jiahao87

"""Script for fine-tuning Pegasus
Example usage:
  # use XSum dataset as example, with first 1000 docs as training data
  from datasets import load_dataset
  dataset = load_dataset("xsum")
  train_texts, train_labels = dataset['train']['document'][:1000], dataset['train']['summary'][:1000]
  
  # use Pegasus Large model as base for fine-tuning
  model_name = 'google/pegasus-large'
  train_dataset, _, _, tokenizer = prepare_data(model_name, train_texts, train_labels)

jkleint

#!/usr/bin/env python
"""
Example of using Keras to implement a 1D convolutional neural network (CNN) for timeseries prediction.
"""

from __future__ import print_function, division

import numpy as np
from keras.layers import Convolution1D, Dense, MaxPooling1D, Flatten
from keras.models import Sequential