High level framework for starting Deep Learning projects
Project description
<a href="http://remicadene.com/bootstrap"><img src="https://github.com/Cadene/bootstrap.pytorch/blob/master/docs/source/_static/img/bootstrap-logo-dark.png" width="50%"/></a>
<a href="https://travis-ci.org/Cadene/bootstrap.pytorch"><img src="https://api.travis-ci.org/Cadene/bootstrap.pytorch.svg?branch=master"/></a>
`Bootstrap` is a high-level framework for starting deep learning projects.
It aims at accelerating research projects and prototyping by providing a powerful workflow focused on your dataset and model only.
And it is:
- Scalable
- Modular
- Shareable
- Extendable
- Uncomplicated
- Built for reproducibility
- Easy to log and plot anything
It's not a wrapper over pytorch, it's a powerful extension.
## Quick tour
To display parsed options from the yaml file:
```
python -m bootstrap.run
-o mnist/options/sgd.yaml
-h
```
To run an experiment (training + evaluation):
```
python -m bootstrap.run
-o mnist/options/sgd.yaml
```
Running an experiment will create 4 files:
- [options.yaml](https://github.com/Cadene/bootstrap.pytorch/blob/master/logs/mnist/sgd/options.yaml) contains the options used for the experiment,
- [logs.txt](https://github.com/Cadene/bootstrap.pytorch/blob/master/logs/mnist/sgd/logs.txt) contains all the information given to the logger.
- [logs.json](https://github.com/Cadene/bootstrap.pytorch/blob/master/logs/mnist/sgd/logs.json) contains the following data: train_epoch.loss, train_batch.loss, eval_epoch.accuracy_top1, etc.
- <a href="https://rawgit.com/Cadene/bootstrap.pytorch/master/logs/mnist/sgd/view.html">view.html</a> contains training and evaluation curves with javascript utilities (plotly).
To save the next experiment in a specific directory:
```
python -m bootstrap.run
-o mnist/options/sgd.yaml
--exp.dir logs/mnist/custom
```
To reload an experiment:
```
python -m bootstrap.run
-o logs/mnist/cuda/options.yaml
--exp.resume last
```
## Documentation
The package reference is available on the [documentation website](http://remicadene.com/bootstrap).
It also contains some notes:
- [Installation](http://remicadene.com/bootstrap/#installation)
- [Concepts](http://remicadene.com/bootstrap/concepts.html)
- [Quickstart](http://remicadene.com/bootstrap/quickstart.html)
- [Directories](http://remicadene.com/bootstrap/directories.html)
- [Examples](http://remicadene.com/bootstrap/examples.html)
## Poster
<a href="http://remicadene.com/bootstrap/_static/img/bootstrap_poster.pdf"><img src="http://remicadene.com/bootstrap/_static/img/bootstrap_poster_mini.png" width="20%"/></a>
<a href="https://travis-ci.org/Cadene/bootstrap.pytorch"><img src="https://api.travis-ci.org/Cadene/bootstrap.pytorch.svg?branch=master"/></a>
`Bootstrap` is a high-level framework for starting deep learning projects.
It aims at accelerating research projects and prototyping by providing a powerful workflow focused on your dataset and model only.
And it is:
- Scalable
- Modular
- Shareable
- Extendable
- Uncomplicated
- Built for reproducibility
- Easy to log and plot anything
It's not a wrapper over pytorch, it's a powerful extension.
## Quick tour
To display parsed options from the yaml file:
```
python -m bootstrap.run
-o mnist/options/sgd.yaml
-h
```
To run an experiment (training + evaluation):
```
python -m bootstrap.run
-o mnist/options/sgd.yaml
```
Running an experiment will create 4 files:
- [options.yaml](https://github.com/Cadene/bootstrap.pytorch/blob/master/logs/mnist/sgd/options.yaml) contains the options used for the experiment,
- [logs.txt](https://github.com/Cadene/bootstrap.pytorch/blob/master/logs/mnist/sgd/logs.txt) contains all the information given to the logger.
- [logs.json](https://github.com/Cadene/bootstrap.pytorch/blob/master/logs/mnist/sgd/logs.json) contains the following data: train_epoch.loss, train_batch.loss, eval_epoch.accuracy_top1, etc.
- <a href="https://rawgit.com/Cadene/bootstrap.pytorch/master/logs/mnist/sgd/view.html">view.html</a> contains training and evaluation curves with javascript utilities (plotly).
To save the next experiment in a specific directory:
```
python -m bootstrap.run
-o mnist/options/sgd.yaml
--exp.dir logs/mnist/custom
```
To reload an experiment:
```
python -m bootstrap.run
-o logs/mnist/cuda/options.yaml
--exp.resume last
```
## Documentation
The package reference is available on the [documentation website](http://remicadene.com/bootstrap).
It also contains some notes:
- [Installation](http://remicadene.com/bootstrap/#installation)
- [Concepts](http://remicadene.com/bootstrap/concepts.html)
- [Quickstart](http://remicadene.com/bootstrap/quickstart.html)
- [Directories](http://remicadene.com/bootstrap/directories.html)
- [Examples](http://remicadene.com/bootstrap/examples.html)
## Poster
<a href="http://remicadene.com/bootstrap/_static/img/bootstrap_poster.pdf"><img src="http://remicadene.com/bootstrap/_static/img/bootstrap_poster_mini.png" width="20%"/></a>
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
bootstrap.pytorch-0.0.10.tar.gz
(26.8 kB
view hashes)
Built Distributions
bootstrap.pytorch-0.0.10-py3.7.egg
(164.8 kB
view hashes)
Close
Hashes for bootstrap.pytorch-0.0.10-py3.7.egg
Algorithm | Hash digest | |
---|---|---|
SHA256 | b641d73af6df79dfd6d85718c083407e335a90dafd1a1b4865df9f544dd45be6 |
|
MD5 | 8e8dc0dc00d5d31caba874d4a2599c72 |
|
BLAKE2b-256 | 70f07111d527db35c66a88014d5c7b6662c684378c8212a7a4ef4c367267b77a |
Close
Hashes for bootstrap.pytorch-0.0.10-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 09949f2059453b189631d95ea2f7c32db5453d9956847e6179f8eadbaa99d402 |
|
MD5 | d8113532ce668a26667ac2d47ed504db |
|
BLAKE2b-256 | b5701c3a0abe1a42873580f19357b3a68514fba2184cf23ea137d5299dae0e62 |