Skip to main content
Join the official 2019 Python Developers SurveyStart the survey!

A tool for training deep learning cross validation training.

Project description

PyPI version

Trainbench

Trainbench is a tool for quickly setup a (Deep Learning) cross-validation training session with configuration instantiation, learned weights, plots, etc, recording at each training epoch.

Compatible with Keras.

⚠️ This package is currently under active development!

Getting started

Install
pip install trainbench
Create a train session
# train.py

# name
name = 'experiment_01'

# parameters to cross validate against
parameters = {
    'fc_size': [256, 512, 1024]    
    # ...
}

def train(parameters):
    xyz = parameters['xyz']
    # ...
    
    
Running from the command line
trainbench .
Checkout your results
<name>/crosses/
               /001 
                   ├─ meta.yml 
                   ├─ history.pkl 
                   └─ weights/
                          ├─ 001.h5
                          └─ 002.h5
                             ...
In Jupyter Notebooks/Python script
from trainbench import Bench

parameters = {

}

def train_fn(parameters):
    pass
    
bench = Bench()
bench.train('experiment_xyz', train_fn, parameters)

Author notes

I'm wrote this tool for my own usage. Feel free to use it at your own will. If you would like to see any additional features / report existing issues please submit a pull request and/or open an issue.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for trainbench, version 0.0.2
Filename, size File type Python version Upload date Hashes
Filename, size trainbench-0.0.2.tar.gz (3.1 kB) File type Source Python version None Upload date Hashes View hashes

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN SignalFx SignalFx Supporter DigiCert DigiCert EV certificate StatusPage StatusPage Status page