Commit f94f1c60 authored by Lysandre's avatar Lysandre
Browse files

Distributed training + tokenizer agnostic mask token

parent 5652f54a
...@@ -39,12 +39,10 @@ from pytorch_transformers import (WEIGHTS_NAME, GPT2Config, GPT2LMHeadModel, GPT ...@@ -39,12 +39,10 @@ from pytorch_transformers import (WEIGHTS_NAME, GPT2Config, GPT2LMHeadModel, GPT
BertConfig, BertForMaskedLM, BertTokenizer, BERT_PRETRAINED_MODEL_ARCHIVE_MAP, BertConfig, BertForMaskedLM, BertTokenizer, BERT_PRETRAINED_MODEL_ARCHIVE_MAP,
RobertaConfig, RobertaForMaskedLM, RobertaTokenizer, ROBERTA_PRETRAINED_MODEL_ARCHIVE_MAP) RobertaConfig, RobertaForMaskedLM, RobertaTokenizer, ROBERTA_PRETRAINED_MODEL_ARCHIVE_MAP)
from pytorch_transformers import AdamW, WarmupLinearSchedule from pytorch_transformers import AdamW, WarmupLinearSchedule
logger = logging.getLogger(__name__)
from utils_lm import WikiTextDataset from utils_lm import WikiTextDataset
logger = logging.getLogger(__name__)
ALL_MODELS = sum((tuple(conf.pretrained_config_archive_map.keys()) for conf in (GPT2Config,)), ())
MODEL_CLASSES = { MODEL_CLASSES = {
'gpt2': (GPT2Config, GPT2LMHeadModel, GPT2Tokenizer), 'gpt2': (GPT2Config, GPT2LMHeadModel, GPT2Tokenizer),
...@@ -68,10 +66,7 @@ def mask_tokens(inputs, tokenizer, args): ...@@ -68,10 +66,7 @@ def mask_tokens(inputs, tokenizer, args):
labels[~masked_indices.bool()] = -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 indices_replaced = torch.bernoulli(torch.full(labels.shape, 0.8)).byte() & masked_indices
if args.model_name == "bert": inputs[indices_replaced.bool()] = tokenizer.convert_tokens_to_ids(tokenizer.mask_token) # 80% of the time, replace masked input tokens with [MASK]
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() 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) random_words = torch.randint(args.num_embeddings, labels.shape, dtype=torch.long)
inputs[indices_random] = random_words[ inputs[indices_random] = random_words[
...@@ -246,10 +241,7 @@ def evaluate(args, model, tokenizer, prefix=""): ...@@ -246,10 +241,7 @@ def evaluate(args, model, tokenizer, prefix=""):
def load_and_cache_examples(args, tokenizer, evaluate=False): def load_and_cache_examples(args, tokenizer, evaluate=False):
if args.local_rank not in [-1, 0]: dataset = WikiTextDataset(args, tokenizer, file="test" if evaluate else "train", directory=args.data_dir)
torch.distributed.barrier() # Make sure only the first process in distributed training process the dataset, and the others will use the cache
dataset = WikiTextDataset(tokenizer, file="test" if evaluate else "train", directory=args.data_dir)
return dataset return dataset
......
...@@ -3,10 +3,27 @@ import os ...@@ -3,10 +3,27 @@ import os
import random import random
import torch import torch
import torch.nn.functional as F import torch.nn.functional as F
import logging
import pickle
logger = logging.getLogger(__name__)
class WikiTextDataset(Dataset): class WikiTextDataset(Dataset):
def __init__(self, tokenizer, file='train', directory='wikitext', max_context_length=1024): def __init__(self, args, tokenizer, file='train', directory='wikitext', max_context_length=512, cache=None):
if args.local_rank not in [-1, 0]:
torch.distributed.barrier() # Make sure only the first process in distributed training process the dataset, and the others will use the cache
cached_features_file = os.path.join(args.data_dir, f'cached_lm_{file}_{args.max_seq_length}')
if os.path.exists(cached_features_file):
logger.info("Loading features from cached file %s", cached_features_file)
with open(cached_features_file, 'rb') as handle:
self.examples = pickle.load(handle)
else:
logger.info("Creating features from dataset file at %s", args.data_dir)
self.max_context_length = max_context_length self.max_context_length = max_context_length
self.examples = [] self.examples = []
...@@ -18,6 +35,14 @@ class WikiTextDataset(Dataset): ...@@ -18,6 +35,14 @@ class WikiTextDataset(Dataset):
while len(tokenized_text) > max_context_length: while len(tokenized_text) > max_context_length:
self.examples.append(tokenized_text[:max_context_length]) self.examples.append(tokenized_text[:max_context_length])
tokenized_text = tokenized_text[max_context_length:] tokenized_text = tokenized_text[max_context_length:]
if args.local_rank in [-1, 0]:
logger.info("Saving features into cached file %s", cached_features_file)
with open(cached_features_file, 'wb') as handle:
pickle.dump(self.examples, handle, protocol=pickle.HIGHEST_PROTOCOL)
if args.local_rank == 0:
torch.distributed.barrier()
def __len__(self): def __len__(self):
return len(self.examples) return len(self.examples)
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment