jkoutsikakis / Pytorch Wrapper
Licence: mit
Provides a systematic and extensible way to build, train, evaluate, and tune deep learning models using PyTorch.
Stars: ✭ 92
Programming Languages
python
139335 projects - #7 most used programming language
Projects that are alternatives of or similar to Pytorch Wrapper
Hyperlearn
50% faster, 50% less RAM Machine Learning. Numba rewritten Sklearn. SVD, NNMF, PCA, LinearReg, RidgeReg, Randomized, Truncated SVD/PCA, CSR Matrices all 50+% faster
Stars: ✭ 1,204 (+1208.7%)
Mutual labels: data-science, tensor
Daily Coding Problem
Series of the problem 💯 and solution ✅ asked by Daily Coding problem👨🎓 website.
Stars: ✭ 90 (-2.17%)
Mutual labels: data-science
Deep Learning Boot Camp
A community run, 5-day PyTorch Deep Learning Bootcamp
Stars: ✭ 1,270 (+1280.43%)
Mutual labels: data-science
Pymc Example Project
Example PyMC3 project for performing Bayesian data analysis using a probabilistic programming approach to machine learning.
Stars: ✭ 90 (-2.17%)
Mutual labels: data-science
Knet.jl
Koç University deep learning framework.
Stars: ✭ 1,260 (+1269.57%)
Mutual labels: data-science
Lda Topic Modeling
A PureScript, browser-based implementation of LDA topic modeling.
Stars: ✭ 91 (-1.09%)
Mutual labels: data-science
Sci Pype
A Machine Learning API with native redis caching and export + import using S3. Analyze entire datasets using an API for building, training, testing, analyzing, extracting, importing, and archiving. This repository can run from a docker container or from the repository.
Stars: ✭ 90 (-2.17%)
Mutual labels: data-science
Vvedenie Mashinnoe Obuchenie
📝 Подборка ресурсов по машинному обучению
Stars: ✭ 1,282 (+1293.48%)
Mutual labels: data-science
Pytorch2c
A Python module for compiling PyTorch graphs to C
Stars: ✭ 86 (-6.52%)
Mutual labels: tensor
Starcraft2 Replay Analysis
A jupyter notebook that provides analysis for StarCraft 2 replays
Stars: ✭ 90 (-2.17%)
Mutual labels: data-science
Kaggle Competitions
There are plenty of courses and tutorials that can help you learn machine learning from scratch but here in GitHub, I want to solve some Kaggle competitions as a comprehensive workflow with python packages. After reading, you can use this workflow to solve other real problems and use it as a template.
Stars: ✭ 86 (-6.52%)
Mutual labels: data-science
H2o Tutorials
Tutorials and training material for the H2O Machine Learning Platform
Stars: ✭ 1,305 (+1318.48%)
Mutual labels: data-science
Topic Modeling Tool
A point-and-click tool for creating and analyzing topic models produced by MALLET.
Stars: ✭ 85 (-7.61%)
Mutual labels: data-science
Epfl
EPFL summaries & cheatsheets over 5 years (computer science, communication systems, data science and computational neuroscience).
Stars: ✭ 90 (-2.17%)
Mutual labels: data-science
Danfojs
danfo.js is an open source, JavaScript library providing high performance, intuitive, and easy to use data structures for manipulating and processing structured data.
Stars: ✭ 1,304 (+1317.39%)
Mutual labels: data-science
Applied Ml
📚 Papers & tech blogs by companies sharing their work on data science & machine learning in production.
Stars: ✭ 17,824 (+19273.91%)
Mutual labels: data-science
PyTorch Wrapper
PyTorch Wrapper is a library that provides a systematic and extensible way to build, train, evaluate, and tune deep learning models using PyTorch. It also provides several ready to use modules and functions for fast model development.
Branch | Build | Coverage |
---|---|---|
Master | ||
Develop |
Installation
From PyPI
pip install pytorch-wrapper
From Source
git clone https://github.com/jkoutsikakis/pytorch-wrapper.git
cd pytorch-wrapper
pip install .
Basic abstract usage pattern
import torch
import pytorch_wrapper as pw
train_dataloader = ...
val_dataloader = ...
dev_dataloader = ...
evaluators = { 'acc': pw.evaluators.AccuracyEvaluator(), ... }
loss_wrapper = pw.loss_wrappers.GenericPointWiseLossWrapper(torch.nn.BCEWithLogitsLoss())
model = ...
system = pw.System(model=model, device=torch.device('cuda'))
optimizer = torch.optim.Adam(system.model.parameters())
system.train(
loss_wrapper,
optimizer,
train_data_loader=train_dataloader,
evaluators=evaluators,
evaluation_data_loaders={'val': val_dataloader},
callbacks=[
pw.training_callbacks.EarlyStoppingCriterionCallback(
patience=3,
evaluation_data_loader_key='val',
evaluator_key='acc',
tmp_best_state_filepath='current_best.weights'
)
]
)
results = system.evaluate(dev_dataloader, evaluators)
predictions = system.predict(dev_dataloader)
system.save_model_state('model.weights')
system.load_model_state('model.weights')
Docs & Examples
The docs can be found here.
There are also the following examples in notebook format:
Note that the project description data, including the texts, logos, images, and/or trademarks,
for each open source project belongs to its rightful owner.
If you wish to add or remove any projects, please contact us at [email protected].