Leon Chen b2764cf337 fix softmax activation for large values

2016-10-10 22:24:32 -04:00

..

activations

fix softmax activation for large values

2016-10-10 22:24:32 -04:00

advanced_activations

move all layer params into attrs object

2016-09-18 23:20:59 -04:00

convolutional

use weblas.pipeline.sgemm for conv layers when gpu flag turned on

2016-10-09 12:55:02 -04:00

core

use weblas.pipeline.sgemm for conv layers when gpu flag turned on

2016-10-09 12:55:02 -04:00

embeddings

remove .only from test

2016-09-17 00:24:58 -04:00

normalization

implement BatchNormalization in all modes, with tests

2016-09-05 22:59:21 -04:00

pooling

implement MaxPooling3D/AveragePooling3D, with tests

2016-09-05 15:35:24 -04:00

recurrent

use weblas.pipeline.sgemm for conv layers when gpu flag turned on

2016-10-09 12:55:02 -04:00

wrappers

use weblas.pipeline.sgemm for conv layers when gpu flag turned on

2016-10-09 12:55:02 -04:00

globals.js

update test console display formatting

2016-08-26 23:36:12 -04:00

index.html

implement TimeDistributed wrapper layer, with tests

2016-09-25 21:39:10 -04:00