https://github.com/chagmgang/simclr_pytorch

A Simple Framework for Contrastive Learning of Visual Representations of pytorch implementation
https://github.com/chagmgang/simclr_pytorch

deep-learning pytorch representation-learning simclr

Last synced: about 1 month ago
JSON representation

A Simple Framework for Contrastive Learning of Visual Representations of pytorch implementation

• Host: GitHub
• URL: https://github.com/chagmgang/simclr_pytorch
• Owner: chagmgang
• Created: 2020-03-16T12:48:47.000Z (over 6 years ago)
• Default Branch: master
• Last Pushed: 2020-03-23T00:11:01.000Z (over 6 years ago)
• Last Synced: 2025-08-10T10:29:34.694Z (11 months ago)
• Topics: deep-learning, pytorch, representation-learning, simclr
• Language: Python
• Homepage:
• Size: 2.53 MB
• Stars: 3
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# A Simple Framework for Contrastive Learning of Visual Representations with Pytorch

* Pytorch Impementation [A Simple Framework for Contrastive Learning of Visual Representations](https://arxiv.org/abs/2002.05709)

## DataAugmentation

|   |  negative |  negative |  negative | negative  |

|---|---|---|---|---|

|  positive |  ![6](source/positive/6.png) |  ![7](source/positive/7.png) |  ![8](source/positive/8.png) |  ![9](source/positive/9.png) |

|  positive |  ![1](source/positive/1.png) |  ![2](source/positive/2.png) |  ![3](source/positive/3.png) |  ![4](source/positive/4.png) |

## Dependency

```

python-opencv

torch

torchvision

matplotlib

pandas

PIL

```

## Directory

```

├── data

│   ├── 1.jpg

│   ├── 2.jpg

│   ├── 3.jpg

│   ├── ...

│   ├── 5.jpg

│   └── 6.jpg

├── .gitignore

├── dataset_wrapper.py

├── model.py

├── nt_xent_loss.py

├── README.md

└── simclr.py

```

## RUN

```

(pytorch) xxx:xxx xxx$ python simclr.py 

----------------

epoch : 0

loss  : 1.5892890691757202

lr    : 0.0003

----------------

epoch : 1

loss  : 1.5985093116760254

lr    : 0.00029699999999999996

----------------

...

----------------

epoch : 5

loss  : 0.6665001213550568

lr    : 0.00028529701496999996

----------------

epoch : 6

loss  : 0.5047109872102737

lr    : 0.0002824440448203

----------------

epoch : 7

loss  : 0.4476690739393234

lr    : 0.00027961960437209696

----------------

```

## Reference 

* [A Simple Framework for Contrastive Learning of Visual Representations](https://arxiv.org/abs/2002.05709)

* [https://github.com/sthalles/SimCLR](https://github.com/sthalles/SimCLR)