Releases: Dobiasd/frugally-deep
Releases · Dobiasd/frugally-deep
v0.18.2
16 May 20:07
Compare
Sorry, something went wrong.
No results found
Fixed default negative slope for leaky-relu activation from 0.3 to 0.2
v0.18.1
28 Apr 15:58
Compare
Sorry, something went wrong.
No results found
Added support for Normalization layers with invert=True
v0.18.0
05 Apr 07:56
Compare
Sorry, something went wrong.
No results found
Added support for multiple new activation functions/layers
v0.17.1
11 Mar 03:25
Compare
Sorry, something went wrong.
No results found
Added support for negative_slope in LeakyReLU layer
v0.17.0
10 Mar 11:04
Compare
Sorry, something went wrong.
No results found
Added support for Conv1DTranspose and Conv2DTranspose layers
Augmented Python scripts with type annotations
v0.16.3
10 Feb 13:46
Compare
Sorry, something went wrong.
No results found
Fix invalid-node-index error when using autoencoder models assembled in a specific way
v0.16.2
20 Jan 07:00
Compare
Sorry, something went wrong.
No results found
Switch from tf.keras to the separate keras package
v0.16.1
19 Jan 17:57
Compare
Sorry, something went wrong.
No results found
Update TensorFlow to version 2.18
Some minor cleanups
v0.16.0
16 Apr 14:16
Compare
Sorry, something went wrong.
No results found
Updates TensorFlow from 2.15 to 2.16, which resulted in some non-minor changes.
Dropped support for Bidirectional, GRU, LSTM, and stateful models. They might be re-added again in the future though.
v0.15.31
21 Feb 10:18
Compare
Sorry, something went wrong.
No results found
Simplified the internal implementation of sum_tensors
Removed some redundant internal stuff
Auto-formatted all code