Commits · fa025551dffc1dda23cbeb8ccbde0a0c0c8743bd · Shashank Suhas / seminar-breakout

21 Feb, 2017 6 commits
- refactor predictors. · fa025551
  Yuxin Wu authored Feb 22, 2017
  
  fa025551
- tutorial update · dbfa9982
  Yuxin Wu authored Feb 22, 2017
  
  dbfa9982
- code clean-up in predict/ · fe2b4f97
  Yuxin Wu authored Feb 21, 2017
  
  fe2b4f97
- some change in gradproc, and fix #158 · 8df83a93
  Yuxin Wu authored Feb 21, 2017
  
  8df83a93
- some docs change · 10f55570
  Yuxin Wu authored Feb 21, 2017
  
  10f55570
- docs update · 74e3eeef
  Yuxin Wu authored Feb 19, 2017
  
  74e3eeef
20 Feb, 2017 14 commits
- remove some typos from the documentation (#155) · a9958037
  Patrick Wieschollek authored Feb 20, 2017
```
* remove some typos from the documentation

* add missing updates

* revert datapoint spelling

* renaming variable

`grep -rn . -e "max_pertub_pixel" --include=*.py` shows no
dependencies in examples.
```
  a9958037
- make MultiTowerPredictorTrainer a default + some doc change. (#156) · 35833026
  Yuxin Wu authored Feb 20, 2017
  
  35833026
- just docs · 2b076314
  Yuxin Wu authored Feb 20, 2017
  
  2b076314
- docs update · a2934281
  Yuxin Wu authored Feb 20, 2017
  
  a2934281
- move summary_op from trainer to callbacks. fix #125 · a71ff4d7
  Yuxin Wu authored Feb 20, 2017
  
  a71ff4d7
- add callback <-> hook adapter. fix #147 · 658529d5
  Yuxin Wu authored Feb 20, 2017
  
  658529d5
- docs of callbacks (#147) · 4c926eb7
  Yuxin Wu authored Feb 20, 2017
  
  4c926eb7
- Merge branch 'dev' · f3644ce9
  Yuxin Wu authored Feb 20, 2017
  
  f3644ce9
- Merge branch 'master' into dev · 0f2eaeea
  Yuxin Wu authored Feb 20, 2017
  
  0f2eaeea
- some clean-ups, and add an alias `hooked_sess` (#147) · a5652699
  Yuxin Wu authored Feb 20, 2017
  
  a5652699
- rename extra_fetch to before_run (#147) · ccf4a5a0
  Yuxin Wu authored Feb 20, 2017
  
  ccf4a5a0
- split trigger_step and after_run (#147) · 39d08d47
  Yuxin Wu authored Feb 20, 2017
  
  39d08d47
- distinguish between sess.run call and run_step call. fix WGAN examples. (#147) · f80843dc
  Yuxin Wu authored Feb 20, 2017
  
  f80843dc
- fix Deconv. fix #154 · 05f7ba8f
  Yuxin Wu authored Feb 19, 2017
  
  05f7ba8f
19 Feb, 2017 4 commits
- use hooks to run step triggers. examples unfixed. (#147) · eee05770
  Yuxin Wu authored Feb 20, 2017
  
  eee05770
- bug fix in DQN.py · 136174c9
  Yuxin Wu authored Feb 19, 2017
  
  136174c9
- update DQN code · 6452b444
  Yuxin Wu authored Feb 19, 2017
  
  6452b444
- speedup DQN. · 0870401c
  Yuxin Wu authored Feb 19, 2017
  
  0870401c
18 Feb, 2017 6 commits
- update readme · cab0c4f3
  Yuxin Wu authored Feb 19, 2017
  
  cab0c4f3
- a performance claim · 1b8cd1a4
  Yuxin Wu authored Feb 19, 2017
  
  1b8cd1a4
- support NCHW in deconv. change some more examples. (#150) · 08cecd44
  Yuxin Wu authored Feb 18, 2017
  
  08cecd44
- add NCHW for Conv2D,Pooling and BN. fix #150 · 880b767d
  Yuxin Wu authored Feb 18, 2017
  
  880b767d
- simplify code of two saverrestore · 4be41492
  Yuxin Wu authored Feb 18, 2017
  
  4be41492
- EMA callback don't create variables itself. add old SaverRestore to be fast · cc89b105
  Yuxin Wu authored Feb 18, 2017
  
  cc89b105
17 Feb, 2017 2 commits
- simplify ModeSaver. just save without rename · 20d1af11
  Yuxin Wu authored Feb 17, 2017
  
  20d1af11
- use tf.train.MonitoredSession. enforce a boundary between graph finalize & session create · 3e97f126
  Yuxin Wu authored Feb 17, 2017
  
  3e97f126
16 Feb, 2017 4 commits
- add infogan models, and some small changes. · 224025e3
  Yuxin Wu authored Feb 17, 2017
  
  224025e3
- trainer.get_predictor support tower-tensor as input · 55f2f5da
  Yuxin Wu authored Feb 17, 2017
  
  55f2f5da
- support old checkpoint format. · 6640f9bb
  Yuxin Wu authored Feb 16, 2017
  
  6640f9bb
- move expreplay out of RL. · b6df5567
  Yuxin Wu authored Feb 16, 2017
  
  b6df5567
15 Feb, 2017 4 commits
- simplify WGAN. fix bug in viz. · 63004976
  Yuxin Wu authored Feb 16, 2017
  
  63004976
- add a WGAN example · 13c96b94
  Yuxin Wu authored Feb 15, 2017
  
  13c96b94
- refactor in viz. stack_patches now handles nonuniform patches as well. · 00fdb263
  Yuxin Wu authored Feb 15, 2017
  
  00fdb263
- an optimizer which accumulates gradients. fix #141 · d6a89f0b
  Yuxin Wu authored Feb 15, 2017
  
  d6a89f0b