Build in the Lab
The lab was something we created to let you add a bit of training to our models. We created the lab primarily for our models which use prompt engineering. The idea behind the lab was, let's remove all of the training data that we have given the model and let you put what you want with 5 examples of both the inputs and the desired outputs. This gives you the opportunity to get a better output and something more tailored to your business without having to go down the full custom model process.
When you click, you will arrive at a page looking like this. The two options are to create a new model on the left and to view your creations on the right. By clicking on the left, you can then choose which model you'd like to add your data for and you will be given the opportunity to add 5 examples of both the inputs and the outputs. Once this process is complete you will be able to review your information before hitting submit.
Each model created within The Lab is manually approved by our team. This gives us a chance to check for any bugs or errors and determine if the examples fit the NLP provider and their requirements in terms of filters. Note - we try not to filter when necessary and make our positions clear here.
After submitting, you will be able to see the model in the 'View Experiments' tab on the right side. There will be a notice at the bottom of your model which will say 'Approved: Yes' or 'Approved: No'. Please note that new models can take up to 1 hour to be approved. Once it is approved, you will be able to use the model by clicking on it and start to see the improvements with the help of your own data.