Skip to content
Projects
Groups
Snippets
Help
Loading...
Help
Support
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in
Toggle navigation
S
seminar-breakout
Project overview
Project overview
Details
Activity
Releases
Repository
Repository
Files
Commits
Branches
Tags
Contributors
Graph
Compare
Issues
0
Issues
0
List
Boards
Labels
Milestones
Merge Requests
0
Merge Requests
0
CI / CD
CI / CD
Pipelines
Jobs
Schedules
Analytics
Analytics
CI / CD
Repository
Value Stream
Wiki
Wiki
Members
Members
Collapse sidebar
Close sidebar
Activity
Graph
Create a new issue
Jobs
Commits
Issue Boards
Open sidebar
Shashank Suhas
seminar-breakout
Commits
dc709e942bd2b2da874e12033d5b2306a98d9572
Switch branch/tag
seminar-breakout
25 Oct, 2017
3 commits
update some examples
· dc709e94
Yuxin Wu
authored
Oct 25, 2017
dc709e94
fix new linting rules
· 0c65c338
Yuxin Wu
authored
Oct 25, 2017
0c65c338
docs update & StagingInputWrapper -> StagingInput
· 9ecf87de
Yuxin Wu
authored
Oct 25, 2017
9ecf87de
21 Oct, 2017
6 commits
Merge branch 'master' into model-redesign
· 73c66c18
Yuxin Wu
authored
Oct 21, 2017
73c66c18
[Trainerv2] Swap trainer directory. change two examples.
· ba4e3178
Yuxin Wu
authored
Oct 21, 2017
ba4e3178
Checkpoint before migrating examples
· 9268bc8c
Yuxin Wu
authored
Oct 21, 2017
9268bc8c
make old GANs work with new trainer
· 0be707e9
Yuxin Wu
authored
Oct 20, 2017
0be707e9
[Trainerv2] use v2 inference interface in v1 trainer
· 9a711e72
Yuxin Wu
authored
Oct 20, 2017
9a711e72
Allow TowerContext to use both ns_name and vs_name
· e121701a
Yuxin Wu
authored
Oct 20, 2017
e121701a
20 Oct, 2017
6 commits
Fix session creator bug for distributed trainer.
· 0addcdc6
Yuxin Wu
authored
Oct 20, 2017
0addcdc6
Allow sesscreate to finalize the graph.
· c99d013a
Yuxin Wu
authored
Oct 20, 2017
c99d013a
[Trainerv2] Add get_predictor support
· 93a177bf
Yuxin Wu
authored
Oct 20, 2017
93a177bf
[Trainerv2] Add TowerTrainer on top of singlecost
· e0b13533
Yuxin Wu
authored
Oct 19, 2017
e0b13533
clean-up '-summary' in stats name. clean-up some use of get_scalar_var
· 0f90d4c2
Yuxin Wu
authored
Oct 19, 2017
0f90d4c2
use sess.make_callable for predictors
· a55d81ca
Yuxin Wu
authored
Oct 19, 2017
a55d81ca
19 Oct, 2017
11 commits
Call predictor with positional arguments
· f363d2e8
Yuxin Wu
authored
Oct 19, 2017
f363d2e8
fix import
· a6a2aba4
Yuxin Wu
authored
Oct 19, 2017
a6a2aba4
Move gradient average utilities
· f409fbf0
Yuxin Wu
authored
Oct 19, 2017
f409fbf0
Create the old-style trainer when requested
· d53b5c4c
Yuxin Wu
authored
Oct 19, 2017
d53b5c4c
Add a function to apply general prefetch policies
· 4e644290
Yuxin Wu
authored
Oct 19, 2017
4e644290
Use `logger.get_logger_dir()` to access logger directory.
· fff3f2d3
Yuxin Wu
authored
Oct 19, 2017
fff3f2d3
fix bug of last commit
· 1be49dc9
Yuxin Wu
authored
Oct 19, 2017
1be49dc9
Builder with get_grads instead of input+get_cost
· d2939cf8
Yuxin Wu
authored
Oct 19, 2017
d2939cf8
Allow double-wrapping TowerFuncWrapper
· 2c5b1bec
Yuxin Wu
authored
Oct 18, 2017
2c5b1bec
deprecate TrainConfig(predict_tower). The device should be passed directly to InferenceRunner
· d9817a56
Yuxin Wu
authored
Oct 18, 2017
d9817a56
[Trainerv2] Let InferenceRunner run with new Trainer
· da98e447
Yuxin Wu
authored
Oct 18, 2017
da98e447
18 Oct, 2017
14 commits
a better TowerTensorHandle to access tensors (currently for predictor only) (#318)
· e5ff50e7
Yuxin Wu
authored
Oct 18, 2017
e5ff50e7
bugfix of last commit.
· 9911b234
Yuxin Wu
authored
Oct 18, 2017
9911b234
Fix imports; Add call_only_once decorator;
· 260936ff
Yuxin Wu
authored
Oct 18, 2017
260936ff
remove DataParallelFeedInput, as dataparallel inference now relies on QueueInput
· 9fff46d5
Yuxin Wu
authored
Oct 18, 2017
9fff46d5
[Trainerv2] fix interface for distributed trainer.
· 12cd6e6c
Yuxin Wu
authored
Oct 18, 2017
12cd6e6c
initial commit of new trainer interface (#318)
· a673974c
Yuxin Wu
authored
Oct 18, 2017
a673974c
Share documents between builder & trainer
· 82187086
Yuxin Wu
authored
Oct 18, 2017
82187086
bump versions
· 5b8ed8be
Yuxin Wu
authored
Oct 18, 2017
5b8ed8be
StagingInputWrapper takes a list of int
· e6493857
Yuxin Wu
authored
Oct 18, 2017
e6493857
Fix some use of loop attributes. These attributes should be available only from before_train
· 4c5cdf9b
Yuxin Wu
authored
Oct 18, 2017
4c5cdf9b
make _monitors private as well.
· 4126a583
Yuxin Wu
authored
Oct 17, 2017
4126a583
fix __doc__ assignment in Py3
· e10d43e7
Yuxin Wu
authored
Oct 17, 2017
e10d43e7
update docs
· c33a3ccb
Yuxin Wu
authored
Oct 17, 2017
c33a3ccb
fix missing import
· caf9ee8d
Yuxin Wu
authored
Oct 17, 2017
caf9ee8d
