Generation

class Manteia.Generation.Generation(model=None, documents=None, seed=None)

This is the class to gnerate text in categorie a NLP task.

Args:

model_name (string, optional, defaults to ‘bert’):

give the name of a model.

documents (list, optional, defaults to None):

A list of documents.

labels (float, optional, defaults to None):

A list of labels.

Example:

from Manteia.Generation import Generation 
from Manteia.Dataset import Dataset
from Manteia.Model import *

ds=Dataset('Short_Jokes')

model       = Model(model_name ='gpt2-medium')
text_loader = Create_DataLoader_generation(ds.documents_train[:3000])
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)