "1.16.0+", "Checkpoint File": "bert_mlp_degrad_acc_87.44_f1_86.99.zip", "Author": "H, Yang (hy345@exeter.ac.uk)"}}, "TAD": {"tad-sst2": {"id": "", "Training Model": "TAD", "Training Dataset": "SST2", "Language": "English", "Description": "Trained on RTX3090", "Available Version": "1.15+", "Checkpoint File": "TAD-SST2.zip", "Author": "H, Yang (yangheng@m.scnu.edu.cn)"}, "tad-agnews10k": {"id": "", "Training Model": "TAD", "Training Dataset": "AGNews", "Language": "English", "Description": "Trained on RTX3090", "Available Version": "1.15+", "Checkpoint File": "TAD-AGNews10K.zip", "Author": "H, Yang (yangheng@m.scnu.edu.cn)"}, "tad-amazon": {"id": "", "Training Model": "TAD", "Training Dataset": "AGNews", "Language": "English", "Description": "Trained on RTX3090", "Available Version": "1.15+", "Checkpoint File": "TAD-Amazon.zip", "Author": "H, Yang (yangheng@m.scnu.edu.cn)"}}, "CDD": {"promise": {"id": "", "Training Model": "CodeT5-small", "Training Dataset": "Promise", "Language": "Code", "Description": "Trained on RTX3090", "Available Version": "1.16.0+", "Checkpoint File": "bert_mlp_all_cpdp_acc_75.33_f1_73.52.zip", "Author": "H, Yang (hy345@exeter.ac.uk)"}}, "UPPERTASKCODE": {"promise": {"id": "", "Training Model": "CodeT5-small", "Training Dataset": "DatasetName", "Language": "", "Description": "Trained on RTX3090", "Available Version": "1.16.0+", "Checkpoint File": "lstm_degrad_acc_85.26_f1_84.62.zip", "Author": "H, Yang (hy345@exeter.ac.uk)"}}}}

examples-v2/aspect_polarity_classification/ensemble_inference.py

@@ -39,7 +39,13 @@ def ensemble_performance(dataset, print_result=False):
 if __name__ == "__main__":
     # Training the models before ensemble inference, take Laptop14 as an example
 
-    for dataset in [Laptop14, Restaurant14, Restaurant15, Restaurant16, MAMS]:
+    for dataset in [
+        APC.APCDatasetList.Laptop14,
+        APC.APCDatasetList.Restaurant14,
+        APC.APCDatasetList.Restaurant15,
+        APC.APCDatasetList.Restaurant16,
+        APC.APCDatasetList.MAMS
+    ]:
         # Training
         pass
     # Ensemble inference

examples-v2/aspect_polarity_classification/train_apc.py

@@ -31,14 +31,14 @@
     APC.APCDatasetList.MAMS,
 ]:
     for model in [
-        APC.APCModelList.FAST_LSA_T_V2,
-        APC.APCModelList.FAST_LSA_S_V2,
+        # APC.APCModelList.FAST_LSA_T_V2,
+        # APC.APCModelList.FAST_LSA_S_V2,
         APC.APCModelList.BERT_SPC_V2,
         # APC.APCModelList.BERT_SPC
     ]:
         for pretrained_bert in [
-            "microsoft/deberta-v3-base",
-            # "bert-base-uncased",
+            # "microsoft/deberta-v3-base",
+            "bert-base-uncased",
             # 'roberta-base',
             # 'microsoft/deberta-v3-large',
         ]:
@@ -47,7 +47,7 @@
             config.pretrained_bert = pretrained_bert
             # config.pretrained_bert = 'roberta-base'
             config.evaluate_begin = 0
-            config.max_seq_len = 512
+            config.max_seq_len = 80
             config.num_epoch = 30
             # config.log_step = 5
             config.log_step = -1

examples-v2/aspect_term_extraction/train_atepc.py

@@ -17,8 +17,8 @@
 config.evaluate_begin = 0
 config.max_seq_len = 128
 config.batch_size = 16
-# config.pretrained_bert = 'yangheng/deberta-v3-base-absa'
-config.pretrained_bert = "microsoft/mdeberta-v3-base"
+config.pretrained_bert = 'yangheng/deberta-v3-base-absa'
+# config.pretrained_bert = "microsoft/mdeberta-v3-base"
 config.log_step = -1
 config.l2reg = 1e-8
 config.num_epoch = 20
@@ -28,12 +28,12 @@
 config.cache_dataset = True
 config.cross_validate_fold = -1
 
