BERT + RoBERTa masking tokens handling + GPU device update.

examples/run_generative_finetuning.py

@@ -65,11 +65,15 @@ def set_seed(args):
 def mask_tokens(inputs, tokenizer, args):
     labels = inputs.clone()
     masked_indices = torch.bernoulli(torch.full(labels.shape, args.mlm_probability)).byte()
-    labels[~masked_indices] = -1  # We only compute loss on masked tokens
+    labels[~masked_indices.bool()] = -1  # We only compute loss on masked tokens
     indices_replaced = torch.bernoulli(torch.full(labels.shape, 0.8)).byte() & masked_indices
-    inputs[indices_replaced] = tokenizer.vocab["[MASK]"]  # 80% of the time, replace masked input tokens with [MASK]
-    indices_random = torch.bernoulli(torch.full(labels.shape, 0.5)).byte() & masked_indices & ~indices_replaced
-    random_words = torch.randint(args.num_embeddings, labels.shape, dtype=torch.long, device=args.device)
+
+    if args.model_name == "bert":
+        inputs[indices_replaced.bool()] = tokenizer.vocab["[MASK]"]  # 80% of the time, replace masked input tokens with [MASK]
+    elif args.model_name == "roberta":
+        inputs[indices_replaced.bool()] = tokenizer.encoder["<mask>"]  # 80% of the time, replace masked input tokens with <mask>
+    indices_random = (torch.bernoulli(torch.full(labels.shape, 0.5)).byte() & masked_indices & ~indices_replaced).bool()
+    random_words = torch.randint(args.num_embeddings, labels.shape, dtype=torch.long)
     inputs[indices_random] = random_words[
         indices_random]  # 10% of the time, replace masked input tokens with random word
     return inputs, labels
@@ -132,14 +136,15 @@ def train(args, train_dataset, model, tokenizer):
     for _ in train_iterator:
         epoch_iterator = tqdm(train_dataloader, desc="Iteration", disable=args.local_rank not in [-1, 0])
         for step, batch in enumerate(epoch_iterator):
-            batch.to(args.device)
-            model.train()
             inputs, labels = mask_tokens(batch, tokenizer, args) if args.mlm else (batch, batch)
+            inputs = inputs.to(args.device)
+            labels = labels.to(args.device)
+            model.train()
             outputs = model(inputs, masked_lm_labels=labels) if args.mlm else model(inputs, labels=labels)
             loss = outputs[0]  # model outputs are always tuple in pytorch-transformers (see doc)
 
             if args.n_gpu > 1:
-                loss = loss.mean() # mean() to average on multi-gpu parallel training
+                loss = loss.mean()  # mean() to average on multi-gpu parallel training
             if args.gradient_accumulation_steps > 1:
                 loss = loss / args.gradient_accumulation_steps
 
@@ -214,7 +219,7 @@ def evaluate(args, model, tokenizer, prefix=""):
     nb_eval_steps = 0
     for batch in tqdm(eval_dataloader, desc="Evaluating"):
         model.eval()
-        batch.to(args.device)
+        batch = batch.to(args.device)
 
         with torch.no_grad():
             outputs = model(batch)
@@ -285,9 +290,9 @@ def main():
     parser.add_argument("--do_lower_case", action='store_true',
                         help="Set this flag if you are using an uncased model.")
 
-    parser.add_argument("--per_gpu_train_batch_size", default=8, type=int,
+    parser.add_argument("--per_gpu_train_batch_size", default=4, type=int,
                         help="Batch size per GPU/CPU for training.")
-    parser.add_argument("--per_gpu_eval_batch_size", default=8, type=int,
+    parser.add_argument("--per_gpu_eval_batch_size", default=4, type=int,
                         help="Batch size per GPU/CPU for evaluation.")
     parser.add_argument('--gradient_accumulation_steps', type=int, default=1,
                         help="Number of updates steps to accumulate before performing a backward/update pass.")
@@ -299,7 +304,7 @@ def main():
                         help="Epsilon for Adam optimizer.")
     parser.add_argument("--max_grad_norm", default=1.0, type=float,
                         help="Max gradient norm.")
-    parser.add_argument("--num_train_epochs", default=3.0, type=float,
+    parser.add_argument("--num_train_epochs", default=1.0, type=float,
                         help="Total number of training epochs to perform.")
     parser.add_argument("--max_steps", default=-1, type=int,
                         help="If > 0: set total number of training steps to perform. Override num_train_epochs.")

examples/utils_lm.py

@@ -6,8 +6,7 @@ import torch.nn.functional as F
 
 
 class WikiTextDataset(Dataset):
-	def __init__(self, tokenizer, file='train', directory='wikitext', max_context_length=512, device='cpu'):
-		self.device = device
+	def __init__(self, tokenizer, file='train', directory='wikitext', max_context_length=512):
 		self.max_context_length = max_context_length
 
 		self.examples = []
@@ -32,7 +31,7 @@ class WikiTextDataset(Dataset):
 		return len(self.examples)
 
 	def __getitem__(self, item):
-		return torch.tensor(self.examples[item], device=self.device)
+		return torch.tensor(self.examples[item])
 
 	@staticmethod
 	def collate(values):