Merge branch 'alpha0.0.1a4' into master

Author: codertimo

LICENSE

@@ -186,7 +186,7 @@
       same "printed page" as the copyright notice for easier
       identification within third-party archives.
 
-   Copyright 2018 Junseong Kim, Scatter Labs, BERT contributors
+   Copyright 2018 Junseong Kim, Scatter Lab, BERT contributors
 
    Licensed under the Apache License, Version 2.0 (the "License");
    you may not use this file except in compliance with the License.

README.md

@@ -60,7 +60,7 @@ bert-vocab -c data/corpus.small -o data/vocab.small
 
 ### 2. Train your own BERT model
 ```shell
-bert -c data/dataset.small -v data/vocab.small -o output/bert.model
+bert -c data/corpus.small -v data/vocab.small -o output/bert.model
 ```
 
 ## Language Model Pre-training

bert_pytorch/__main__.py

@@ -10,27 +10,30 @@
 def train():
     parser = argparse.ArgumentParser()
 
-    parser.add_argument("-c", "--train_dataset", required=True, type=str)
-    parser.add_argument("-t", "--test_dataset", type=str, default=None)
-    parser.add_argument("-v", "--vocab_path", required=True, type=str)
-    parser.add_argument("-o", "--output_path", required=True, type=str)
-
-    parser.add_argument("-hs", "--hidden", type=int, default=256)
-    parser.add_argument("-l", "--layers", type=int, default=8)
-    parser.add_argument("-a", "--attn_heads", type=int, default=8)
-    parser.add_argument("-s", "--seq_len", type=int, default=20)
-
-    parser.add_argument("-b", "--batch_size", type=int, default=64)
-    parser.add_argument("-e", "--epochs", type=int, default=10)
-    parser.add_argument("-w", "--num_workers", type=int, default=5)
-    parser.add_argument("--with_cuda", type=bool, default=True)
-    parser.add_argument("--log_freq", type=int, default=10)
-    parser.add_argument("--corpus_lines", type=int, default=None)
-
-    parser.add_argument("--lr", type=float, default=1e-3)
-    parser.add_argument("--adam_weight_decay", type=float, default=0.01)
-    parser.add_argument("--adam_beta1", type=float, default=0.9)
-    parser.add_argument("--adam_beta2", type=float, default=0.999)
+    parser.add_argument("-c", "--train_dataset", required=True, type=str, help="train dataset for train bert")
+    parser.add_argument("-t", "--test_dataset", type=str, default=None, help="test set for evaluate train set")
+    parser.add_argument("-v", "--vocab_path", required=True, type=str, help="built vocab model path with bert-vocab")
+    parser.add_argument("-o", "--output_path", required=True, type=str, help="ex)output/bert.model")
+
+    parser.add_argument("-hs", "--hidden", type=int, default=256, help="hidden size of transformer model")
+    parser.add_argument("-l", "--layers", type=int, default=8, help="number of layers")
+    parser.add_argument("-a", "--attn_heads", type=int, default=8, help="number of attention heads")
+    parser.add_argument("-s", "--seq_len", type=int, default=20, help="maximum sequence len")
+
+    parser.add_argument("-b", "--batch_size", type=int, default=64, help="number of batch_size")
+    parser.add_argument("-e", "--epochs", type=int, default=10, help="number of epochs")
+    parser.add_argument("-w", "--num_workers", type=int, default=5, help="dataloader worker size")
+
+    parser.add_argument("--with_cuda", type=bool, default=True, help="training with CUDA: true, or false")
+    parser.add_argument("--log_freq", type=int, default=10, help="printing loss every n iter: setting n")
+    parser.add_argument("--corpus_lines", type=int, default=None, help="total number of lines in corpus")
+    parser.add_argument("--cuda_devices", type=int, nargs='+', default=None, help="CUDA device ids")
+    parser.add_argument("--on_memory", type=bool, default=True, help="Loading on memory: true or false")
+
+    parser.add_argument("--lr", type=float, default=1e-3, help="learning rate of adam")
+    parser.add_argument("--adam_weight_decay", type=float, default=0.01, help="weight_decay of adam")
+    parser.add_argument("--adam_beta1", type=float, default=0.9, help="adam first beta value")
+    parser.add_argument("--adam_beta2", type=float, default=0.999, help="adam first beta value")
 
     args = parser.parse_args()
 
@@ -39,11 +42,12 @@ def train():
     print("Vocab Size: ", len(vocab))
 
     print("Loading Train Dataset", args.train_dataset)
-    train_dataset = BERTDataset(args.train_dataset, vocab, seq_len=args.seq_len, corpus_lines=args.corpus_lines)
+    train_dataset = BERTDataset(args.train_dataset, vocab, seq_len=args.seq_len,
+                                corpus_lines=args.corpus_lines, on_memory=args.on_memory)
 
     print("Loading Test Dataset", args.test_dataset)
-    test_dataset = BERTDataset(args.test_dataset, vocab,
-                               seq_len=args.seq_len) if args.test_dataset is not None else None
+    test_dataset = BERTDataset(args.test_dataset, vocab, seq_len=args.seq_len, on_memory=args.on_memory) \
+        if args.test_dataset is not None else None
 
     print("Creating Dataloader")
     train_data_loader = DataLoader(train_dataset, batch_size=args.batch_size, num_workers=args.num_workers)
@@ -56,7 +60,7 @@ def train():
     print("Creating BERT Trainer")
     trainer = BERTTrainer(bert, len(vocab), train_dataloader=train_data_loader, test_dataloader=test_data_loader,
                           lr=args.lr, betas=(args.adam_beta1, args.adam_beta2), weight_decay=args.adam_weight_decay,
-                          with_cuda=args.with_cuda, log_freq=args.log_freq)
+                          with_cuda=args.with_cuda, cuda_devices=args.cuda_devices, log_freq=args.log_freq)
 
     print("Training Start")
     for epoch in range(args.epochs):

bert_pytorch/dataset/dataset.py

@@ -5,19 +5,37 @@
 
 
 class BERTDataset(Dataset):
-    def __init__(self, corpus_path, vocab, seq_len, encoding="utf-8", corpus_lines=None):
+    def __init__(self, corpus_path, vocab, seq_len, encoding="utf-8", corpus_lines=None, on_memory=True):
         self.vocab = vocab
         self.seq_len = seq_len
 
+        self.on_memory = on_memory
+        self.corpus_lines = corpus_lines
+        self.corpus_path = corpus_path
+        self.encoding = encoding
+
         with open(corpus_path, "r", encoding=encoding) as f:
-            self.datas = [line[:-1].split("\t")
-                          for line in tqdm.tqdm(f, desc="Loading Dataset", total=corpus_lines)]
+            if self.corpus_lines is None and not on_memory:
+                for _ in tqdm.tqdm(f, desc="Loading Dataset", total=corpus_lines):
+                    self.corpus_lines += 1
+
+            if on_memory:
+                self.lines = [line[:-1].split("\t")
+                              for line in tqdm.tqdm(f, desc="Loading Dataset", total=corpus_lines)]
+                self.corpus_lines = len(self.lines)
+
+        if not on_memory:
+            self.file = open(corpus_path, "r", encoding=encoding)
+            self.random_file = open(corpus_path, "r", encoding=encoding)
+
+            for _ in range(random.randint(self.corpus_lines if self.corpus_lines < 1000 else 1000)):
+                self.random_file.__next__()
 
     def __len__(self):
-        return len(self.datas)
+        return self.corpus_lines
 
     def __getitem__(self, item):
-        t1, (t2, is_next_label) = self.datas[item][0], self.random_sent(item)
+        t1, t2, is_next_label = self.random_sent(item)
         t1_random, t1_label = self.random_word(t1)
         t2_random, t2_label = self.random_word(t2)
 
@@ -49,16 +67,18 @@ def random_word(self, sentence):
         for i, token in enumerate(tokens):
             prob = random.random()
             if prob < 0.15:
-                # 80% randomly change token to make token
-                if prob < 0.15 * 0.8:
+                prob /= 0.15
+
+                # 80% randomly change token to mask token
+                if prob < 0.8:
                     tokens[i] = self.vocab.mask_index
 
