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

added warning

Browse Source
This commit is contained in:
lukovnikov
2019-02-27 17:11:42 +01:00
parent 4d79e0d386
commit 35410da758
2 changed files with 6 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
pytorch_pretrained_bert/optimization.py
View File

@@ -159,8 +159,8 @@ class BertAdam(Optimizer):
# warning for exceeding t_total (only active with warmup_linear
if group['schedule'] == "warmup_linear" and progress > 1. and not warned_for_t_total:
logger.warning(
"Training beyond specified 't_total' steps. Learning rate set to {}. "
"Please set 't_total' of {} correctly.".format(lr_scheduled, self.__class__.__name__))
"Training beyond specified 't_total' steps with schedule '{}'. Learning rate set to {}. "
"Please set 't_total' of {} correctly.".format(group['schedule'], lr_scheduled, self.__class__.__name__))
warned_for_t_total = True
# end warning
else:

8
pytorch_pretrained_bert/optimization_openai.py
View File

@@ -29,14 +29,14 @@ def warmup_cosine(x, warmup=0.002):
return 0.5 * (1.0 + torch.cos(math.pi * x))
def warmup_constant(x, warmup=0.002):
""" Linearly increases learning rate over `warmup`*`t_total` (as provided to BertAdam) training steps.
""" Linearly increases learning rate over `warmup`*`t_total` (as provided to OpenAIAdam) training steps.
Learning rate is 1. afterwards. """
if x < warmup:
return x/warmup
return 1.0
def warmup_linear(x, warmup=0.002):
""" Specifies a triangular learning rate schedule where peak is reached at `warmup`*`t_total`-th (as provided to BertAdam) training step.
""" Specifies a triangular learning rate schedule where peak is reached at `warmup`*`t_total`-th (as provided to OpenAIAdam) training step.
After `t_total`-th training step, learning rate is zero. """
if x < warmup:
return x/warmup
@@ -142,8 +142,8 @@ class OpenAIAdam(Optimizer):
# warning for exceeding t_total (only active with warmup_linear
if group['schedule'] == "warmup_linear" and progress > 1. and not warned_for_t_total:
logger.warning(
"Training beyond specified 't_total' steps. Learning rate set to {}. "
"Please set 't_total' of {} correctly.".format(lr_scheduled, self.__class__.__name__))
"Training beyond specified 't_total' steps with schedule '{}'. Learning rate set to {}. "
"Please set 't_total' of {} correctly.".format(group['schedule'], lr_scheduled, self.__class__.__name__))
warned_for_t_total = True
# end warning
else: