SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

Merge branch 'master' into python_2

Browse Source
This commit is contained in:
Thomas Wolf
2019-02-06 00:13:20 +01:00
committed by GitHub
parent 448937c00d 822915142b
commit 848aae49e1
8 changed files with 271 additions and 175 deletions
Download Patch File Download Diff File Expand all files Collapse all files

34
examples/run_classifier.py
View File

@@ -34,8 +34,8 @@ from tqdm import tqdm, trange
from pytorch_pretrained_bert.file_utils import PYTORCH_PRETRAINED_BERT_CACHE
from pytorch_pretrained_bert.modeling import BertForSequenceClassification
from pytorch_pretrained_bert.optimization import BertAdam
from pytorch_pretrained_bert.tokenization import BertTokenizer
from pytorch_pretrained_bert.optimization import BertAdam, warmup_linear
logging.basicConfig(format = '%(asctime)s - %(levelname)s - %(name)s - %(message)s',
datefmt = '%m/%d/%Y %H:%M:%S',
@@ -299,11 +299,6 @@ def accuracy(out, labels):
outputs = np.argmax(out, axis=1)
return np.sum(outputs == labels)
def warmup_linear(x, warmup=0.002):
if x < warmup:
return x/warmup
return 1.0 - x
def main():
parser = argparse.ArgumentParser()
@@ -419,7 +414,7 @@ def main():
raise ValueError("Invalid gradient_accumulation_steps parameter: {}, should be >= 1".format(
args.gradient_accumulation_steps))
args.train_batch_size = int(args.train_batch_size / args.gradient_accumulation_steps)
args.train_batch_size = args.train_batch_size // args.gradient_accumulation_steps
random.seed(args.seed)
np.random.seed(args.seed)
@@ -447,11 +442,13 @@ def main():
tokenizer = BertTokenizer.from_pretrained(args.bert_model, do_lower_case=args.do_lower_case)
train_examples = None
num_train_steps = None
num_train_optimization_steps = None
if args.do_train:
train_examples = processor.get_train_examples(args.data_dir)
num_train_steps = int(
len(train_examples) / args.train_batch_size / args.gradient_accumulation_steps * args.num_train_epochs)
num_train_optimization_steps = int(
len(train_examples) / args.train_batch_size / args.gradient_accumulation_steps) * args.num_train_epochs
if args.local_rank != -1:
num_train_optimization_steps = num_train_optimization_steps // torch.distributed.get_world_size()
# Prepare model
model = BertForSequenceClassification.from_pretrained(args.bert_model,
@@ -477,9 +474,6 @@ def main():
{'params': [p for n, p in param_optimizer if not any(nd in n for nd in no_decay)], 'weight_decay': 0.01},
{'params': [p for n, p in param_optimizer if any(nd in n for nd in no_decay)], 'weight_decay': 0.0}
]
t_total = num_train_steps
if args.local_rank != -1:
t_total = t_total // torch.distributed.get_world_size()
if args.fp16:
try:
from apex.optimizers import FP16_Optimizer
@@ -500,7 +494,7 @@ def main():
optimizer = BertAdam(optimizer_grouped_parameters,
lr=args.learning_rate,
warmup=args.warmup_proportion,
t_total=t_total)
t_total=num_train_optimization_steps)
global_step = 0
nb_tr_steps = 0
@@ -511,7 +505,7 @@ def main():
logger.info("***** Running training *****")
logger.info(" Num examples = %d", len(train_examples))
logger.info(" Batch size = %d", args.train_batch_size)
logger.info(" Num steps = %d", num_train_steps)
logger.info(" Num steps = %d", num_train_optimization_steps)
all_input_ids = torch.tensor([f.input_ids for f in train_features], dtype=torch.long)
all_input_mask = torch.tensor([f.input_mask for f in train_features], dtype=torch.long)
all_segment_ids = torch.tensor([f.segment_ids for f in train_features], dtype=torch.long)
@@ -545,10 +539,12 @@ def main():
nb_tr_examples += input_ids.size(0)
nb_tr_steps += 1
if (step + 1) % args.gradient_accumulation_steps == 0:
# modify learning rate with special warm up BERT uses
lr_this_step = args.learning_rate * warmup_linear(global_step/t_total, args.warmup_proportion)
for param_group in optimizer.param_groups:
param_group['lr'] = lr_this_step
if args.fp16:
# modify learning rate with special warm up BERT uses
# if args.fp16 is False, BertAdam is used that handles this automatically
lr_this_step = args.learning_rate * warmup_linear(global_step/num_train_optimization_steps, args.warmup_proportion)
for param_group in optimizer.param_groups:
param_group['lr'] = lr_this_step
optimizer.step()
optimizer.zero_grad()
global_step += 1