numenta nupic.embodied issues

numenta / nupic.embodied

GNU Affero General Public License v3.0

7 stars 7 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Extra experiments per journal review feedback

#67 abhi-iyer closed 2 years ago
0
Changes to wandb arguments and requirements.txt

#66 lucasosouza closed 2 years ago
0
Update paper_experiments.py

#65 abhi-iyer closed 2 years ago
0
implements batchnorm and layernorm for MLPs

#64 lucasosouza opened 2 years ago
0
Notebook with latest analysis

#63 lucasosouza closed 2 years ago
0
Readme for Multitask

#62 lucasosouza closed 2 years ago
1
Paper configs

#61 abhi-iyer closed 2 years ago
0
Notebook for MTRL core results

#60 karangrewal closed 2 years ago
1
notebook to export data from wandb and other small changes

#59 lucasosouza closed 2 years ago
1
adds option to set seed for weight initialization

#58 lucasosouza closed 2 years ago
0
Reversing preprocessing linear layer bias change

#57 lucasosouza closed 2 years ago
1
Fix to override weight initialization, making it mean centered

#56 lucasosouza closed 2 years ago
0
overrides weight initialization, removes bias in preprocessed layers

#55 lucasosouza closed 2 years ago
0
bug fixes and new debugging notebook

#54 lucasosouza closed 2 years ago
0
changes checkpoint frequency

#53 lucasosouza closed 2 years ago
0
Support for policy hooks for data visualization (single and composite), env updates in trainer, saving/restoring eval env state

#52 lucasosouza closed 2 years ago
2
Visualizations

#51 abhi-iyer closed 3 years ago
1
correct fix to restore flag

#50 lucasosouza closed 3 years ago
0
fix bug when not restoring

#49 lucasosouza closed 3 years ago
0
adds save and restore capabilities to trainer and algorithm

#48 lucasosouza closed 3 years ago
2
Update multiseg_experiments.py

#47 abhi-iyer closed 3 years ago
0
Refactors policy/q-functions, pushes old code to deprecated, adds new optimizations/modular dendritic layers

#46 abhi-iyer closed 3 years ago
0
Refactors training loop, removes bidirectional message passing between trainer and algo

#45 lucasosouza closed 3 years ago
3
FP16 training

#44 abhi-iyer closed 3 years ago
0
change eval frequency to 10

#43 lucasosouza closed 3 years ago
0
multitask: API refactoring , style changes to adhere to PEP8, local Ray GPU sampler

#42 lucasosouza closed 3 years ago
0
Memory profiler

#41 lucasosouza closed 3 years ago
2
Don't log model files while debugging

#40 vkakerbeck closed 3 years ago
0
Changes references from "robot-arm" to "disagreement", small fixes in np to tensor conversion

#39 lucasosouza closed 3 years ago
0
Small fixes on model loading

#38 vkakerbeck closed 3 years ago
1
Reorganization into two frameworks, disagreement and multitask + small fixes to disagreement

#37 lucasosouza closed 3 years ago
0
EOF and Segmentation Error after upgrading to python=3.8 and replacing baselines with stable_baselines

#36 lucasosouza opened 3 years ago
2
Changes to make it compatible with python 3.7

#35 lucasosouza closed 3 years ago
0
Small changes after first test of the new implementation

#34 vkakerbeck closed 3 years ago
0
Refactoring of learn.py and Trainer class

#33 lucasosouza closed 3 years ago
0
Makes auxiliary task independent of dynamic models

#32 lucasosouza closed 3 years ago
0
Bug fixes for ppo update & backprob speed issue

#31 vkakerbeck closed 3 years ago
0
Small fixes and some comments and logging for debiggung speed issue

#30 vkakerbeck closed 3 years ago
0
Torchifies disagreement experiments

#29 lucasosouza closed 3 years ago
0
Minor Logging fixes, Adding 2 new hyperparameter

#28 vkakerbeck closed 3 years ago
0
Code cleanup, refactor of the update function, and improvements to backprop through reward

#27 lucasosouza closed 3 years ago
0
Initial draft of backprop through reward

#26 lucasosouza closed 3 years ago
0
Implement gradient interference in MultiTask project

#25 lucasosouza closed 2 years ago
0
Change experiment names for SingleSegReLU1 and SingleSegReLU2

#24 akashvelu closed 3 years ago
0
New experiment configs

#23 akashvelu closed 3 years ago
0
Refactor Dendritic MLP and Policies/QFunctions to use nupic.research's dendritic mlp

#22 akashvelu closed 3 years ago
0
Add MT50 runner script with MTSAC and different experiment configs for MT50.

#21 akashvelu closed 3 years ago
0
Changes evaluation frequency

#20 lucasosouza closed 3 years ago
0
Add Custom (MT)SAC files

#19 akashvelu closed 3 years ago
0
Include network creation commits, fix min and max std bug for dendritic policy.

#18 akashvelu closed 3 years ago
0