Commits · 8df2d7c599822c4eb246d217553310ab40bd3596 · Shashank Suhas / seminar-breakout

04 Aug, 2017 2 commits
- use ps_device=cpu when #gpu>=4 · 8df2d7c5
  Yuxin Wu authored Aug 04, 2017
  
  8df2d7c5
- Let InputSource.setup() returns the callbacks, to simplify trainer implementations · 8419ee3f
  Yuxin Wu authored Aug 03, 2017
  
  8419ee3f
03 Aug, 2017 8 commits
- [zmq op] get rid of protobuf dependency (#221) · 398cb933
  Yuxin Wu authored Aug 03, 2017
  
  398cb933
- fix DummyConstantInput · d899c925
  Yuxin Wu authored Aug 03, 2017
  
  d899c925
- Check ModelDescBase instead of ModelDesc in PredictConfig (fix #361) · 8059ee40
  Yuxin Wu authored Aug 03, 2017
  
  8059ee40
- typos & clean-ups (including #358) · 930af0b6
  Yuxin Wu authored Aug 03, 2017
  
  930af0b6
- fix BatchRenorm (fix #360) · ebf1d570
  Yuxin Wu authored Aug 03, 2017
  
  ebf1d570
- Fix missing global. Update git hooks. · 8b487b90
  Yuxin Wu authored Aug 02, 2017
  
  8b487b90
- Let get_predictor take GPU id, and remove get_predictors · 875f4d7d
  Yuxin Wu authored Aug 02, 2017
  
  875f4d7d
- comments for common trainers · cbb26847
  Yuxin Wu authored Aug 02, 2017
  
  cbb26847
02 Aug, 2017 8 commits
- Use default session config in predict · 92ee69dc
  Yuxin Wu authored Aug 02, 2017
  
  92ee69dc
- fix rgb/bgr in ResNet script · 149ad4eb
  Yuxin Wu authored Aug 02, 2017
  
  149ad4eb
- fix build · 02a5d258
  Yuxin Wu authored Aug 02, 2017
  
  02a5d258
- Support npz format · 300d840a
  Yuxin Wu authored Aug 02, 2017
  
  300d840a
- update trainer doc (#353, #359) · ef8d4e49
  Yuxin Wu authored Aug 02, 2017
  
  ef8d4e49
- update documentation · f85c3003
  Yuxin Wu authored Aug 02, 2017
  
  f85c3003
- update trainer doc · b785bf77
  Yuxin Wu authored Aug 02, 2017
  
  b785bf77
- Print something for #218 · 2d856f94
  Yuxin Wu authored Aug 01, 2017
  
  2d856f94
01 Aug, 2017 8 commits
- fix build.. · 7ea758cd
  Yuxin Wu authored Jul 31, 2017
  
  7ea758cd
- fix build · e9326105
  Yuxin Wu authored Jul 31, 2017
  
  e9326105
- check cv2 cuda support during import · 7ae014f1
  Yuxin Wu authored Jul 31, 2017
  
  7ae014f1
- hide deprecated methods from documentation · 1bf2737f
  Yuxin Wu authored Jul 31, 2017
  
  1bf2737f
- update docs · a9cba3c6
  Yuxin Wu authored Jul 31, 2017
  
  a9cba3c6
- docs and GAN trainer upgrade · fde338ea
  Yuxin Wu authored Jul 31, 2017
  
  fde338ea
- minor changes in examples and TrainConfig · 353bb6c5
  Yuxin Wu authored Jul 31, 2017
  
  353bb6c5
- improve documents · 82afb459
  Yuxin Wu authored Jul 31, 2017
  
  82afb459
31 Jul, 2017 11 commits
- update rtd to use py3.5 · 01c25a9f
  Yuxin Wu authored Jul 31, 2017
  
  01c25a9f
- Add 'name' to layer signature in documentation · dc448bda
  Yuxin Wu authored Jul 31, 2017
  
  dc448bda
- allow empty model & data in config · 0b226adb
  Yuxin Wu authored Jul 31, 2017
  
  0b226adb
- check config.data/config.model in trainers · e791b9a5
  Yuxin Wu authored Jul 31, 2017
  
  e791b9a5
- update docs · 784e2b7b
  Yuxin Wu authored Jul 31, 2017
  
  784e2b7b
- bump tensorpack version · 5d0d6d16
  Yuxin Wu authored Jul 31, 2017
  
  5d0d6d16
- expose setup_graph in all multigpu trainers; TowerContext(use_vs=True) instead of passing a vs_name · 26b4ea44
  Yuxin Wu authored Jul 31, 2017
  
  26b4ea44
- fix SimpleTrainer. expose setup_graph for AsyncMultiGPUTrainer · 06bb5142
  Yuxin Wu authored Jul 31, 2017
  
  06bb5142
- Expose SimpleTrainer.build_graph; Provide a default Trainer.run_step · 8837d748
  Yuxin Wu authored Jul 31, 2017
  
  8837d748
- cleaner scope logic in TowerContext · 064ea7c7
  Yuxin Wu authored Jul 30, 2017
  
  064ea7c7
- fix resnet · 844d8e69
  Yuxin Wu authored Jul 30, 2017
  
  844d8e69
30 Jul, 2017 3 commits
- only freeze trainable variables (fix #351) · 2a1af832
  Yuxin Wu authored Jul 30, 2017
  
  2a1af832
- fix build · 2d5b44e6
  Yuxin Wu authored Jul 30, 2017
  
  2d5b44e6
- clean-up docs and imports · 5134338e
  Yuxin Wu authored Jul 30, 2017
  
  5134338e