목록Python/TA (5)

무회blog

200814,test bert 001

In [1]: # conda install -c conda-forge ipywidgets In [2]: from fast_bert.data_cls import BertDataBunch DATA_PATH = './../Downloads/fastai/fast-bert-1.8.0/sample_data/imdb_movie_reviews/data/' LABEL_PATH = './../Downloads/fastai/fast-bert-1.8.0/sample_data/imdb_movie_reviews/label/' OUTPUT_DIR = './../Downloads/fastai/fast-bert-1.8.0/sample_data/imdb_movie_reviews/output/'..
Python/TA 2020. 8. 14. 01:21
python: pytorch, model , train 시키기 ,

In [1]: import torch from transformers import BertForSequenceClassification model = BertForSequenceClassification.from_pretrained('bert-base-uncased') model.train() Some weights of the model checkpoint at bert-base-uncased were not used when initializing BertForSequenceClassification: ['cls.predictions.bias', 'cls.predictions.transform.dense.weight', 'cls.predictions.transform.dense.bias', 'cls...
Python/TA 2020. 8. 13. 18:31
python: pytorch, 질의응답관련 , bert 사용

In [4]: ## BertForMaskedLM from transformers import BertTokenizer, BertForMaskedLM import torch tokenizer = BertTokenizer.from_pretrained('bert-base-uncased') model = BertForMaskedLM.from_pretrained('bert-base-uncased') input_ids = tokenizer("Hello, my dog is cute", return_tensors="pt")["input_ids"] # print(input_ids) outputs = model(input_ids, labels=input_ids) loss, prediction_scores = outputs..
Python/TA 2020. 8. 13. 18:29
python: pytorch,transformer , pipline , 提问解答사용해보기

In [6]: from __future__ import print_function import ipywidgets as widgets from transformers import pipeline print('success') success In [7]: nlp_sentence_classif = pipeline('sentiment-analysis') nlp_sentence_classif('Such a nice weather outside !') Out[7]: [{'label': 'POSITIVE', 'score': 0.9997655749320984}] In [12]: nlp_token_class = pipeline('ner') nlp_token_class('Hugging Face is a French co..
Python/TA 2020. 8. 13. 18:18
python: pytorch,Bert, 로 토큰나이징 하기

In [1]: import torch from transformers import AutoModel,AutoTokenizer, BertTokenizer print(torch.__version__) torch.set_grad_enabled(False) 1.6.0 Out[1]: In [2]: # 모델 저장하기 , Store the model we want to use MODEL_NAME = "bert-base-cased" # We need to create the model and tokenizer model = AutoModel.from_pretrained(MODEL_NAME) tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME) print('model : ',m..
Python/TA 2020. 8. 13. 18:15
Prev 1 Next