Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/Naagar/Normalizing_Flow_3x3_inv

Implementation of CInC Flow paper, TensorFlow.
https://github.com/Naagar/Normalizing_Flow_3x3_inv

Last synced: about 2 months ago
JSON representation

Implementation of CInC Flow paper, TensorFlow.

Host: GitHub
URL: https://github.com/Naagar/Normalizing_Flow_3x3_inv
Owner: Naagar
Created: 2021-01-22T21:04:57.000Z (over 3 years ago)
Default Branch: main
Last Pushed: 2023-07-14T08:29:57.000Z (about 1 year ago)
Last Synced: 2024-07-12T12:41:02.404Z (2 months ago)
Language: Jupyter Notebook
Homepage:
Size: 2.55 MB
Stars: 7
Watchers: 2
Forks: 0
Open Issues: 0
Metadata Files:
- Readme: README.md

Awesome Lists containing this project

awesome-normalizing-flows - [Code

README

# [CInC Flow: Characterizable Invertible 3X3 Convolution](https://arxiv.org/abs/2107.01358)
[OpenReview](https://openreview.net/forum?id=kl1ds_AeLRM)

PAPER: [link](https://arxiv.org/abs/2107.01358)

Accepted: 4th workshop TPM 2021 (UAI-21)
![TPM-21](images/tpm21.png)

Implementation of improvements for generative normalizing flows and more specifically Glow.

We extend the 1x1 convolutions used in glow to convolutions with any kernel size and we introduce a new coupling layer.

This work is adapted from [Emerging Convolutions for Generative Normalizing Flows](https://github.com/ehoogeboom/emerging):
```
Emiel Hoogeboom, Rianne van den Berg, and Max Welling. Emerging Convolutions for Generative Normalizing Flows. International Conference on Machine Learning, 2019.
```
## Method
![Why CInC?](images/CInC_Flow.png)

## Flow module

![module](images/norm_flow.png)

## Image interpolation sample
![Changing Hair Color, original: right](images/blond_hair_id142.png )
![Changing smile](images/high_cheekbones_id7.png)
![Removing the glasses](images/glasses_id151.png)
![Gradually changing the age parameter](images/id42_aging.png)

## Requirements
The ```pip_installs``` script can be used to install all the required packages using pip.

## Download datasets
CIFAR10 is automatically downloaded.
Galaxy images need to be downloaded [here](https://github.com/SpaceML/merger_transfer_learning).

ImageNet 32x32 and 64x64 was downloaded from the link on the Glow github: `https://storage.googleapis.com/glow-demo/data/{dataset_name}-tfr.tar`
with `imagenet-oord` as dataset_name.

## How to start reading the code?
The quad coupling layer is defined on line 409 of the ```model.py``` file.
The convolution is defined on line 463 of the ```conv2d/conv2d.py``` file. The corresponding inversion operation can be found in ```conv2d/inverses/inverse_cython.py``` and ```conv2d\inverses\inverse_op_cython.pyx```.

## Experiments
To get infos regarding the parameter use ```python3 train.py -h```.

##### CIFAR-10 results
Emerging:
```
mpiexec -n 2 python3 train.py --problem cifar10 --image_size 32 --n_level 3 --depth 32 --flow_permutation 3 --flow_coupling 1 --seed 2 --learnprior --lr 0.001 --n_bits_x 8 --epochs 4001
```

Glow:
```
mpiexec -n 2 python3 train.py --problem cifar10 --image_size 32 --n_level 3 --depth 32 --flow_permutation 2 --flow_coupling 1 --seed 2 --learnprior --lr 0.001 --n_bits_x 8 --epochs 4001
```

3x3 convolution:
```
mpiexec -n 2 python3 train.py --problem cifar10 --image_size 32 --n_level 3 --depth 32 --flow_permutation 7 --flow_coupling 1 --seed 2 --learnprior --lr 0.001 --n_bits_x 8 --epochs 4001
```

3x3 convolution and quad-coupling:
```
mpiexec -n 2 python3 train.py --problem cifar10 --image_size 32 --n_level 3 --depth 32 --flow_permutation 7 --flow_coupling 2 --seed 2 --learnprior --lr 0.001 --n_bits_x 8 --epochs 4001
```

##### ImageNet 32x32 results
This command lines assumes that the variable ```DATA_PATH``` contains the
path to the imagenet dataset.

