https://github.com/s4m-mo/detametricssdk
The Python SDK for the Kaggle and Colab-compatible TensorBoard alternative.
https://github.com/s4m-mo/detametricssdk
alternative dashboard deta deta-space machine-learning python tensorboard tensorflow
Last synced: 30 days ago
JSON representation
The Python SDK for the Kaggle and Colab-compatible TensorBoard alternative.
- Host: GitHub
- URL: https://github.com/s4m-mo/detametricssdk
- Owner: s4m-mo
- License: mit
- Created: 2023-06-30T12:12:43.000Z (almost 3 years ago)
- Default Branch: main
- Last Pushed: 2024-06-23T12:54:10.000Z (almost 2 years ago)
- Last Synced: 2025-06-13T07:43:56.169Z (about 1 year ago)
- Topics: alternative, dashboard, deta, deta-space, machine-learning, python, tensorboard, tensorflow
- Language: Python
- Homepage: https://deta.space/discovery/@sam-the-programmer/detametrics
- Size: 14.6 KB
- Stars: 3
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
> This project is now archived, since Deta Space is now shut down. This will no longer work.
# DetaMetricsSDK
A Python SDK for interacting with my TensorBoard alternative that runs on [Deta Space](https://deta.space).
- [DetaMetricsSDK](#detametricssdk)
- [Installation](#installation)
- [Usage](#usage)
- [TensorFlow Callback](#tensorflow-callback)
- [Base API](#base-api)
---
## Installation
```powershell
pip install detametrics
```
## Usage
- Install in Deta Space from https://deta.space/discovery/@sam-the-programmer/detametrics
- All API classes need an URL ID passed as a parameter, which can be found in the app UI.
- They also need a Deta Space App API key, which can be gotten from...
- Click the 3 dots on the app on your Deta Space Canvas
- Click **"Keys"**
- Add an API Key, and input it in the box on the App UI and in your Python code as seen below
### TensorFlow Callback
Requires Tensorflow to be installed.
```python
from detametrics.tf import DetaMetricsTFCallback
# use in your keras model
model = ...
model.fit(
..., # other params here
callbacks=[DetaMetricsTFCallback("MY_URL_ID", "MY_API_KEY", log_batch=True)] # if log_batch is True, it logs every batch, else only every epoch
)
```
### Base API
Get your URL id from the UI on the [Deta Space app](https://deta.space/discovery/@sam-the-programmer/detametrics).
```python
from detametrics import DetaMetrics
metrics = DetaMetrics("MY_URL_ID", "MY_API_KEY")
metrics.clear() # WARNING: clears all past metrics
metrics.set("GraphName", "LineName", 0.1)
metrics.set("GraphName", "LineName", 123.45)
metrics.set("GraphName", "LineName", 78.9)
```