def main(args):
ds=Dataset('pubmed_rct20k')
Statistic(ds.documents_train,ds.labels_train)
return 0
def main(args):
ds = Dataset('20newsgroups')
documents = ds.documents_train
labels = ds.labels_train
pp = Preprocess(documents=documents, labels=labels, nb_sample=500)
documents = pp.documents
labels = pp.labels
cl = Classification(documents_train=documents, labels_train=labels)
cl.list_labels = pp.list_labels
cl.process()
print(cl.predict(documents[:5]))
return 0
def main(args):
ds = Dataset('Short_Jokes')
model = Model(model_name='gpt2')
text_loader = Create_DataLoader_generation(ds.documents_train[:10000],
batch_size=32)
model.load_type()
model.load_tokenizer()
model.load_class()
model.devices()
model.configuration(text_loader)
gn = Generation(model)
gn.model.fit_generation(text_loader)
output = model.predict_generation('What did you expect ?')
output_text = decode_text(output, model.tokenizer)
print(output_text)
return 0
from Manteia.Dataset import Dataset
ds = Dataset('Amazon Review Full', test=True, desc=True)
print('Train : ')
print(ds.documents_train[:5])
print(ds.labels_train[:5])
print('Test : ')
print(ds.documents_test[:5])
print(ds.labels_test[:5])
print('Description :')
print(ds.description)
from Manteia.Dataset import Dataset
ds=Dataset('pubmed_rct20k')
print('Train : ')
print(ds.documents_train[:5])
print(ds.labels_train[:5])
from Manteia.Dataset import Dataset
ds = Dataset('Yelp Review Full', test=True, desc=True)
print('Train : ')
print(ds.documents_train[:5])
print(ds.labels_train[:5])
print('Test : ')
print(ds.documents_test[:5])
print(ds.labels_test[:5])
print('Description :')
print(ds.description)
#modifier->parametre du notebook->GPU
#import nltk
#nltk.download('wordnet')
from Manteia.Classification import Classification
from Manteia.Model import *
from Manteia.Dataset import Dataset
from Manteia.Preprocess import list_labels
from Manteia.Augmentation import *
from sklearn.model_selection import train_test_split,KFold
ds=Dataset('drugscom')
ds.documents_train=np.array(ds.documents_train[:100])
ds.labels_train=np.array(ds.labels_train[:100])
model = Model(model_name ='bert',early_stopping=True)
model.load_type()
model.load_tokenizer()
list_label=list_labels(ds.labels_train)
print(list_label)
model.num_labels=len(list_label)
model.load_class()
model.save('model_init')
#validation croisée
nb_pass=4
def coss_validation_idx(nb_pass,nb_docs):
docs_idx = [idx for idx in range(nb_docs)]
train_idx, test_idx = [], []
for pli in range(nb_pass):
from Manteia.Dataset import Dataset
ds = Dataset('20newsgroups')
print('Train : ')
print(ds.documents_train[:5])
print(ds.labels_train[:5])
from Manteia.Dataset import Dataset
ds = Dataset('Yelp Review Polarity', test=True, desc=True)
print('Train : ')
print(ds.documents_train[:5])
print(ds.labels_train[:5])
print(ds.documents_test[:5])
print(ds.labels_test[:5])
print(ds.description)
from Manteia.Dataset import Dataset
ds = Dataset('Yahoo! Answers', test=True, desc=True)
print('Train : ')
print(ds.documents_train[:5])
print(ds.labels_train[:5])
print('Test : ')
print(ds.documents_test[:5])
print(ds.labels_test[:5])
print('Description :')
print(ds.description)
print('List labels :')
print(ds.list_labels)
from Manteia.Dataset import Dataset
ds=Dataset('drugscom')
print('Train : ')
print(ds.documents_train[:5])
print(ds.labels_train[:5])
from Manteia.Dataset import Dataset
ds = Dataset('Amazon Review Polarity', test=True, desc=True)
print('Train : ')
print(ds.documents_train[:5])
print(ds.labels_train[:5])
print(ds.documents_test[:5])
print(ds.labels_test[:5])
print(ds.description)
from Manteia.Dataset import Dataset
ds=Dataset('agnews')
print('Train : ')
print(ds.documents_train[:5])
print(ds.labels_train[:5])
from Manteia.Dataset import Dataset
ds=Dataset('DBPedia',path='test/test/',test=True,desc=True,classe=True)
print('Train : ')
print(ds.documents_train[:5])
print(ds.labels_train[:5])
print('Test : ')
print(ds.documents_test[:5])
print(ds.labels_test[:5])
print('Description :')
print(ds.description)
print('List labels :')
print(ds.list_labels)
from Manteia.Dataset import Dataset
ds = Dataset('SST-5', dev=True)
print('Dev : ')
print(ds.documents_dev[:5])
print(ds.labels_dev[:5])
from Manteia.Dataset import Dataset
ds = Dataset('Tweeter Airline Sentiment')
print('Train : ')
print(ds.documents_train[:5])
print(ds.labels_train[:5])
from Manteia.Dataset import Dataset
ds = Dataset('Short_Jokes')
print('Train : ')
print(ds.documents_train[:5])
from Manteia.Dataset import Dataset
ds = Dataset('SST-2')
print('Train : ')
print(ds.documents_train[:5])
print(ds.labels_train[:5])
from Manteia.Dataset import Dataset
ds = Dataset('trec')
print('Train : ')
print(ds.documents_train[:5])
print(ds.labels_train[:5])
