1.16.19a0

pyabsa/__init__.py

@@ -7,7 +7,7 @@
 # Copyright (C) 2021. All Rights Reserved.
 
 
-__version__ = '1.16.18'
+__version__ = '1.16.19a0'
 
 __name__ = 'pyabsa'
 

pyabsa/core/apc/models/ensembler.py

@@ -80,7 +80,7 @@ def __init__(self, opt, load_dataset=True, **kwargs):
             if hasattr(APCModelList, models[i].__name__):
                 try:
 
-                    if kwargs.pop('offline', False):
+                    if kwargs.get('offline', False):
                         self.tokenizer = AutoTokenizer.from_pretrained(find_cwd_dir(self.opt.pretrained_bert.split('/')[-1]), do_lower_case='uncased' in self.opt.pretrained_bert)
                         self.bert = AutoModel.from_pretrained(find_cwd_dir(self.opt.pretrained_bert.split('/')[-1])) if not self.bert else self.bert  # share the underlying bert between models
                     else:

pyabsa/core/apc/prediction/sentiment_classifier.py

@@ -74,7 +74,7 @@ def __init__(self, model_arg=None, cal_perplexity=False, **kwargs):
 
                 with open(config_path, mode='rb') as f:
                     self.opt = pickle.load(f)
-                    self.opt.device = get_device(kwargs.pop('auto_device', True))[0]
+                    self.opt.device = get_device(kwargs.get('auto_device', True))[0]
 
                 if state_dict_path or model_path:
                     if state_dict_path:
@@ -85,7 +85,7 @@ def __init__(self, model_arg=None, cal_perplexity=False, **kwargs):
                     with open(tokenizer_path, mode='rb') as f:
                         if hasattr(APCModelList, self.opt.model.__name__):
                             try:
-                                if kwargs.pop('offline', False):
+                                if kwargs.get('offline', False):
                                     self.tokenizer = AutoTokenizer.from_pretrained(find_cwd_dir(self.opt.pretrained_bert.split('/')[-1]), do_lower_case='uncased' in self.opt.pretrained_bert)
                                 else:
                                     self.tokenizer = AutoTokenizer.from_pretrained(self.opt.pretrained_bert, do_lower_case='uncased' in self.opt.pretrained_bert)
@@ -109,7 +109,7 @@ def __init__(self, model_arg=None, cal_perplexity=False, **kwargs):
 
                             self.tokenizer = tokenizer
 
-                if kwargs.pop('verbose', False):
+                if kwargs.get('verbose', False):
                     print('Config used in Training:')
                     print_args(self.opt)
 
@@ -153,7 +153,7 @@ def __init__(self, model_arg=None, cal_perplexity=False, **kwargs):
         #     torch.backends.cudnn.benchmark = False
 
         self.opt.initializer = self.opt.initializer
-        self.opt.eval_batch_size = kwargs.pop('eval_batch_size', 128)
+        self.opt.eval_batch_size = kwargs.get('eval_batch_size', 128)
 
         if self.cal_perplexity:
             try:

pyabsa/core/atepc/prediction/aspect_extractor.py

@@ -59,10 +59,10 @@ def __init__(self, model_arg=None, **kwargs):
 
                 with open(config_path, mode='rb') as f:
                     self.opt = pickle.load(f)
-                    self.opt.device = get_device(kwargs.pop('auto_device', True))[0]
+                    self.opt.device = get_device(kwargs.get('auto_device', True))[0]
                 if state_dict_path:
                     try:
-                        if kwargs.pop('offline', False):
+                        if kwargs.get('offline', False):
                             bert_base_model = AutoModel.from_pretrained(find_cwd_dir(self.opt.pretrained_bert.split('/')[-1]))
                         else:
                             bert_base_model = AutoModel.from_pretrained(self.opt.pretrained_bert)
@@ -76,7 +76,7 @@ def __init__(self, model_arg=None, **kwargs):
                     self.model = torch.load(model_path, map_location='cpu')
                     self.model.opt = self.opt
                 try:
-                    if kwargs.pop('offline', False):
+                    if kwargs.get('offline', False):
                         self.tokenizer = AutoTokenizer.from_pretrained(find_cwd_dir(self.opt.pretrained_bert.split('/')[-1]))
                     else:
                         self.tokenizer = AutoTokenizer.from_pretrained(self.opt.pretrained_bert, do_lower_case='uncased' in self.opt.pretrained_bert)
@@ -102,7 +102,7 @@ def __init__(self, model_arg=None, **kwargs):
         # np.random.seed(self.opt.seed)
         # torch.manual_seed(self.opt.seed)
 
-        if kwargs.pop('verbose', False):
+        if kwargs.get('verbose', False):
             print('Config used in Training:')
             print_args(self.opt)
 
