Skip to main content

A python package that lets you remotely monitor your deep learning training metrics through it's companion app.

Project description

TensorDash

TensorDash is an application that lets you remotely monitor your deep learning model's metrics and notifies you when your model training is completed or crashed.

Why Tensordash?

  1. Watch your model train in real-time
  2. Remotely get details on the training and validation metrics
  3. Get notified when your model has completed trainng or when it has crashed.
  4. Get detailed graphs on your model’s metrics.

Installation

Installing the Python Package

pip install tensor-dash

Installing the Android App

Install the android app from the play store.

How to use

  1. If you are using the app for he first time, sign up by clicking on the "create an account" button.
  2. After signing up, sign in to your account.

For Tensorflow and Keras

  1. In your python code, import Tensordash library There are multiple ways to Link your Account , The Following are displayed.

Specify only Model Name :

from tensordash.tensordash import Tensordash
histories = Tensordash(ModelName = '<YOUR_MODEL_NAME>')
Enter Email : ...........
Enter Tensordash Password : ********

Specify Model Name and Email address :

from tensordash.tensordash import Tensordash
histories = Tensordash(
	ModelName = '<YOUR_MODEL_NAME>',
	email = '<YOUR_EMAIL_ID>')
Enter Tensordash Password : ********

Specify Model Name, Email address and password :

from tensordash.tensordash import Tensordash
histories = Tensordash(
	ModelName = '<YOUR_MODEL_NAME>',
	email = '<YOUR_EMAIL_ID>', 
	password = '<YOUR PASSWORD>')

In the app, if you have multiple models you would be able to identify your model by YOUR_MODEL_NAME, so this name has to be unique.

  1. Now you can monitor your model values and status using crash analysis. Simply use a try-catch block as shown below.
try:
    model.fit(
	X_train, 
	y_train, 
	epochs = epochs, 
	validation_data = validation_data, 
	batch_size = batch_size, 
	callbacks = [histories])

except:
    histories.sendCrash()

OR

Alternatively, if you do not want to use crash analysis then you can just monitor by just adding histories object to callback

model.fit(
	X_train, 
	y_train, 
	epochs = epochs, 
	validation_data = validation_data, 
	batch_size = batch_size, 
	callbacks = [histories])

For Fast.ai

  1. In your python code, import Tensordash library There are multiple ways to Link your Account , The Following are displayed.

Specify only Model Name :

from tensordash.fastdash import Fastdash

learn = cnn_learner(data, models.resnet18, metrics=accuracy)
my_cb = Fastdash(learn, ModelName = '<YOUR_MODEL_NAME>')
Enter Email : ...........
Enter Tensordash Password : ********

Specify Model Name and Email address :

from tensordash.fastdash import Fastdash
my_cb = Fastdash(
	ModelName = '<YOUR_MODEL_NAME>',
	email = '<YOUR_EMAIL_ID>')
Enter Tensordash Password : ********

Specify Model Name, Email address and password :

from tensordash.fastdash import Fastdash
my_cb = Tensordash(
	ModelName = '<YOUR_MODEL_NAME>',
	email = '<YOUR_EMAIL_ID>', 
	password = '<YOUR PASSWORD>')
  1. Now you can monitor your model values and status using crash analysis. Simply use a try-catch block as shown below.
try:
    learn.fit(epochs, learning_rate, callbacks = my_cb)
except:
    my_cb.sendCrash()

OR

Alternatively, if you do not want to use crash analysis then you can just monitor by just adding my_cb object to callback

learn.fit(epochs, learning_rate, callbacks = my_cb)

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 tensor-dash, version 1.5
Filename, size File type Python version Upload date Hashes
Filename, size tensor_dash-1.5-py3-none-any.whl (6.4 kB) File type Wheel Python version py3 Upload date Hashes View hashes
Filename, size tensor-dash-1.5.tar.gz (4.4 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 DigiCert DigiCert EV certificate StatusPage StatusPage Status page