def read_one_example(self, inputs):
""" inputs keys: sequence_a and sequence_b """
sequence_a = utils.get_sequence_a(inputs)
sequence_b = inputs.get("sequence_b", None)
sequence_a_sub_tokens = self.subword_tokenizer.tokenize(sequence_a)
bert_input = [self.CLS_TOKEN
] + sequence_a_sub_tokens + [self.SEP_TOKEN]
if sequence_b:
sequence_b_sub_tokens = self.subword_tokenizer.tokenize(sequence_b)
bert_input += sequence_b_sub_tokens + [self.SEP_TOKEN]
if len(bert_input) > self.sequence_max_length:
bert_input = bert_input[:self.sequence_max_length -
1] + [self.SEP_TOKEN]
token_type = utils.make_bert_token_type(bert_input,
SEP_token=self.SEP_TOKEN)
features = []
features.append({
"bert_input": bert_input,
"token_type": {
"feature": token_type,
"text": ""
}, # TODO: fix hard-code
})
return features, {}
def read_one_example(self, inputs):
""" inputs keys: sequence_a and sequence_b """
sequence_a = utils.get_sequence_a(inputs)
sequence_b = inputs.get("sequence_b", None)
bert_feature = BertFeature()
bert_feature.set_input_with_speical_token(
sequence_a,
sequence_b,
self.tokenizer,
max_seq_length=self.sequence_max_length,
data_type="predict",
cls_token=self.cls_token,
sep_token=self.sep_token,
input_type=self.input_type,
)
features = [bert_feature.to_dict()]
helper = {}
return features, helper
def _read(self, file_path, data_type=None):
"""
.json file structure should be something like this:
{
"data": [
{
"sequence_a": "what a wonderful day!",
"sequence_b": "what a great day!",
"score": 0.9
},
...
]
}
"""
data = self._get_data(file_path, data_type=data_type)
helper = {
"file_path": file_path,
"examples": {},
"cls_token": self.CLS_TOKEN,
"sep_token": self.SEP_TOKEN,
"unk_token": self.UNK_TOKEN,
"model": {},
"predict_helper": {}
}
features, labels = [], []
for example in tqdm(data, desc=data_type):
sequence_a = utils.get_sequence_a(example)
sequence_b = example.get("sequence_b", None)
sequence_a_sub_tokens = self.subword_tokenizer.tokenize(sequence_a)
sequence_b_sub_tokens = None
bert_input = [self.CLS_TOKEN
] + sequence_a_sub_tokens + [self.SEP_TOKEN]
if sequence_b is not None:
sequence_b_sub_tokens = self.subword_tokenizer.tokenize(
sequence_b)
bert_input += sequence_b_sub_tokens + [self.SEP_TOKEN]
if (self.sequence_max_length is not None and data_type == "train"
and len(bert_input) > self.sequence_max_length):
continue
if "uid" in example:
data_uid = example["uid"]
else:
data_uid = str(uuid.uuid1())
feature_row = {
"id": data_uid,
"bert_input": bert_input,
}
features.append(feature_row)
score = example[self.label_key]
label_row = {
"id": data_uid,
"score": score,
}
labels.append(label_row)
helper["examples"][data_uid] = {
"sequence_a": sequence_a,
"sequence_a_sub_tokens": sequence_a_sub_tokens,
"sequence_b": sequence_b,
"sequence_b_sub_tokens": sequence_b_sub_tokens,
"score": score,
}
if self.is_test and len(features) >= 10:
break
return make_batch(features, labels), helper
def _read(self, file_path, data_type=None):
"""
.json file structure should be something like this:
{
"data": [
{
"sequence_a": "what a wonderful day!",
"sequence_b": "what a great day!",
"score": 0.9
},
...
]
}
"""
data = self._get_data(file_path, data_type=data_type)
helper = Helper(**{
"file_path": file_path,
"cls_token": self.cls_token,
"sep_token": self.sep_token,
})
features, labels = [], []
for example in tqdm(data, desc=data_type):
sequence_a = utils.get_sequence_a(example)
sequence_b = example.get("sequence_b", None)
sequence_a_tokens = self.tokenizer.tokenize(sequence_a)
sequence_b_tokens = None
if sequence_b:
sequence_b_tokens = self.tokenizer.tokenize(sequence_b)
bert_input = utils.make_bert_input(
sequence_a,
sequence_b,
self.tokenizer,
max_seq_length=self.sequence_max_length,
data_type=data_type,
cls_token=self.cls_token,
sep_token=self.sep_token,
input_type=self.input_type,
)
if bert_input is None:
continue
if "uid" in example:
data_uid = example["uid"]
else:
data_uid = str(uuid.uuid1())
feature_row = {
"id": data_uid,
"bert_input": bert_input,
}
features.append(feature_row)
score = example[self.label_key]
label_row = {
"id": data_uid,
"score": score,
}
labels.append(label_row)
helper.set_example(data_uid, {
"sequence_a": sequence_a,
"sequence_a_tokens": sequence_a_tokens,
"sequence_b": sequence_b,
"sequence_b_tokens": sequence_b_tokens,
"score": score,
})
if self.is_test and len(features) >= 10:
break
return utils.make_batch(features, labels), helper.to_dict()
def _read(self, file_path, data_type=None):
"""
.json file structure should be something like this:
{
"data": [
{
"sequence": "what a wonderful day!",
"emotion": "happy"
},
...
],
"emotion": [ // class_key
"angry",
"happy",
"sad",
...
]
}
"""
data = self._get_data(file_path, data_type=data_type)
class_idx2text, class_text2idx = self._get_class_dicts(data=data)
helper = Helper(**{
"file_path": file_path,
"class_idx2text": class_idx2text,
"class_text2idx": class_text2idx,
"cls_token": self.cls_token,
"sep_token": self.sep_token,
"dataset": SeqClsBertDataset,
"metric_key": self.METRIC_KEY,
})
helper.set_model_parameter({
"num_classes": len(class_idx2text),
})
helper.set_predict_helper({
"class_idx2text": class_idx2text,
})
features, labels = [], []
for example in tqdm(data, desc=data_type):
sequence_a = utils.get_sequence_a(example)
sequence_b = example.get("sequence_b", None)
sequence_a_tokens = self.tokenizer.tokenize(sequence_a)
sequence_b_tokens = None
if sequence_b:
sequence_b_tokens = self.tokenizer.tokenize(sequence_b)
bert_input = utils.make_bert_input(
sequence_a,
sequence_b,
self.tokenizer,
max_seq_length=self.sequence_max_length,
data_type=data_type,
cls_token=self.cls_token,
sep_token=self.sep_token,
input_type=self.input_type,
)
if bert_input is None:
continue
if "uid" in example:
data_uid = example["uid"]
else:
data_uid = str(uuid.uuid1())
# token_type(segment_ids) will be added in dataset
feature_row = {
"id": data_uid,
"bert_input": bert_input,
}
features.append(feature_row)
class_text = example[self.class_key]
label_row = {
"id": data_uid,
"class_idx": class_text2idx[class_text],
"class_text": class_text,
}
labels.append(label_row)
helper.set_example(data_uid, {
"sequence_a": sequence_a,
"sequence_a_tokens": sequence_a_tokens,
"sequence_b": sequence_b,
"sequence_b_tokens": sequence_b_tokens,
"class_idx": class_text2idx[class_text],
"class_text": class_text,
})
if self.is_test and len(features) >= 10:
break
return utils.make_batch(features, labels), helper.to_dict()
