def model_load(self, path):
s3_model_url = 'https://distilbert-finetuned-model.s3.eu-west-2.amazonaws.com/pytorch_model.bin'
path_to_model = download_model(s3_model_url, model_name="pytorch_model.bin")
config = DistilBertConfig.from_pretrained(path + "/config.json")
tokenizer = DistilBertTokenizer.from_pretrained(path, do_lower_case=self.do_lower_case)
model = DistilBertForQuestionAnswering.from_pretrained(path_to_model, from_tf=False, config=config)
return model, tokenizer
def model_load(self, path):
s3_model_url = 'https://storage.googleapis.com/bertpepper/pepperqa/pytorch_model.bin'
print('******************************************')
path_to_model = download_model(s3_model_url,
model_name="pytorch_model.bin")
qa_pipeline = pipeline("question-answering",
model=path,
tokenizer=path)
return qa_pipeline
embedding_output = self.embeddings(
input_ids=input_ids, position_ids=position_ids,
token_type_ids=token_type_ids
)
encoder_outputs = self.encoder(
embedding_output,
attention_mask=extended_attention_mask,
head_mask=head_mask,
encoder_hidden_states=encoder_hidden_states,
encoder_attention_mask=encoder_extended_attention_mask,
)
sequence_output = encoder_outputs[0]
pooled_output = self.pooler(sequence_output)
outputs = (sequence_output, pooled_output,) + encoder_outputs[
1:
] # add hidden_states and attentions if they are here
return outputs # sequence_output, pooled_output, (hidden_states), (attentions)
if __name__ == "__main__":
import logging
from download import download_model
logging.basicConfig(level=logging.INFO)
download_model('medical_character_bert')
path = "pretrained-models/medical_character_bert/"
model = CharacterBertModel.from_pretrained(path)
logging.info('%s', model)
def download(uid, path):
path.mkdir(parents=True, exist_ok=True)
try:
download_model(uid, str(path))
except Exception as exc:
print(f"Failed to download {uid}: {exc}")