Compare versions of your GenAI model (evaluate & fine-tune)

how it works - To evaluate your new version model we can compare new results against the original model responses and target global human feedback.

The customer needs to have an AWS S3 bucket and we will do the heavy lifting of comparing the models versions .This will be done via a secured CloudFront Distribution placed in front of the private bucket.

Upload your data - Customer proprietary data (up to 2K results of prompts for the 2 version models that will be compared )

Choose a comparison question - here are some examples but you can create your own a. Which text more positive/negative/helpful/creative/polite/informative etc.… Which image is more relasttic, pretty, accurate, etc... b. Which is better at answering the prompt requirements? c. Which is better Localized / native language / culture related? Define your crowd – locality/ global/ language - limited up to 5K humans reach. Get the results in tasq.ai platform. Price: $750 - one time comparison, including a detailed report and access to dashboard. $1,000 - Annual license, price per month – up to 2 comparison per month, including a detailed reports and access to dashboard