Nuts and bolts deep learning library to make training neural networks easier. Features:
- Logging functionality not at set time intervals but as a percentage of the total training time
- Convenient specification for Random and Grid hyperparameter search: queued with a single GPU, split over multiple local GPUs, or over a Slurm Cluster
- Clean implementation of popular methods/problems in CV such as Vanilla Image Classification, Regression, the PiModel for Semi-Supervised learning, and SN-GAN; all through the common Trainer abstraction
To install, run pip install git+https://github.com/mfinzi/olive-oil-ml. Dependencies will be checked and installed from the setup.py file.
- Python 3.7+
- PyTorch 1.3.0+
- torchvision
- tqdm 4.40+
- natsort
- (optional) tensorboardX
To get a feel for the library, try training a (Classifier, Regressor, PiModel, GAN) model from our recipes. For classification try running
python oil/recipes/simpleTrial.py --dataset CIFAR100 --num_epochs 10
Or, to train a conditional GAN model:
python oil/recipes/simpleCGan.py --dataset SVHN --lr 2e-4
Or train a PiModel semisupervised on CIFAR10 using only 1k labels:
python oil/recipes/simplePi.py --dataset CIFAR10 --train 1000
You can use -h to see the full range of arguments available. Command line arguments and defaults are automatically inferred
from the code used to construct the trial, so you can make a new trial (that uses some exotic data augmentation strategy for example) and the command line parser will generated for you, see the example recipes for how this works.
Example: Search over hyperparameters for CNN classifier on Cifar100
python oil/recipes/exampleHyperSearch.py --dataset CIFAR100 --bs [50,32,64] --k [64,96] --num_epochs 100See example code for programmatic way of specifying the hyperparameter search. Automatically parallelizes the search over multiple GPUs if available.