Commits

Branch selector

 master

User selector

All users

All time

Commit History

Commits on Jan 2, 2018

• Merge pull request #257 from shark8me/save-gradients-flag

  Show description for 94b1430
  

  Chris Nuernberger

  authored
  94b1430

  Copy full SHA for 94b1430

Commits on Dec 29, 2017

• support for turning on gradient outputs in :compute-graph's :buffers for debugging

  shark8me

  committed
  9500c78

  Copy full SHA for 9500c78

Commits on Dec 28, 2017

• Merge pull request #256 from thinktopic/tensor-losses

  Show description for 3d295cc
  

  Chris Nuernberger

  authored
  3d295cc

  Copy full SHA for 3d295cc

• Removed the majority of the complexity of the math subsystem

  

  Chris Nuernberger

  committed
  20933d1

  Copy full SHA for 20933d1

• Started removing functionality.

  

  Chris Nuernberger

  committed
  51e1351

  Copy full SHA for 51e1351

• Updating center-loss to use a hypersphere for placing centers

  

  Chris Nuernberger

  committed
  553831f

  Copy full SHA for 553831f

• starting to add radius to center-loss.

  

  Chris Nuernberger

  committed
  da61f96

  Copy full SHA for da61f96

Commits on Dec 21, 2017

• Merge pull request #253 from iboss-ptk/master

  Show description for bd8616d
  harold

  authored
  bd8616d

  Copy full SHA for bd8616d

• Update brew cuda installation instruction

  iboss-ptk

  committed
  ed24867

  Copy full SHA for ed24867

Commits on Dec 6, 2017

• Merge pull request #251 from shark8me/experiment-classification-api

  Show description for 87b7b88
  

  Chris Nuernberger

  authored
  87b7b88

  Copy full SHA for 87b7b88

• Merge pull request #249 from gigasquid/remove-first-dropout-from-mnist-example

  Show description for 1aab2a9
  

  Chris Nuernberger

  authored
  1aab2a9

  Copy full SHA for 1aab2a9

• Merge pull request #247 from gigasquid/selu-activation-and-tensor-operations

  Show description for c8e64df
  

  Chris Nuernberger

  authored
  c8e64df

  Copy full SHA for c8e64df

Commits on Nov 29, 2017

• metrics for train and test loss via the experiment API

  shark8me

  committed
  7cc076d

  Copy full SHA for 7cc076d

Commits on Nov 28, 2017

• Updated centerloss test and fixed centerloss.

  

  Chris Nuernberger

  committed
  de13a55

  Copy full SHA for de13a55

• Trying to get the last loss, center-loss ported over.

  

  Chris Nuernberger

  committed
  cbf90e0

  Copy full SHA for cbf90e0

Commits on Nov 27, 2017

• tweak example to get rid of first dropout layer

  gigasquid

  committed
  42f3d2b

  Copy full SHA for 42f3d2b

• Removing the source of the issue.

  

  Chris Nuernberger

  committed
  edad52a

  Copy full SHA for edad52a

• Trying to fix travis.

  

  Chris Nuernberger

  committed
  792cd4f

  Copy full SHA for 792cd4f

• More indexed tensor testing.

  

  Chris Nuernberger

  committed
  df35236

  Copy full SHA for df35236

• Updated comment, preparing for more thorough (broadcasting) tests.

  

  Chris Nuernberger

  committed
  9139c08

  Copy full SHA for 9139c08

• Indirect access implemented and lightly tested on both cpu,gpu.

  

  Chris Nuernberger

  committed
  3eeae17

  Copy full SHA for 3eeae17

Commits on Nov 26, 2017

• Indirect addressing added step 1.

  

  Chris Nuernberger

  committed
  f64a637

  Copy full SHA for f64a637

• First small bit of indexed dimensions moved forward.

  

  Chris Nuernberger

  committed
  10b8a59

  Copy full SHA for 10b8a59

Commits on Nov 24, 2017

• Tests passing.

  

  Chris Nuernberger

  committed
  81a6ea4

  Copy full SHA for 81a6ea4

• Working through adding indirect tensor addressing.

  

  Chris Nuernberger

  committed
  e10cc37

  Copy full SHA for e10cc37

Commits on Nov 17, 2017

• selu activation & more high level tensor operations

  Show description for 273378f
  gigasquid

  committed
  273378f

  Copy full SHA for 273378f

Commits on Nov 9, 2017

• [experiment/train] Mild improvement to simple loss print. (#246)

  harold

  authored
  195a188

  Copy full SHA for 195a188

Commits on Nov 2, 2017

• Merge pull request #244 from gigasquid/swish-activation

  Show description for 666bb48
  

  Chris Nuernberger

  authored
  666bb48

  Copy full SHA for 666bb48

Commits on Nov 1, 2017

• Merge pull request #243 from gigasquid/non-snapshot-examples

  Show description for 319ba73
  

  Chris Nuernberger

  authored
  319ba73

  Copy full SHA for 319ba73

Commits on Oct 29, 2017

• Swish activation function

  gigasquid

  committed
  34c4c93

  Copy full SHA for 34c4c93

Commits on Oct 25, 2017

• Unified all the examples so that they are clojure 1.9 and

  Show description for 4111ca9
  gigasquid

  committed
  4111ca9

  Copy full SHA for 4111ca9

Commits on Oct 24, 2017

• Merge pull request #241 from gigasquid/resnet-50-example-enhance

  Show description for d5da3ef
  

  Chris Nuernberger

  authored
  d5da3ef

  Copy full SHA for d5da3ef

• don’t use snapshots

  gigasquid

  committed
  308b22f

  Copy full SHA for 308b22f

• revert back original espacing

  gigasquid

  committed
  3284b9d

  Copy full SHA for 3284b9d

• fix spelling errors

  gigasquid

  committed
  a38e531

  Copy full SHA for a38e531