SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

Added RoBERTa tokenizer to AutoTokenizer

Browse Source
This commit is contained in:
LysandreJik
2019-08-15 17:07:07 -04:00
parent e24e19ce3b
commit 83dba0b67b
2 changed files with 8 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
pytorch_transformers/modeling_auto.py
View File

@@ -127,7 +127,7 @@ class AutoConfig(object):
raise ValueError("Unrecognized model identifier in {}. Should contains one of " raise ValueError("Unrecognized model identifier in {}. Should contains one of "
"'bert', 'openai-gpt', 'gpt2', 'transfo-xl', 'xlnet', " "'bert', 'openai-gpt', 'gpt2', 'transfo-xl', 'xlnet', "
"'xlm'".format(pretrained_model_name_or_path)) "'xlm', 'roberta'".format(pretrained_model_name_or_path))
class AutoModel(object): class AutoModel(object):
@@ -242,4 +242,4 @@ class AutoModel(object):
raise ValueError("Unrecognized model identifier in {}. Should contains one of " raise ValueError("Unrecognized model identifier in {}. Should contains one of "
"'bert', 'openai-gpt', 'gpt2', 'transfo-xl', 'xlnet', " "'bert', 'openai-gpt', 'gpt2', 'transfo-xl', 'xlnet', "
"'xlm'".format(pretrained_model_name_or_path)) "'xlm', 'roberta'".format(pretrained_model_name_or_path))

7
pytorch_transformers/tokenization_auto.py
View File

@@ -24,6 +24,7 @@ from .tokenization_gpt2 import GPT2Tokenizer
from .tokenization_transfo_xl import TransfoXLTokenizer from .tokenization_transfo_xl import TransfoXLTokenizer
from .tokenization_xlnet import XLNetTokenizer from .tokenization_xlnet import XLNetTokenizer
from .tokenization_xlm import XLMTokenizer from .tokenization_xlm import XLMTokenizer
from .tokenization_roberta import RobertaTokenizer
logger = logging.getLogger(__name__) logger = logging.getLogger(__name__)
@@ -44,6 +45,7 @@ class AutoTokenizer(object):
- contains `transfo-xl`: TransfoXLTokenizer (Transformer-XL model) - contains `transfo-xl`: TransfoXLTokenizer (Transformer-XL model)
- contains `xlnet`: XLNetTokenizer (XLNet model) - contains `xlnet`: XLNetTokenizer (XLNet model)
- contains `xlm`: XLMTokenizer (XLM model) - contains `xlm`: XLMTokenizer (XLM model)
- contains `roberta`: RobertaTokenizer (RoBERTa model)
This class cannot be instantiated using `__init__()` (throw an error). This class cannot be instantiated using `__init__()` (throw an error).
""" """
@@ -64,6 +66,7 @@ class AutoTokenizer(object):
- contains `transfo-xl`: TransfoXLTokenizer (Transformer-XL model) - contains `transfo-xl`: TransfoXLTokenizer (Transformer-XL model)
- contains `xlnet`: XLNetTokenizer (XLNet model) - contains `xlnet`: XLNetTokenizer (XLNet model)
- contains `xlm`: XLMTokenizer (XLM model) - contains `xlm`: XLMTokenizer (XLM model)
- contains `roberta`: RobertaTokenizer (XLM model)
Params: Params:
**pretrained_model_name_or_path**: either: **pretrained_model_name_or_path**: either:
@@ -94,7 +97,9 @@ class AutoTokenizer(object):
return XLNetTokenizer.from_pretrained(pretrained_model_name_or_path, *inputs, **kwargs) return XLNetTokenizer.from_pretrained(pretrained_model_name_or_path, *inputs, **kwargs)
elif 'xlm' in pretrained_model_name_or_path: elif 'xlm' in pretrained_model_name_or_path:
return XLMTokenizer.from_pretrained(pretrained_model_name_or_path, *inputs, **kwargs) return XLMTokenizer.from_pretrained(pretrained_model_name_or_path, *inputs, **kwargs)
elif 'roberta' in pretrained_model_name_or_path:
return RobertaTokenizer.from_pretrained(pretrained_model_name_or_path, *inputs, **kwargs)
raise ValueError("Unrecognized model identifier in {}. Should contains one of " raise ValueError("Unrecognized model identifier in {}. Should contains one of "
"'bert', 'openai-gpt', 'gpt2', 'transfo-xl', 'xlnet', " "'bert', 'openai-gpt', 'gpt2', 'transfo-xl', 'xlnet', "
"'xlm'".format(pretrained_model_name_or_path)) "'xlm', 'roberta'".format(pretrained_model_name_or_path))