Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in / Register
Toggle navigation
Menu
Open sidebar
ModelZoo
ResNet50_tensorflow
Commits
eef72ed6
Unverified
Commit
eef72ed6
authored
Jun 18, 2018
by
Taylor Robie
Committed by
GitHub
Jun 18, 2018
Browse files
remove unused imports and lint (#4475)
* remove unused imports and lint * fix schedule.py * address PR comments
parent
2310bc34
Changes
2
Show whitespace changes
Inline
Side-by-side
Showing
2 changed files
with
13 additions
and
26 deletions
+13
-26
official/transformer/utils/schedule.py
official/transformer/utils/schedule.py
+0
-4
official/utils/accelerator/tpu.py
official/utils/accelerator/tpu.py
+13
-22
No files found.
official/transformer/utils/schedule.py
View file @
eef72ed6
...
@@ -19,12 +19,9 @@ from __future__ import division
...
@@ -19,12 +19,9 @@ from __future__ import division
from
__future__
import
print_function
from
__future__
import
print_function
import
math
import
math
import
time
import
tensorflow
as
tf
import
tensorflow
as
tf
from
official.transformer.utils
import
dataset
_TRAIN
,
_EVAL
=
tf
.
estimator
.
ModeKeys
.
TRAIN
,
tf
.
estimator
.
ModeKeys
.
EVAL
_TRAIN
,
_EVAL
=
tf
.
estimator
.
ModeKeys
.
TRAIN
,
tf
.
estimator
.
ModeKeys
.
EVAL
...
@@ -123,7 +120,6 @@ class Manager(object):
...
@@ -123,7 +120,6 @@ class Manager(object):
Args:
Args:
num_epochs: An integer of the number of epochs to convert to steps.
num_epochs: An integer of the number of epochs to convert to steps.
batch_size: The mini-batch size used.
mode: The estimator ModeKey of the computation
mode: The estimator ModeKey of the computation
Returns:
Returns:
...
...
official/utils/accelerator/tpu.py
View file @
eef72ed6
...
@@ -14,14 +14,7 @@
...
@@ -14,14 +14,7 @@
# ==============================================================================
# ==============================================================================
"""Functions specific to running TensorFlow on TPUs."""
"""Functions specific to running TensorFlow on TPUs."""
import
time
import
tensorflow
as
tf
import
tensorflow
as
tf
from
tensorflow.python.framework
import
dtypes
from
tensorflow.python.framework
import
ops
from
tensorflow.python.ops
import
array_ops
from
tensorflow.python.ops
import
embedding_ops
from
tensorflow.python.ops
import
math_ops
# "local" is a magic word in the TPU cluster resolver; it informs the resolver
# "local" is a magic word in the TPU cluster resolver; it informs the resolver
...
@@ -84,7 +77,7 @@ def construct_scalar_host_call(metric_dict, model_dir, prefix=""):
...
@@ -84,7 +77,7 @@ def construct_scalar_host_call(metric_dict, model_dir, prefix=""):
return
host_call_fn
,
[
global_step_tensor
]
+
other_tensors
return
host_call_fn
,
[
global_step_tensor
]
+
other_tensors
def
embedding_matmul
(
embedding_table
,
values
,
mask
,
name
=
'
embedding_matmul
'
):
def
embedding_matmul
(
embedding_table
,
values
,
mask
,
name
=
"
embedding_matmul
"
):
"""Performs embedding lookup via a matmul.
"""Performs embedding lookup via a matmul.
The matrix to be multiplied by the embedding table Tensor is constructed
The matrix to be multiplied by the embedding table Tensor is constructed
...
@@ -104,21 +97,19 @@ def embedding_matmul(embedding_table, values, mask, name='embedding_matmul'):
...
@@ -104,21 +97,19 @@ def embedding_matmul(embedding_table, values, mask, name='embedding_matmul'):
Rank 3 tensor of embedding vectors.
Rank 3 tensor of embedding vectors.
"""
"""
with
ops
.
name_scope
(
name
):
with
tf
.
name_scope
(
name
):
n_embeddings
,
embedding_dim
=
embedding_table
.
get_shape
().
as_list
()
n_embeddings
=
embedding_table
.
get_shape
().
as_list
()
[
0
]
batch_size
,
padded_size
=
values
.
shape
.
as_list
()
batch_size
,
padded_size
=
values
.
shape
.
as_list
()
emb_idcs
=
array_ops
.
tile
(
emb_idcs
=
tf
.
tile
(
array_ops
.
reshape
(
values
,
(
batch_size
,
padded_size
,
1
)),
(
1
,
1
,
tf
.
reshape
(
values
,
(
batch_size
,
padded_size
,
1
)),
(
1
,
1
,
n_embeddings
))
n_embeddings
))
emb_weights
=
tf
.
tile
(
emb_weights
=
array_ops
.
tile
(
tf
.
reshape
(
mask
,
(
batch_size
,
padded_size
,
1
)),
(
1
,
1
,
n_embeddings
))
array_ops
.
reshape
(
mask
,
(
batch_size
,
padded_size
,
1
)),
col_idcs
=
tf
.
tile
(
(
1
,
1
,
n_embeddings
))
tf
.
reshape
(
tf
.
range
(
n_embeddings
),
(
1
,
1
,
n_embeddings
)),
col_idcs
=
array_ops
.
tile
(
array_ops
.
reshape
(
math_ops
.
range
(
n_embeddings
),
(
1
,
1
,
n_embeddings
)),
(
batch_size
,
padded_size
,
1
))
(
batch_size
,
padded_size
,
1
))
one_hot
=
array_ops
.
where
(
one_hot
=
tf
.
where
(
math_ops
.
equal
(
emb_idcs
,
col_idcs
),
emb_weights
,
tf
.
equal
(
emb_idcs
,
col_idcs
),
emb_weights
,
array_ops
.
zeros
((
batch_size
,
padded_size
,
n_embeddings
)))
tf
.
zeros
((
batch_size
,
padded_size
,
n_embeddings
)))
return
math_ops
.
tensordot
(
one_hot
,
embedding_table
,
1
)
return
tf
.
tensordot
(
one_hot
,
embedding_table
,
1
)
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment