Data Normalization

Data processing and normalization tasks evaluate structured output from messy prose and different structures. This capability is essential for catalogue and product pipelines where data needs to be extracted from unstructured text and formatted consistently. Models must understand the desired output format and extract relevant information accurately.

Duration

Input Tokens

618

Output Tokens

Cost

$0.00

Context

Input

Analyze the following instructions and extract all appliances mentioned. Order is not important

Expected output

{
  "appliances": [
    "vacuum cleaner",
    "microwave",
    "kettle"
  ]
}

Model output

{
  "appliances": [
    "microwave",
    "vacuum cleaner",
    "kettle"
  ]
}