Emerging:
```
mpiexec -n 4 python3 train.py --problem imagenet-oord --image_size 32 --n_level 3 --depth 48 --flow_permutation 3 --flow_coupling 1 --seed 0 --learnprior --lr 0.001 --n_bits_x 8 --data_dir $DATA_PATH
```

Glow:
```
mpiexec -n 4 python3 train.py --problem imagenet-oord --image_size 32 --n_level 3 --depth 48 --flow_permutation 2 --flow_coupling 1 --seed 0 --learnprior --lr 0.001 --n_bits_x 8 --data_dir $DATA_PATH
```

3x3 convolution:
```
mpiexec -n 4 python3 train.py --problem imagenet-oord --image_size 32 --n_level 3 --depth 48 --flow_permutation 7 --flow_coupling 1 --seed 0 --learnprior --lr 0.001 --n_bits_x 8 --data_dir $DATA_PATH
```

3x3 convolution and quad-coupling:
```
mpiexec -n 4 python3 train.py --problem imagenet-oord --image_size 32 --n_level 3 --depth 48 --flow_permutation 7 --flow_coupling 2 --seed 0 --learnprior --lr 0.001 --n_bits_x 8 --data_dir $DATA_PATH
```

##### ImageNet 64x64 results
Emerging:
```
mpiexec -n 4 python3 train.py --problem imagenet-oord --image_size 64 --n_level 4 --depth 48 --flow_permutation 3 --flow_coupling 1 --seed 0 --learnprior --lr 0.001 --n_bits_x 8 --data_dir $DATA_PATH
```

Glow:
```
mpiexec -n 4 python3 train.py --problem imagenet-oord --image_size 64 --n_level 4 --depth 48 --flow_permutation 2 --flow_coupling 1 --seed 0 --learnprior --lr 0.001 --n_bits_x 8 --data_dir $DATA_PATH
```

3x3 convolution:
```
mpiexec -n 4 python3 train.py --problem imagenet-oord --image_size 64 --n_level 4 --depth 48 --flow_permutation 7 --flow_coupling 1 --seed 0 --learnprior --lr 0.001 --n_bits_x 8 --data_dir $DATA_PATH
```

3x3 convolution and quad-coupling:
```
mpiexec -n 4 python3 train.py --problem imagenet-oord --image_size 64 --n_level 4 --depth 48 --flow_permutation 7 --flow_coupling 2 --seed 0 --learnprior --lr 0.001 --n_bits_x 8 --data_dir $DATA_PATH
```

## Sample time results
Emerging:
```
mpiexec -n 1 python3 train.py --problem cifar10 --image_size 32 --n_level 3 --depth 32 --flow_permutation 3 --flow_coupling 1 --seed 2 --learnprior --lr 0.001 --n_bits_x 8 --sample
```

Glow:
```
mpiexec -n 1 python3 train.py --problem cifar10 --image_size 32 --n_level 3 --depth 32 --flow_permutation 2 --flow_coupling 1 --seed 2 --learnprior --lr 0.001 --n_bits_x 8 --sample
```

3x3 convolution:
```
mpiexec -n 1 python3 train.py --problem cifar10 --image_size 32 --n_level 3 --depth 32 --flow_permutation 7 --flow_coupling 1 --seed 2 --learnprior --lr 0.001 --n_bits_x 8 --sample
```

3x3 convolution and quad-coupling:
```
mpiexec -n 1 python3 train.py --problem cifar10 --image_size 32 --n_level 3 --depth 32 --flow_permutation 7 --flow_coupling 2 --seed 2 --learnprior --lr 0.001 --n_bits_x 8 --sample
```
## have a question?, contact: [email protected]