Cleanup and reformat with autopep8

bot.py

@@ -1,10 +1,17 @@
-#!/usr/bin/python3
+from argparse import ArgumentParser
 
 from mastodon import Mastodon
 
+
+parser = ArgumentParser()
+parser.add_argument('-t', '--token', help='Mastodon application access token')
+args = parser.parse_args()
+
+
 mastodon = Mastodon(
-    access_token = 'token.secret',
+    access_token = args.token,
     api_base_url = 'https://social.exozy.me/'
 )
 
-mastodon.status_post('i am pretty horrible but i can keep reading the same dtb thing to be a really cool people there a notification in every fediverse in liberal democracies and games iaposd looks like to enjoy course at that goes and is it in case youaposre always welcome this time ago maybe you can start things that i think you want me to screenshot you use straw as an owl party platform in the algerian population for that have the list i just wanted to be a registry that you are out where the app actually the')
+
+mastodon.status_post('i am pretty horrible but i can keep reading the same dtb thing to be a really cool people there a notification in every fediverse in liberal democracies and games iaposd looks like to enjoy course at that goes and is it in case youaposre always welcome this time ago maybe you can start things that i think you want me to screenshot you use straw as an owl party platform in the algerian population for that have the list i just wanted to be a registry that you are out where the app actually the')

data.py

@@ -1,6 +1,6 @@
-from re import sub
-from html import unescape
 from argparse import ArgumentParser
+from html import unescape
+from re import sub
 
 from psycopg2 import connect
 
@@ -29,6 +29,7 @@ words = [sub(r'[^a-z0-9]', '', word.lower())
 words = [word for word in words if word != '']
 
 
+# Save to output file
 with open(args.output, 'w') as f:
     for word in words:
         f.write(word + '\n')

dataset.py

@@ -25,4 +25,4 @@ class Dataset(torch.utils.data.Dataset):
 
     def __getitem__(self, index):
         return (torch.tensor(self.words_indexes[index:index+self.seq_size]),
-                torch.tensor(self.words_indexes[index+1:index+self.seq_size+1]))
+                torch.tensor(self.words_indexes[index+1:index+self.seq_size+1]))

model.py

@@ -4,14 +4,14 @@ import torch.nn as nn
 class Model(nn.Module):
     def __init__(self, dataset, embedding_size, lstm_size, num_layers, dropout):
         super(Model, self).__init__()
-        
+
         self.seq_size = dataset.seq_size
         self.lstm_size = lstm_size
         self.num_layers = num_layers
 
         n_vocab = len(dataset.uniq_words)
         self.embedding = nn.Embedding(n_vocab, embedding_size)
-        
+
         self.lstm = nn.LSTM(
             input_size=embedding_size,
             hidden_size=lstm_size,
@@ -30,4 +30,4 @@ class Model(nn.Module):
 
     def zero_state(self, batch_size):
         return (torch.zeros(self.num_layers, batch_size, self.lstm_size),
-                torch.zeros(self.num_layers, batch_size, self.lstm_size))
+                torch.zeros(self.num_layers, batch_size, self.lstm_size))