Requirement to run Single-GPU training

facebookresearch / suncet

Code to reproduce the results in the FAIR research papers "Semi-Supervised Learning of Visual Features by Non-Parametrically Predicting View Assignments with Support Samples" https://arxiv.org/abs/2104.13963 and "Supervision Accelerates Pre-training in Contrastive Semi-Supervised Learning of Visual Representations" https://arxiv.org/abs/2006.10803

MIT License

486 stars 67 forks source link

Hi @3bao,

Assuming you've already cloned the repo and installed the following packages:

Python 3.8
PyTorch install 1.7.1
torchvision
CUDA 11.0
Apex with CUDA extension
Other dependencies: PyYaml, numpy, opencv, submitit

After that you can just follow the instructions here In short, you just need to

download CIFAR10 to the appropriate path
create a config file and save it in the configs/paws/ directory

run the command

python main.py
--sel paws_train
--fname configs/paws/3bao_config_file.yaml

You can see an example config file for single-GPU training on CIFAR10 here

facebookresearch / suncet

Requirement to run Single-GPU training #31