Skip to content

Releases: Dobiasd/frugally-deep

v0.15.12-p0

26 Aug 09:46
Compare
Choose a tag to compare
  • Added support for negative_slope and threshold in ReLU layers

v0.15.11-p0

26 Aug 05:27
d6330d3
Compare
Choose a tag to compare
  • Added support for relu6 activation

v0.15.10-p0

11 Aug 04:17
d231b64
Compare
Choose a tag to compare
  • Added support for layers RandomContrast, RandomFlip, RandomHeight, RandomTranslation, RandomWidth, and RandomZoom (all no-ops during prediction)

v0.15.9-p0

07 Aug 05:58
c2e7b1a
Compare
Choose a tag to compare
  • Added support for activations exponential, gelu, and softsign

v0.15.8-p0

01 Aug 14:11
Compare
Choose a tag to compare
  • Added support for RepeatVector layer

v0.15.7-p0

08 Jul 09:28
Compare
Choose a tag to compare
  • Added support for swish activation.
  • Added support for full models inside TimeDistributed layers.

v0.15.6-p0

07 Jul 05:26
Compare
Choose a tag to compare

v0.15.5-p0

06 Jul 12:59
Compare
Choose a tag to compare
  • Added compatibility for models including layers with training=True by ignoring the flag.

v0.15.4-p0

03 Jul 14:52
d2b5758
Compare
Choose a tag to compare
  • Added support for using BatchNormalization layers as the inner layer in TimeDistributed layers. 🤪

v0.15.3-p0

02 Jul 15:02
Compare
Choose a tag to compare
  • Added support for shape inference in Reshape layers.
  • Improved some docs.