Commits

Breadcrumbs

History for

onmaster

User selector

All users

All time

Commit history

Commits on May 1, 2017

• Inceptionv3, compute batch size from --gpu option instead of hard-coded. (#246)

  ppwwyyxx

  committed
  1b06a41

  Copy full SHA for 1b06a41

Commits on Mar 29, 2017

• make augmentors copy=True by default. Also add shallow copy to MapDataComponents (#203, #207)

  ppwwyyxx

  committed
  a674541

  Copy full SHA for a674541

Commits on Mar 17, 2017

• add session_creator option in TrainConfig (#191)

  ppwwyyxx

  committed
  ccae46f

  Copy full SHA for ccae46f

Commits on Feb 12, 2017

• simplify (and fix) some examples

  ppwwyyxx

  committed
  7e5d3a8

  Copy full SHA for 7e5d3a8

• change interface for optimizer. (#137)

  ppwwyyxx

  committed
  d1041a7

  Copy full SHA for d1041a7

Commits on Feb 10, 2017

• use InputDesc instead of InputVar.

  ppwwyyxx

  committed
  5b29bda

  Copy full SHA for 5b29bda

Commits on Feb 9, 2017

• use uint8 in resnet example

  ppwwyyxx

  committed
  c68686e

  Copy full SHA for c68686e

Commits on Jan 29, 2017

• a rename in examples (not a breaking change)

  ppwwyyxx

  committed
  14b3578

  Copy full SHA for 14b3578

Commits on Jan 28, 2017

• update docs about models

  ppwwyyxx

  committed
  e8c6cbf

  Copy full SHA for e8c6cbf

• 1.0.0rc0 change concat_v2 to concat (#121)

  ppwwyyxx

  committed
  9b85e7a

  Copy full SHA for 9b85e7a

Commits on Jan 27, 2017

• replace step_per_epoch by steps_per_epoch in examples

  ppwwyyxx

  committed
  a9dd0b8

  Copy full SHA for a9dd0b8

Commits on Jan 25, 2017

• [breaking] deprecate callbacks=Callbacks() but use a list.

  ppwwyyxx

  committed
  243e957

  Copy full SHA for 243e957

Commits on Jan 8, 2017

• re-organize predict/; fix TF incompatibile change of sparse_softmax_cross_entropy_loss

  ppwwyyxx

  committed
  84790b7

  Copy full SHA for 84790b7

• fix use of deprecated TF functions.

  ppwwyyxx

  committed
  fdc9076

  Copy full SHA for fdc9076

Commits on Jan 6, 2017

• use positional args instead of list for add_param_summary

  ppwwyyxx

  committed
  bbf41d9

  Copy full SHA for bbf41d9

Commits on Jan 5, 2017

• add mnist-slim example. Trainer does not automatically summary total cost any more

  ppwwyyxx

  committed
  1b73d9c

  Copy full SHA for 1b73d9c

• deprecate TrainConfig.dataset and use 'dataflow' instead

  ppwwyyxx

  committed
  651a5ae

  Copy full SHA for 651a5ae

Commits on Jan 3, 2017

• fix TF deprecation of concat/split/pack/unpack

  ppwwyyxx

  committed
  e94abf6

  Copy full SHA for e94abf6

• fix pep8 style in examples/

  ppwwyyxx

  committed
  9b69e86

  Copy full SHA for 9b69e86

• run autopep8 over examples

  ppwwyyxx

  committed
  233b3b9

  Copy full SHA for 233b3b9

Commits on Nov 19, 2016

• name change

  ppwwyyxx

  committed
  79dbd18

  Copy full SHA for 79dbd18

Commits on Nov 18, 2016

• useless comments in example & GAN

  ppwwyyxx

  committed
  1908fbe

  Copy full SHA for 1908fbe

Commits on Nov 17, 2016

• use get_scalar_var for learning rate

  ppwwyyxx

  committed
  608ad4a

  Copy full SHA for 608ad4a

Commits on Nov 6, 2016

• [Incompatible] refactor inferencer. ClassificationError now takes a vector

  ppwwyyxx

  committed
  740e9d8

  Copy full SHA for 740e9d8

Commits on Oct 19, 2016

• example docs and improvements

  ppwwyyxx

  committed
  a041f5a

  Copy full SHA for a041f5a

Commits on Oct 17, 2016

• warning about default relu

  ppwwyyxx

  committed
  6eb0beb

  Copy full SHA for 6eb0beb

• fix image_sampling, and email

  ppwwyyxx

  committed
  dc59ad5

  Copy full SHA for dc59ad5

Commits on Oct 15, 2016

• refine docs and examples

  ppwwyyxx

  committed
  2634a25

  Copy full SHA for 2634a25

Commits on Oct 12, 2016

• update hed example

  ppwwyyxx

  committed
  3c27064

  Copy full SHA for 3c27064

Commits on Oct 7, 2016

• batchdatabyshape

  ppwwyyxx

  committed
  0fa11e6

  Copy full SHA for 0fa11e6

Commits on Oct 6, 2016

• more augmentor

  ppwwyyxx

  committed
  2a44407

  Copy full SHA for 2a44407

Commits on Oct 2, 2016

• inception name

  ppwwyyxx

  committed
  a6ece65

  Copy full SHA for a6ece65

Commits on Aug 31, 2016

• migrate to new model interface without is_training

  ppwwyyxx

  committed
  a6a0889

  Copy full SHA for a6a0889

Commits on Aug 16, 2016

• gym atari

  ppwwyyxx

  committed
  cee7999

  Copy full SHA for cee7999

Commits on Aug 3, 2016

• several bug fix and improvement

  ppwwyyxx

  committed
  c95292d

  Copy full SHA for c95292d