@@ -112,7 +112,7 @@ def __init__(self, model_arg=None, **kwargs):
                 self.opt.gradient_accumulation_steps))
 
         self.eval_dataloader = None
-        self.opt.eval_batch_size = kwargs.pop('eval_batch_size', 128)
+        self.opt.eval_batch_size = kwargs.get('eval_batch_size', 128)
 
         self.to(self.opt.device)
 

pyabsa/core/tad/classic/__bert__/dataset_utils/data_utils_for_inference.py

@@ -14,7 +14,7 @@
 
 class Tokenizer4Pretraining:
     def __init__(self, max_seq_len, opt, **kwargs):
-        if kwargs.pop('offline', False):
+        if kwargs.get('offline', False):
             self.tokenizer = AutoTokenizer.from_pretrained(find_cwd_dir(opt.pretrained_bert.split('/')[-1]),
                                                            do_lower_case='uncased' in opt.pretrained_bert)
         else:

pyabsa/core/tad/prediction/tad_classifier.py

@@ -125,12 +125,12 @@ def __init__(self, model_arg=None, cal_perplexity=False, **kwargs):
 
                 with open(config_path, mode='rb') as f:
                     self.opt = pickle.load(f)
-                    self.opt.device = get_device(kwargs.pop('auto_device', True))[0]
+                    self.opt.device = get_device(kwargs.get('auto_device', True))[0]
 
                 if state_dict_path or model_path:
                     if hasattr(BERTTADModelList, self.opt.model.__name__):
                         if state_dict_path:
-                            if kwargs.pop('offline', False):
+                            if kwargs.get('offline', False):
                                 self.bert = AutoModel.from_pretrained(
                                     find_cwd_dir(self.opt.pretrained_bert.split('/')[-1]))
                             else:
@@ -172,7 +172,7 @@ def __init__(self, model_arg=None, cal_perplexity=False, **kwargs):
 
                         self.tokenizer = tokenizer
 
-                if kwargs.pop('verbose', False):
+                if kwargs.get('verbose', False):
                     print('Config used in Training:')
                     print_args(self.opt)
 
@@ -184,7 +184,7 @@ def __init__(self, model_arg=None, cal_perplexity=False, **kwargs):
                 raise KeyError('The checkpoint you are loading is not from classifier model.')
 
         self.infer_dataloader = None
-        self.opt.eval_batch_size = kwargs.pop('eval_batch_size', 128)
+        self.opt.eval_batch_size = kwargs.get('eval_batch_size', 128)
 
         # if self.opt.seed is not None:
         #     random.seed(self.opt.seed)

pyabsa/core/tc/classic/__bert__/dataset_utils/data_utils_for_inference.py

@@ -14,7 +14,7 @@
 
 class Tokenizer4Pretraining:
     def __init__(self, max_seq_len, opt, **kwargs):
-        if kwargs.pop('offline', False):
+        if kwargs.get('offline', False):
             self.tokenizer = AutoTokenizer.from_pretrained(find_cwd_dir(opt.pretrained_bert.split('/')[-1]),
                                                            do_lower_case='uncased' in opt.pretrained_bert)
         else:

pyabsa/core/tc/prediction/text_classifier.py

@@ -74,12 +74,12 @@ def __init__(self, model_arg=None, cal_perplexity=False, **kwargs):
 
                 with open(config_path, mode='rb') as f:
                     self.opt = pickle.load(f)
-                    self.opt.device = get_device(kwargs.pop('auto_device', True))[0]
+                    self.opt.device = get_device(kwargs.get('auto_device', True))[0]
 
                 if state_dict_path or model_path:
                     if hasattr(BERTTCModelList, self.opt.model.__name__):
                         if state_dict_path:
-                            if kwargs.pop('offline', False):
+                            if kwargs.get('offline', False):
                                 self.bert = AutoModel.from_pretrained(
                                     find_cwd_dir(self.opt.pretrained_bert.split('/')[-1]))
                             else:
@@ -121,7 +121,7 @@ def __init__(self, model_arg=None, cal_perplexity=False, **kwargs):
 
                         self.tokenizer = tokenizer
 
-                if kwargs.pop('verbose', False):
+                if kwargs.get('verbose', False):
                     print('Config used in Training:')
                     print_args(self.opt)
 
@@ -139,7 +139,7 @@ def __init__(self, model_arg=None, cal_perplexity=False, **kwargs):
             self.dataset = GloVeTCDataset(tokenizer=self.tokenizer, opt=self.opt)
 
         self.infer_dataloader = None
-        self.opt.eval_batch_size = kwargs.pop('eval_batch_size', 128)
+        self.opt.eval_batch_size = kwargs.get('eval_batch_size', 128)
 
         # if self.opt.seed is not None:
         #     random.seed(self.opt.seed)