April 22, 2022
5 min read

Create an artificial intelligence model using a CSV file


Using AI to generate content can help you with your writing in a number of ways. For example, you can create a custom AI that will write content for you based on your specific needs. This includes the topics that it covers, the tone of voice, and the style of writing. You will also be able to set up how often it writes content for you and what kind of information it should include in its articles.

If you are not looking for an AI that is customized, there are also many off-the-shelf options available. These services allow you to input your topic and then they will generate relevant articles for you with just a few clicks. However, Contentable.ai allows you to create content for your specific use case from your own data with a click of a button.

To create a model using using a csv file, follow these steps:

Create an account

The first thing to do is to create an account. In order to do this, you will need to enter your email address, and a password. Or simply just sign in with your Gmail account. You can change these later but it is important that you have them right now so that you can get started.

Prepare your CSV file

The next step is to prepare a csv file for you specific use case. The csv should contain at leat 100 rows for the model to be somehow accurate. The more data you provide the better the model will perform. For an example, let's say you would like to create product description for your products based on the name of the product. Your csv file should look something like this

Train your AI model

Next, head into your contentable dashboard and nagivate to the Contents section. Select the CSV file from your local disk.

After that, the schema of your file would appear on the table above. You can use the table to select the inputs and outputs to your model. For example if you would like to train a model that create the product description based on product name, you check the corresponding boxes as follows:

Once you are ready move to the Datasets section. In this section, you would be able to modify a few configuration for your model.

  • Number of samples to query: This is the number of samples we initialy take from your dataset
  • Number of samples: We limit the number of samples to this parameter after shuffling them
  • Exclude null input/output values: This makes sure that your model is not trained on null input/output values

Here is an example to help understand these parameters. With the default parameters above, if your dataset has 1000 entries, here is the process we follow to prepare your dataset. We first filter the samples to not have any not null values. We only consider the first 200 samples from the result and then randomly sample 100 samples from those for the model.

After uploading the dataset, you will be able to train the model on your dataset with a click of a button. Please contact us for any assitance.

Written by

99.9% teams love Collab. Not convinced you’re one?We love a challenge.