dgcnz relaxed-equivariance-dynamics issues

dgcnz / relaxed-equivariance-dynamics

Code for "Effect of equivariance on training dynamics"

2 stars 0 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

feat: update poetry and add gcnn smallest to test hessian

#79 dgcnz closed 3 months ago
0
`aten::grid_sampler_3d_backward` is not implemented

#78 dgcnz opened 3 months ago
1
Revert the bad merge made when fixing math rendering on github

#77 dgcnz closed 3 months ago
0
Update blogpost.md

#76 Nesta-gitU closed 3 months ago
0
Add full equivariance model

#75 dgcnz closed 3 months ago
0
Implement Octahedral gconv

#74 dgcnz closed 3 months ago
0
Meta issue alpha reproduction

#73 Nesta-gitU closed 3 months ago
0
Vectorize rgroup get_weight_constraint

#72 dgcnz closed 3 months ago
2
but patience back to three but put min epochs to 10, because the star…

#71 Nesta-gitU closed 3 months ago
0
Meta issue alpha reproduction

#70 Nesta-gitU closed 3 months ago
0
Meta issue alpha reproduction

#69 Nesta-gitU closed 3 months ago
0
fix trainer configs

#68 Nesta-gitU closed 3 months ago
0
switch trainer patience to 10 epochs

#67 Nesta-gitU closed 3 months ago
0
gamma =0.95 not 0.9 rest is correct

#66 Nesta-gitU closed 3 months ago
0
Meta issue reproduction

#65 Nesta-gitU closed 4 months ago
0
Implement octahedral kernel on gconv

#64 dgcnz closed 3 months ago
4
Research application of octahedral rotations on vector fields

#63 dgcnz opened 4 months ago
1
[Meta Issue] (Wang, 2022) Table 1 Reproduction

#62 dgcnz closed 3 months ago
5
[Meta Issue] Wang 2022 Extension: Cross evaluation by changing model equivariance (alpha) and dataset equivariance

#61 dgcnz closed 3 months ago
2
Rgroup weight constrained

#60 Nesta-gitU closed 4 months ago
0
train_rotation fails with: RuiCnRGCNN.__init__() got an unexpected keyword argument 'alpha'

#59 dgcnz closed 3 months ago
1
debug/fix broadcasting issue when mse_loss

#58 dgcnz closed 4 months ago
3
Add gconv configs

#57 Nesta-gitU closed 4 months ago
0
Train relaxed models on smoke plume for different alphas. Use rotation equivariance full and partial.

#56 Nesta-gitU closed 3 months ago
1
Train relaxed models on smoke plume for different alphas. Use rotation equivariance full and partial.

#55 Nesta-gitU closed 4 months ago
1
add dowload function for the datasets that are up to date with latest…

#54 Nesta-gitU closed 4 months ago
0
feat: add config for wang2022_rsteer reproduction

#53 dgcnz closed 4 months ago
0
Port model training optimizer parameters into configs for [wang 2022] reproduction

#52 dgcnz closed 3 months ago
1
Add base config and setup for wang2022 table 1 reproduction

#51 dgcnz closed 4 months ago
0
Add setup/configs for porting models from [Wang 2022] Table 1

#50 dgcnz closed 4 months ago
0
Draft blog post

#49 dgcnz closed 3 months ago
1
Implement script/config files to reproduce row of Table 1 from [Wang 2022]

#48 dgcnz closed 3 months ago
1
Port models from [Wang 2022] Table 1 reproduction into repo

#47 dgcnz closed 4 months ago
3
Port SmokePlume dataset generator to avoid test leakage

#46 dgcnz closed 4 months ago
2
Implement SmokePlume datamodule from existing data

#45 dgcnz closed 4 months ago
1
Add [wang 2022] rsteerable models to repo

#44 dgcnz closed 4 months ago
0
Cleanup hessian spectra early experiments

#43 dgcnz closed 4 months ago
0
Implement JHTDB dataloader

#42 dgcnz closed 4 months ago
0
add the steerable models to the repo

#41 Nesta-gitU closed 4 months ago
0
clean up the hessian spectra code

#40 Nesta-gitU closed 4 months ago
0
Add e2cnn to conda environment

#39 MeneerTS closed 4 months ago
1
Add sample gcnn pipeline and configs to use train.py

#38 dgcnz closed 4 months ago
0
Create script that takes wandb checkpoint and creates/uploads hessian spectra plot

#37 dgcnz closed 4 months ago
1
Implement super-resolution model architecture

#36 dgcnz closed 3 months ago
13
Initial reproduction of table 1 from [Wang 2022]

#35 dgcnz closed 4 months ago
4
Setup Relaxed Group Convolutional Network

#34 dgcnz closed 4 months ago
0
Setup Regular Group Convolution Network

#33 dgcnz closed 4 months ago
0
Write Background: Training dynamics

#32 dgcnz closed 3 months ago
0
Write Background: Relaxed Group Equivariance

#31 dgcnz closed 3 months ago
0
Write Results

#30 dgcnz closed 3 months ago
0