Streamlit is an open-source app framework to create web apps using python without any front-end experience. If you don’t have any python skills, no problem, we will share templates so you can create AI apps without any code.
One of the most incredible features we love about Streamlit is deployment. They offer free cloud hosting to deploy your apps. Yes, you read well; It is completely free. The only thing you need is a GitHub account. If you have never done coding, you probably are not familiar with Github, but again do not worry; it is pretty easy to set up an account, it is free, and we will explain everything in detail in this tutorial.
Let’s create a Github account, it is free. Go to https://github.com/, click on Sign up and follow all the steps.
Once you are logged in Github, go to the following repo template:
https://github.com/cogniflow/image-streamlit
Click on Use this template to create a new repo in your account which is basically a copy.
You can then choose the name of your repo and add some description. You can choose either Public or Private but take into account that Streamlit Cloud allows only 1 private app but unlimited public apps. Finally, click on “Create repository from template”
Now it is time to go to Streamlit Cloud and use your Github account to sign up.
Go to https://share.streamlit.io/signup, click on Continue with Github and agree to allow the permissions needed to connect Github with Streamlit Cloud.
In your Streamlit Cloud dashboard, click on “New app” and select “From existing repo”.
Then select the repo you created before (1):
Then click on Advanced settings… (2), set your Cogniflow API Key and your model url. You can get the values from the Cogniflow model integration tab:
Finally, click on “Deploy”! (3)
After 1 or 2 mins later you are going to see your app live. In this example I used a model to detect if an image is Coke or Pepsi. You can use this model from our Community here:
https://app.cogniflow.ai/experiment/99e3bea4-540a-4b31-aee4-64b2ee4081e9
This app works uploading an image and clicking on “Get prediction from AI” to see the results. Some examples below.
Voila! You have created your first Image Recognition app.
Please check out our other repos in Github to create apps to do Speech Recognition and Object Detection.
Happy training!
.svg){kind=small}