58 Commits

Author SHA1 Message Date
5e8f7cb69f Update parent_layers.py 2023-11-04 14:48:10 -04:00
e1d9db27c6 Update parent_layers.py 2023-11-04 14:32:22 -04:00
61cce04d7a Fixed title spacing 2023-04-14 13:39:39 -04:00
775d597517 Fixed bug with scaling title. 2023-04-13 19:30:21 -04:00
ca7978929a Changed default functionality of activation functions so they are all lined up with eachother. 2023-04-09 20:15:07 -04:00
3967cbea69 Reset default color scheme to dark mode. Removed unnecessary print statements. 2023-04-03 00:30:21 -04:00
a09c9da2f6 Got colorschemes setup (not globally yet). Fixed bug with layer placing. 2023-04-03 00:23:09 -04:00
90c9dba217 Removed remover argument from ReplacementTrasform which resulted in "Could not find old_mobject in Scene" when cleaning up the scene. (#34) 2023-03-28 16:54:24 -04:00
b1c838a45f Added residual layer example. Fixed some bugs in the process. 2023-02-05 12:26:36 -05:00
7538e2b620 Merge branch 'main' of github.com:helblazer811/ManimML 2023-02-02 22:00:07 -05:00
134be057fb Made mcmc example. Added ability to view matplotlib plots. 2023-02-02 21:59:31 -05:00
42defde750 Dropout Last Layer - option to remove node on the last layer and seed dropouts for reproducible animations (#30)
Option to remove node on the last layer and seed dropouts for reproducible animations
2023-02-02 19:27:19 -05:00
9698907cbf Fixed padding bug with max pooling 2023-02-02 09:42:08 -05:00
1406acd43e Refactored makefile, moved around files to more appropriate places, deleted unused files,
reformatted code.
2023-02-01 22:33:42 -05:00
27d235de25 Added the ability to make residual connections.
Note: still need to add the residual plus icon.
2023-02-01 12:40:43 -05:00
4b06ce1622 Added padding.
Fixed a bug with ImageLayerToConvolutional2D

Padding example
2023-01-31 23:37:55 -05:00
c14972fa4b Refactored some code in neural network, added max pooling to feed forward 2023-01-31 10:30:49 -05:00
b767b9abee Simplified namespace so layers can be imported from manim_ml.neural_network 2023-01-27 14:20:11 -05:00
46958ea293 Working max pooling visualization. 2023-01-26 18:39:38 -05:00
c824409eee Fixed bug with feed forward layer activation. 2023-01-25 17:41:20 -05:00
301b230c73 Added sigmoid activation, and ability to do activation above linear layers. 2023-01-25 17:36:44 -05:00
f56620f047 Bug fixes and linting for the activation functions addition. 2023-01-25 08:40:32 -05:00
ce184af78e General changes, got basic visualization of an activation function working for a
convolutinoal layer.
2023-01-24 15:44:48 -05:00
5291d9db8a Refactored parameters for Convolutional2D to make them clearner looking. 2023-01-15 18:16:48 +09:00
27523ad337 Fixed bug with the size of images in ImageToConvolutional2DLayer 2023-01-15 18:00:40 +09:00
99dbda915b Finshed adding construct_layer methods for refactor. 2023-01-15 16:52:22 +09:00
42b6e37b16 Removed Conv2D because it can be done using just Conv3D and renamed Conv3D to Conv2D to correspond to the spatial conv dimenson not the scene dimension, which is more inline with convention. 2023-01-15 14:35:26 +09:00
ba63116b37 Reformatted the code using black, allowd for different orientation NNs, made an option for highlighting the active filter in a CNN forward pass. 2023-01-09 15:52:37 +09:00
e33f98373a Working initial visualization of a static decision tree. 2023-01-04 20:50:01 -05:00
1d9cad8587 Working dropout animation 2023-01-02 15:24:51 -05:00
3d6e8072e1 Used Black to reformat the code in the repository. 2023-01-01 23:24:59 -05:00
f8701c8a91 Updated examples 2022-12-30 22:32:44 -05:00
b1a85ea782 Overall working 3D convolution visualization. 2022-12-30 21:41:50 -05:00
ff4c1ffded Got image mobject to conv3d transformation working (not great, but passable) 2022-12-29 22:45:15 -05:00
245ce9424b Made a workaround to make sure the filters in the CNN 2 CNN layers
don't appear at the beggining of a forward pass animation. I think
this has to do with a bug in Succession in the core ManimCommunity
library.
2022-12-29 15:11:09 -05:00
8cee86e884 Working initial visualization of a CNN. 2022-12-29 14:09:16 -05:00
330ba170a0 "Got a basic working animation of a CNN in 3D. 2022-12-26 02:01:56 -05:00
58aec269cf Convolutional Layers 2022-05-15 13:42:21 -04:00
2ef4dcab44 Init bug fix 2022-05-07 00:42:13 -04:00
8d45c13f57 Added working convolutional layer. 2022-04-29 14:36:14 -04:00
9310b48c56 Finished oracle guidance video. Integrated various changes necessary to complete this. 2022-04-28 01:37:52 -04:00
63427be139 Added ability to pass layer_args dictionary to each forward pass, which allows
for arguments to be passed through to each neural network layer when running a neural
network forward pass.
2022-04-25 16:28:11 -04:00
7d04bf55ec Added updated VAEScene to the readme. 2022-04-23 23:10:15 -04:00
0152be64b0 Added GAN visualization. 2022-04-22 19:08:28 -04:00
ffd31701bf Working remove neural network animation 2022-04-21 23:18:58 -04:00
11bbd59bb6 Vector layer (Still work in progress) 2022-04-19 02:10:53 -04:00
229c27fa3f Added support for Neural Network overhead title and per-layer title. 2022-04-19 00:36:15 -04:00
05f512f185 Added ListGroup class for better management of a group of objects. 2022-04-18 23:56:31 -04:00
0febbe547d Refactored LabeledColorImage 2022-04-16 00:37:19 -04:00
2306ab39d1 Paired Query Layers and tests 2022-04-16 00:23:10 -04:00