-# chinese_sets = ATEPC.ATEPCDatasetList.Chinese_Zhang
-chinese_sets = ATEPC.ATEPCDatasetList.Multilingual
+chinese_sets = ATEPC.ATEPCDatasetList.Chinese_Zhang
+# chinese_sets = ATEPC.ATEPCDatasetList.Multilingual
 
 aspect_extractor = ATEPC.ATEPCTrainer(
     config=config,
-    # from_checkpoint="",   # not necessary for most situations
+    from_checkpoint="english",   # not necessary for most situations
     dataset=chinese_sets,
     checkpoint_save_mode=1,
     auto_device=True,

examples-v2/text_classification/train_text_classification.py

@@ -13,7 +13,7 @@
 classification_config_english.num_epoch = 20
 classification_config_english.batch_size = 16
 classification_config_english.evaluate_begin = 0
-classification_config_english.max_seq_len = 100
+classification_config_english.max_seq_len = 512
 classification_config_english.learning_rate = 5e-5
 classification_config_english.dropout = 0
 classification_config_english.seed = {42, 14, 5324}
@@ -22,7 +22,7 @@
 # classification_config_english.use_amp = True
 classification_config_english.cache_dataset = False
 
-SST2 = "evoprompt"
+SST2 = "sst2"
 sent_classifier = TC.TCTrainer(
     config=classification_config_english,
     dataset=SST2,

pyabsa/__init__.py

@@ -7,7 +7,7 @@
 # Copyright (C) 2021. All Rights Reserved.
 
 __name__ = "pyabsa"
-__version__ = "2.3.4rc0"
+__version__ = "2.3.4"
 
 
 from pyabsa.utils.notification_utils.notification_utils import (

pyabsa/tasks/ABSAInstruction/data_utils.py

@@ -22,12 +22,12 @@
 
 class InstructDatasetLoader:
     def __init__(
-        self,
-        train_df_id,
-        test_df_id,
-        train_df_ood=None,
-        test_df_ood=None,
-        sample_size=1,
+            self,
+            train_df_id,
+            test_df_id,
+            train_df_ood=None,
+            test_df_ood=None,
+            sample_size=1,
     ):
         self.train_df_id = train_df_id.sample(frac=sample_size, random_state=1999)
         self.test_df_id = test_df_id

pyabsa/tasks/ABSAInstruction/instruction.py

@@ -85,10 +85,10 @@ def __init__(self, bos_instruction=None, eos_instruction=None):
 
     def prepare_input(self, input_text, aspects):
         return (
-            self.bos_instruction
-            + input_text
-            + f"The aspects are: {aspects}"
-            + self.eos_instruction
+                self.bos_instruction
+                + input_text
+                + f"The aspects are: {aspects}"
+                + self.eos_instruction
         )
 
 
@@ -123,10 +123,10 @@ def __init__(self, bos_instruction=None, eos_instruction=None):
 
     def prepare_input(self, input_text, aspects):
         return (
-            self.bos_instruction
-            + input_text
-            + f"The aspects are: {aspects}"
-            + self.eos_instruction
+                self.bos_instruction
+                + input_text
+                + f"The aspects are: {aspects}"
+                + self.eos_instruction
         )
 
