https://github.com/spro/sconce-python

https://github.com/spro/sconce-python

Last synced: 5 months ago
JSON representation

• Host: GitHub
• URL: https://github.com/spro/sconce-python
• Owner: spro
• Created: 2017-04-14T18:25:57.000Z (about 9 years ago)
• Default Branch: master
• Last Pushed: 2017-06-13T06:15:31.000Z (almost 9 years ago)
• Last Synced: 2026-01-18T17:54:09.413Z (5 months ago)
• Language: Python
• Size: 2.93 KB
• Stars: 16
• Watchers: 1
• Forks: 2
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# sconce

Sconce is a dashboard for monitoring and comparing data in real time. It was built to be an easy way to visualize the training progress of different machine learning models.

![](https://i.imgur.com/cb5ExqZ.png)

## Installation

Install with `pip`:

```

pip install sconce

```

## Usage

To start, create a sconce Job with a name and (optionally) parameters. Jobs with the same name will be displayed together, so you can see how different parameters affect results.

```python

import sconce

# Create a sconce job with a name and parameters

job = sconce.Job('my-neural-network', {

    'n_layers': 6,

    'hidden_size': 250,

    'learning_rate': 0.001

})

```

To plot data, call `job.record(x, y)`. Instead of posting every data point, Sconce will average the values and plot them every `plot_every` calls.

```python

# Record x, y values

for x in range(1000):

    y = train()

    job.record(x, y)

```

Then visit http://sconce.prontotype.us/jobs/my-neural-network to view your results in real time.

## Configuration

If you want more or less granularity, change `job.plot_every` and `job.print_every`:

```python

job.plot_every = 10

job.print_every = 100

```

## Private jobs

**TODO:** Register for an account and API key at http://sconce.prontotype.us/ to create private boards.

```python

sconce.login('spro', 'asdf1234uiop5768')

```