Commits

Breadcrumbs

History for

on6900d93

User selector

All users

Datepicker

All time

Commit History

Commits on Mar 5, 2018

• Fixes issue 160 (BrainJS#162)

  Show description for 6900d93
  freddyC

  authored
  6900d93

  Copy full SHA for 6900d93

Commits on Mar 4, 2018

• fix legacy to support when no trainOpts are sent in, with test

  Show description for 4d94d16
  robertleeplummerjr

  committed
  4d94d16

  Copy full SHA for 4d94d16

Commits on Feb 18, 2018

• Updated to always throw an error.

  Show description for b8a5831
  freddyC

  committed
  b8a5831

  Copy full SHA for b8a5831

• Handles User Object Validation BrainJS#142

  Show description for 56ed8fa
  freddyC

  committed
  56ed8fa

  Copy full SHA for 56ed8fa

Commits on Feb 2, 2018

• Missed a _updateTrainingOpts in the merge, remade browser and dist

  perkyguy

  committed
  bc18f46

  Copy full SHA for bc18f46

• Merge in changes on develop

  perkyguy

  committed
  86edaf3

  Copy full SHA for 86edaf3

Commits on Feb 1, 2018

• Moved the logic for not training anymore to _trainingTick which now returns a boolean to show whether or not it ran. This allows for no additional closures, but still keeping train and `trainAs…

  Show description for 2a8fb2f
  perkyguy

  committed
  2a8fb2f

  Copy full SHA for 2a8fb2f

Commits on Jan 31, 2018

• train and trainAsync now have a common _prepTraining method which sets things up for training. This returns what should happen during a training tick. This is essentially to keep the training metho…

  Show description for 8ff2e3b
  perkyguy

  committed
  8ff2e3b

  Copy full SHA for 8ff2e3b

Commits on Jan 30, 2018

• updateTrainingOptions changed to an "internal only" function by adding in the _ prefix. Also, made the operation a little less fragile by updating any options which are in the default object, rathe…

  Show description for c7a3336
  perkyguy

  committed
  c7a3336

  Copy full SHA for c7a3336

• Updates the json test as well as doing a make.

  freddyC

  committed
  0d1a110

  Copy full SHA for 0d1a110

• Updates

  Show description for e9f1401
  freddyC

  committed
  e9f1401

  Copy full SHA for e9f1401

• Possible suggestions for clarity

  freddyC

  committed
  91d00a4

  Copy full SHA for 91d00a4

Commits on Jan 29, 2018

• BrainJS#135 Used a filter rather than map().every() to get the errors with individual settings.

  perkyguy

  committed
  5c36b39

  Copy full SHA for 5c36b39

• Added some basic handling for when a net is run without training. Checks for the side effects, rather than if train() was ran.

  perkyguy

  committed
  23bedfd

  Copy full SHA for 23bedfd

Commits on Jan 27, 2018

• Merge branch 'develop' into add-thaw

  Show description for b828d84
  robertleeplummerjr

  committed
  b828d84

  Copy full SHA for b828d84

Commits on Jan 26, 2018

• move toFunction into its own test, for easier debugging

  Show description for df5d6e3
  robertleeplummerjr

  committed
  df5d6e3

  Copy full SHA for df5d6e3

Commits on Jan 23, 2018

• Removed goodies

  Show description for b4cd266
  freddyC

  committed
  b4cd266

  Copy full SHA for b4cd266

• Remember that time

  Show description for 4782432
  freddyC

  committed
  4782432

  Copy full SHA for 4782432

Commits on Jan 22, 2018

• Loggity

  Show description for 6854435
  freddyC

  committed
  6854435

  Copy full SHA for 6854435

Commits on Jan 21, 2018

• Updates all the things.

  Show description for 69cfeae
  freddyC

  committed
  69cfeae

  Copy full SHA for 69cfeae

• Committing before I change to promise so I don't break things too bad

  freddyC

  committed
  5e7a4e5

  Copy full SHA for 5e7a4e5

Commits on Jan 20, 2018

• Rebased onto master and small update for initialize to not send data but sizes

  freddyC

  committed
  7556c1c

  Copy full SHA for 7556c1c

• updates training to train async or not and fixes the tests to conform

  freddyC

  committed
  c5933bb

  Copy full SHA for c5933bb

• This adds thaw.js (one failing test)

  Show description for ce2bff7
  freddyC

  committed
  ce2bff7

  Copy full SHA for ce2bff7

Commits on Jan 19, 2018

• Spread output of runInput for both NN and NNGPU, NNGPU does a toArray(this.gpu) first. This prevents object sharing issues.

  perkyguy

  committed
  8e84766

  Copy full SHA for 8e84766

Commits on Jan 13, 2018

• fix BrainJS#102.

  Show description for 5f48479
  robertleeplummerjr

  committed
  5f48479

  Copy full SHA for 5f48479

Commits on Dec 31, 2017

• fixes BrainJS#110, makes data formatter a set of singletons and don't overwrite input, but rather use rawInput

  Show description for 11ae3e8
  robertleeplummerjr

  committed
  11ae3e8

  Copy full SHA for 11ae3e8

Commits on Dec 27, 2017

• change data.constructor !== Array to Array.isArray

  Show description for 44a86f5
  

  root

  committed
  44a86f5

  Copy full SHA for 44a86f5

Commits on Dec 26, 2017

• refine defaults into class

  robertleeplummerjr

  committed
  d470849

  Copy full SHA for d470849

• cleanup of unused, unneeded properties, and default checks

  robertleeplummerjr

  committed
  18e523b

  Copy full SHA for 18e523b

• Ensure proper names are passed down through to gpu createKernelMap

  Show description for 731875d
  robertleeplummerjr

  committed
  731875d

  Copy full SHA for 731875d

Commits on Sep 3, 2017

• handle part of BrainJS#82, erroneous keepNetworkIntact logic changed and renamed to reinforce

  robertleeplummerjr

  committed
  d78d4d3

  Copy full SHA for d78d4d3

• Merge pull request BrainJS#87 from harthur-org/other-activation-functions

  Show description for 4c9baad
  robertleeplummerjr

  authored
  4c9baad

  Copy full SHA for 4c9baad

Commits on Sep 2, 2017

• stronger logic

  robertleeplummerjr

  committed
  6c129f9

  Copy full SHA for 6c129f9

Commits on Sep 1, 2017

• finalize other activation types

  robertleeplummerjr

  committed
  70ab1ee

  Copy full SHA for 70ab1ee

Pagination

Previous

Next