 
@@ -161,8 +161,8 @@ def __init__(self, bos_instruction=None, eos_instruction=None):
 
     def prepare_input(self, input_text, aspects):
         return (
-            self.bos_instruction
-            + input_text
-            + f"The aspects are: {aspects}"
-            + self.eos_instruction
+                self.bos_instruction
+                + input_text
+                + f"The aspects are: {aspects}"
+                + self.eos_instruction
         )

pyabsa/tasks/ABSAInstruction/model.py

@@ -154,12 +154,12 @@ def predict(self, text, **kwargs):
         return ensemble_result
 
     def get_labels(
-        self,
-        tokenized_dataset,
-        trained_model_path=None,
-        predictor=None,
-        batch_size=4,
-        sample_set="train",
+            self,
+            tokenized_dataset,
+            trained_model_path=None,
+            predictor=None,
+            batch_size=4,
+            sample_set="train",
     ):
         """
         Get the predictions from the trained model.
@@ -315,7 +315,7 @@ def train(self, tokenized_datasets, **kwargs):
         return trainer
 
     def get_labels(
-        self, tokenized_dataset, predictor=None, batch_size=4, sample_set="train"
+            self, tokenized_dataset, predictor=None, batch_size=4, sample_set="train"
     ):
         """
         Get the predictions from the trained model.

pyabsa/tasks/AspectPolarityClassification/dataset_utils/__classic__/classic_glove_apc_utils.py

@@ -53,7 +53,7 @@ def syntax_distance_alignment(tokens, dist, max_seq_len, tokenizer):
 
                     text = text[1:]
                     dep_dist = dep_dist[1:]
-                    bert_tokens = bert_tokens[len(tmp_tokens) :]
+                    bert_tokens = bert_tokens[len(tmp_tokens):]
                 else:
                     text = text[1:]
                     bert_tokens = bert_tokens[1:]
@@ -109,7 +109,7 @@ def prepare_input_for_apc(config, tokenizer, text_left, text_right, aspect):
 
     text_raw = text_left + " " + aspect + " " + text_right
     text_spc = (
-        bos_token + " " + text_raw + " " + eos_token + " " + aspect + " " + eos_token
+            bos_token + " " + text_raw + " " + eos_token + " " + aspect + " " + eos_token
     )
     text_indices = text_to_sequence(tokenizer, text_spc, config.max_seq_len)
     text_raw_bert_indices = text_to_sequence(
@@ -186,7 +186,7 @@ def get_syntax_distance(text_raw, aspect, tokenizer, config):
 
 
 def get_lca_ids_and_cdm_vec(
-    config, bert_spc_indices, aspect_indices, aspect_begin, syntactical_dist=None
+        config, bert_spc_indices, aspect_indices, aspect_begin, syntactical_dist=None
 ):
     SRD = config.SRD
     cdm_vec = np.zeros((config.max_seq_len), dtype=np.int64)
@@ -206,7 +206,7 @@ def get_lca_ids_and_cdm_vec(
 
 
 def get_cdw_vec(
-    config, bert_spc_indices, aspect_indices, aspect_begin, syntactical_dist=None
+        config, bert_spc_indices, aspect_indices, aspect_begin, syntactical_dist=None
 ):
     SRD = config.SRD
     cdw_vec = np.zeros((config.max_seq_len), dtype=np.float32)
@@ -246,15 +246,15 @@ def build_spc_mask_vec(config, text_ids):
 
 
 def build_sentiment_window(
-    examples, tokenizer, similarity_threshold, input_demands=None
+        examples, tokenizer, similarity_threshold, input_demands=None
 ):
     copy_side_aspect("left", examples[0], examples[0], examples, input_demands)
     for idx in range(1, len(examples)):
         if is_similar(
-            examples[idx - 1]["text_indices"],
-            examples[idx]["text_indices"],
-            tokenizer=None,
-            similarity_threshold=similarity_threshold,
+                examples[idx - 1]["text_indices"],
+                examples[idx]["text_indices"],
+                tokenizer=None,
+                similarity_threshold=similarity_threshold,
         ):
             copy_side_aspect(
                 "right", examples[idx - 1], examples[idx], examples, input_demands