def create_model(self, transformer="longformer"):
if transformer == "distilbert":
from transformers import DistilBertForQuestionAnswering
self.model = DistilBertForQuestionAnswering.from_pretrained(
"distilbert-base-uncased")
elif transformer == "bert":
from transformers import BertForQuestionAnswering
self.model = BertForQuestionAnswering.from_pretrained(
"bert-base-uncased")
elif transformer == "roberta":
from transformers import RobertaForQuestionAnswering
self.model = RobertaForQuestionAnswering.from_pretrained(
"roberta-base")
elif transformer == "roberta_squad":
from transformers import RobertaForQuestionAnswering
self.model = RobertaForQuestionAnswering.from_pretrained(
"deepset/roberta-base-squad2")
elif transformer == "longformer":
from transformers import LongformerForQuestionAnswering
self.model = LongformerForQuestionAnswering.from_pretrained(
"allenai/longformer-base-4096")
elif transformer == "bart":
from transformers import BartForQuestionAnswering
self.model = BartForQuestionAnswering.from_pretrained(
"facebook/bart-base")
elif transformer == "electra":
from transformers import ElectraForQuestionAnswering
self.model = ElectraForQuestionAnswering.from_pretrained(
"google/electra-small-discriminator")
else:
print(
"The model you chose is not available in this version. You can try to manually change the code or manually overwrite the variable self.model"
)
print(
"The available choices are 'distilbert' , 'bert' , 'roberta' , 'longformer' , 'bart' , 'electra' "
)
def configure_tokenizer_model_roberta(args, logger, is_preprocess=False):
logger.info("***** Loading tokenizer *****")
tokenizer = RobertaTokenizer.from_pretrained(args.config_name,
do_lower_case=args.do_lower_case,
cache_dir=args.init_dir,)
logger.info("***** Loading configuration *****")
config = RobertaConfig.from_pretrained(args.config_name, cache_dir=args.init_dir)
logger.info("Loading pretrained model from {}".format(args.init_dir))
if is_preprocess:
model = AutoModel.from_pretrained(args.model_name_or_path,
config=config,
cache_dir=args.init_dir)
else:
model = RobertaForQuestionAnswering.from_pretrained(args.init_dir, config=config,
cache_dir=args.init_dir)
return tokenizer, model
def __init__(self, hparams):
super().__init__(hparams)
self.net = RobertaForQuestionAnswering.from_pretrained(
self.hparams.model)