steffen/self-replicating-neural-networks
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Readme Update und Residuals GN Training

Browse Source
This commit is contained in:
Steffen Illium 2022-02-03 18:03:01 +01:00
parent 6b1efd0c49
commit 382afa8642
1 changed files with 6 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
README.md
View File

@ -33,13 +33,16 @@
- [x] Training mit kleineren GNs
- Accuracy leidet enorm (_0.56_)
![image info](./figures/training_lineplot.png)
![image info](./figures/lesser_neurons_training_lineplot.png)
- Es entstehen mehr SRNN
- Der Dropout Effekt wird stärker (diff_ohne_SRNN = _0.0_)
![image info](./figures/dropout_stacked_barplot.png)
![image info](./figures/lesser_neurons_dropout_stacked_barplot.png)
- [ ] Weiter Trainieren -> 500 Epochs?
- [ ] Loss Gewichtung anpassen
- [ ] Training ohne Residual Skip Connection | - Running
- [x] Training ohne Residual Skip Connection
- Ist kacke
![image info](./figures/no_res_training_lineplot.png)
![image info](./figures/no_res_dropout_stacked_barplot.png)
- [ ] Test mit Baseline Dense Network
- [ ] mit vergleichbaren Neuron Count
- [ ] mit gesamt Weight Count