Fix the issue with C2 cont build
Summary: Issue was introduced in D21258652. We need to make sure it compiles with opt mode. We may still have some left over py2 packages. Let's just use some format work with both.
Test Plan: ci
Reviewed By: xush6528
Differential Revision: D21457394
fbshipit-source-id: cde79a0fc6b4feba307bd9d45e1a1d4a42de9263
diff --git a/caffe2/python/optimizer.py b/caffe2/python/optimizer.py
index c7207f9..94b31f1 100644
--- a/caffe2/python/optimizer.py
+++ b/caffe2/python/optimizer.py
@@ -522,7 +522,7 @@
output_effective_lr_and_update=False,
pruning_options=None, swa_options=None, weight_scale=None, **kwargs):
for k, v in locals().items():
- logger.info(f'AdagradOptimizer: input arguments: {k}: {v}')
+ logger.info('AdagradOptimizer: input arguments: {}: {}'.format(k, v))
super(AdagradOptimizer, self).__init__()
self.alpha = alpha
@@ -667,7 +667,7 @@
)
if self.engine in FP16_ENGINES:
- assert self.weight_decay == 0, f'weight decay is not tested for engine: {self.engine}'
+ assert self.weight_decay == 0, 'weight decay is not tested for engine: {}'.format(self.engine)
shapes, types = workspace.InferShapesAndTypes([param_init_net])
assert str(param) in shapes, shapes
@@ -743,16 +743,20 @@
weight_decay = 0.
if isinstance(grad, core.GradientSlice):
if len(param_shape) == 1:
- logger.warn(f"APPLYING weight decay on 1d sparse param: {str(param)}.shape is {param_shape}")
+ logger.warn("APPLYING weight decay on 1d sparse param: {}.shape is {}".format(
+ str(param), param_shape))
weight_decay = self.weight_decay
else:
# Skip weight decay for 1d parameters
if len(param_shape) == 1:
weight_decay = 0.
- logger.warn(f"SKIPPING weight decay on 1d dense param: {str(param)}.shape is {param_shape}")
+ logger.warn("SKIPPING weight decay on 1d dense param: {}.shape is {}".format(
+ str(param), param_shape))
+
else:
weight_decay = self.weight_decay
- logger.info(f"weight_decay for {str(param)} (shape:{param_shape}): {weight_decay}")
+ logger.info("weight_decay for {} (shape:{}): {}".format(
+ str(param), param_shape, weight_decay))
if isinstance(grad, core.GradientSlice):
assert self.decay == 1.,\
@@ -764,18 +768,18 @@
if self.rowWise:
if self.use_mask is True:
op = 'MaskedRowWiseSparseAdagrad'
- assert weight_decay == 0, f'weight decay is not implemented for {op} yet'
+ assert weight_decay == 0, 'weight decay is not implemented for {} yet'.format(op)
input_args += [mask_blob, mask_changed_blob]
else:
op = 'RowWiseSparseAdagrad'
else:
if self.use_mask is True:
op = 'MaskedSparseAdagrad'
- assert weight_decay == 0, f'weight decay is not implemented for {op} yet'
+ assert weight_decay == 0, 'weight decay is not implemented for {} yet'.format(op)
input_args += [mask_blob, mask_changed_blob]
else:
op = 'SparseAdagrad'
- logger.info(f"using {op} for {str(param)}")
+ logger.info("using {} for {}".format(op, str(param)))
if self.prune_delays:
input_args += [lr_iteration, last_mask_updated_iter]
@@ -818,7 +822,7 @@
output_args += [mask_blob, last_mask_updated_iter]
if self.use_mask:
- assert weight_decay == 0, f'weight decay is not implemented for use_mask yet'
+ assert weight_decay == 0, 'weight decay is not implemented for use_mask yet'
net.MaskedAdagrad(
input_args,
output_args,
