ceshine

class Model(nn.Module):
    def __init__(self, nb_words, hidden_size=128, embedding_size=128, n_layers=1,
                 wdrop=0.25, edrop=0.1, idrop=0.25, batch_first=True):
        super(Model, self).__init__()
        # Modified LockedDropout that support batch first arrangement
        self.lockdrop = LockedDropout(batch_first=batch_first)
        self.idrop = idrop
        self.edrop = edrop
        self.n_layers = n_layers
        self.embedding = nn.Embedding(nb_words, embedding_size)

ceshine

checkpoint::checkpoint("2017-10-26")
pacman::p_load(data.table)
pacman::p_load(caret)
pacman::p_load(ggplot2)

set.seed(998)

mushrooms <- fread("mushrooms.csv", stringsAsFactors=T)
mushrooms[, .N, by=class]
mushrooms[, eval("veil-type") := NULL]

ceshine

# http://pytorch.org/docs/master/_modules/torch/optim/sgd.html#SGD

class SGD(Optimizer):

    def __init__(self, params, lr=required, momentum=0, dampening=0,
                 weight_decay=0, nesterov=False):
        # ...

    def __setstate__(self, state):
        # ...

ceshine

# Reference: http://pytorch.org/docs/master/_modules/torch/optim/sgd.html#SGD

class SGD(Optimizer):

    def __init__(self, params, lr=required, momentum=0, dampening=0,
                 weight_decay=0, nesterov=False):
        # ...

    def __setstate__(self, state):
        # ...

ceshine

""" Tested with Python 3.6 """
import re

import pandas as pd
import spacy
import joblib
from tqdm import tqdm

nlp = spacy.load('en')

ceshine

import pandas as pd
import joblib
from sklearn.model_selection import train_test_split


LABELS = ["toxic", "severe_toxic", "obscene",
          "threat", "insult", "identity_hate"]
EMPTY_ID = len(LABELS)

ceshine

import re
import logging

import numpy as np
import pandas as pd
import spacy
import torch
from torchtext import data

NLP = spacy.load('en')

ceshine

import pathlib

output_folder = "tmp/folder"
pathlib.Path(output_folder).mkdir(parents=True, exist_ok=True)

ceshine

"""Improved dataset loader for Toxic Comment dataset from Kaggle
Tested against:
* Python 3.6
* Numpy 1.14.0
* Pandas 0.22.0
* PyTorch 0.4.0a0+f83ca63 (should be very close to 0.3.0)
* torchtext 0.2.1
* spacy 2.0.5
* joblib 0.11
"""

ceshine

tf.reset_default_graph()
graph = tf.Graph()
with graph.as_default():
    tf.set_random_seed(10)
    # tf Graph input
    X = tf.placeholder("float", [None, timesteps, num_input])
    Y = tf.placeholder("float", [None, num_classes])
    is_training = tf.placeholder("bool")
    
    # Define weights