                 # 10% randomly change token to random token
-                elif prob * 0.8 <= prob < prob * 0.9:
+                elif prob < 0.9:
                     tokens[i] = random.randrange(len(self.vocab))
 
                 # 10% randomly change token to current token
-                elif prob >= prob * 0.9:
+                else:
                     tokens[i] = self.vocab.stoi.get(token, self.vocab.unk_index)
 
                 output_label.append(self.vocab.stoi.get(token, self.vocab.unk_index))
@@ -70,8 +90,36 @@ def random_word(self, sentence):
         return tokens, output_label
 
     def random_sent(self, index):
+        t1, t2 = self.get_corpus_line(index)
+
         # output_text, label(isNotNext:0, isNext:1)
         if random.random() > 0.5:
-            return self.datas[index][1], 1
+            return t1, t2, 1
+        else:
+            return t1, self.get_random_line(), 0
+
+    def get_corpus_line(self, item):
+        if self.on_memory:
+            return self.lines[item][0], self.lines[item][1]
         else:
-            return self.datas[random.randrange(len(self.datas))][1], 0
+            line = self.file.__next__()
+            if line is None:
+                self.file.close()
+                self.file = open(self.corpus_path, "r", encoding=self.encoding)
+                line = self.file.__next__()
+
+            t1, t2 = line[:-1].split("\t")
+            return t1, t2
+
+    def get_random_line(self):
+        if self.on_memory:
+            return self.lines[random.randrange(len(self.lines))][1]
+
+        line = self.file.__next__()
+        if line is None:
+            self.file.close()
+            self.file = open(self.corpus_path, "r", encoding=self.encoding)
+            for _ in range(random.randint(self.corpus_lines if self.corpus_lines < 1000 else 1000)):
+                self.random_file.__next__()
+            line = self.random_file.__next__()
+        return line[:-1].split("\t")[1]

bert_pytorch/model/embedding/position.py

@@ -13,7 +13,7 @@ def __init__(self, d_model, max_len=512):
         pe.require_grad = False
 
         position = torch.arange(0, max_len).float().unsqueeze(1)
-        div_term = (torch.arange(0, d_model, 2) * -(math.log(10000.0) / d_model)).float().exp()
+        div_term = (torch.arange(0, d_model, 2).float() * -(math.log(10000.0) / d_model)).exp()
 
         pe[:, 0::2] = torch.sin(position * div_term)
         pe[:, 1::2] = torch.cos(position * div_term)

bert_pytorch/trainer/pretrain.py

@@ -22,7 +22,7 @@ class BERTTrainer:
     def __init__(self, bert: BERT, vocab_size: int,
                  train_dataloader: DataLoader, test_dataloader: DataLoader = None,
                  lr: float = 1e-4, betas=(0.9, 0.999), weight_decay: float = 0.01,
-                 with_cuda: bool = True, log_freq: int = 10):
+                 with_cuda: bool = True, cuda_devices=None, log_freq: int = 10):
         """
         :param bert: BERT model which you want to train
         :param vocab_size: total word vocab size
@@ -45,9 +45,9 @@ def __init__(self, bert: BERT, vocab_size: int,
         self.model = BERTLM(bert, vocab_size).to(self.device)
 
         # Distributed GPU training if CUDA can detect more than 1 GPU
-        if torch.cuda.device_count() > 1:
+        if with_cuda and torch.cuda.device_count() > 1:
             print("Using %d GPUS for BERT" % torch.cuda.device_count())
-            self.model = nn.DataParallel(self.model)
+            self.model = nn.DataParallel(self.model, device_ids=cuda_devices)
 
         # Setting the train and test data loader
         self.train_data = train_dataloader

requirements.txt

@@ -1,3 +1,3 @@
 tqdm
 numpy
-torch>=0.4.0
+torch>=0.4.0

setup.py

@@ -3,10 +3,10 @@
 import os
 import sys
 
-__version__ = "0.0.1a3"
+__version__ = "0.0.1a4"
 
 with open("requirements.txt") as f:
-    require_packages = [line[:-1] for line in f]
+    require_packages = [line[:-1] if line[-1] == "\n" else line for line in f]
 
 with open("README.md", "r", encoding="utf-8") as f:
     long_description = f.read()