I played with Floydhub for the time being
Introduction
Floydhub is a service like Heroku for Deep Learning that allows you to run programs using TensorFlow and Chainer on the Cloud. Features of Floydhub
--Chainer, TensorFlow, Keras, theano, etc. can be used --No credit card required to create an account
This time, I will try Get Started from creating an account for the time being.
Create an account
You can create an account for free. You don't need a credit card now. Sign Up from https://www.floydhub.com to move to the Welcome Page below.
Get Started Setup floyd-cli After creating an account, you will receive an email and authenticate. On the terminal, change to a working directory of your choice and then install floyd-cli.
pip install -U floyd-cli
Log in.
floyd login
Then
Authentication token page will now open in your browser. Continue? [Y/n]:
You will be asked, so enter y and proceed to open the Welcome page. Copy the authentication token at the bottom of the Welcome Page and paste it on your terminal and you should see Login Successful.
Run First Project Next, let's run Tensorflow.
Just follow the instructions on the Welcome Page.
$ git clone https://github.com/floydhub/tensorflow-examples.git
Cloning into 'tensorflow-examples'...
remote: Counting objects: 74, done.
remote: Total 74 (delta 0), reused 0 (delta 0), pack-reused 74
Unpacking objects: 100% (74/74), done.
$ cd tensorflow-examples/3_NeuralNetworks/
$ floyd init dynamic-rnn
Project "dynamic-rnn" initialized in current directory
$ floyd run "python dynamic_rnn.py"
After running floyd run "python dynamic_rnn.py", go to Experiments and you will see the program running.
By the way, it is Failed because the program name was mistakenly executed. ○ floyd run "python dynamic_rnn.py" × floyd run "python dynamic-rnn.py"
$ floyd logs -t <RUN_ID>
You can check the progress of learning with. The learning result seems to be in Data.
end.
Recommended Posts