> For the complete documentation index, see [llms.txt](https://docs.datumo.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.datumo.com/documentation/documentation-en/evaluation/how-to-use/3.-benchmark-set-management.md).
# 3. Benchmark set management
## What is a Benchmark Set?
> A Benchmark Set is a collection of reference data for evaluating tasks, and it is an evaluation dataset composed of multiple data points with evaluation purposes and representativeness.
***
## Benchmark Set Upload
You can build a Benchmark Set by uploading the evaluation data you have.
### 1. Upload Dataset
`Upload File` to analyze.
Drag the Dataset file to upload, or click the upload area to select a file.\
`csv` or `xlsx` Files in this format can be uploaded.
When uploading a file `Next` the button becomes enabled.
The column names of the uploaded file are displayed at the top.
### 2. Edit Data
`Edit` You can edit by pressing the button.
Select a specific cell in the part you want to edit and make changes.
If there are any changes `Save` the button becomes enabled.
### 3. Delete Data
`Edit` You can delete the data you want to remove by selecting the button.
Select the data you want to delete.\
(When selecting the ID column, it changes to a checkbox.)
When data is selected `Delete Data` the button becomes enabled.
### 4. Add Data
`Edit` You can add data by pressing the button.
`+Add Data` Select
`+Add Data` When selected, a new column is added to the very last column.
After entering the additional data, `Save` When clicked, the added column is saved.
***
## AI-based data generation and editing
You can use AI to generate Benchmark Set data or modify existing data.
### 1. Select a generation template
`AI Tool` Click to select the template you want to use.
When there is no data `Augmenter` Everything except is disabled.
### 2. Enter generation conditions
Enter the column to generate and the required information.
### 3. Add Draft and generate data
If the test results are not satisfactory, add a Draft to improve the instructions, and if the results are satisfactory, generate the data.
`+ Create New Draft` 를 클릭하여 Draft를 추가 할 수 있습니다.
`+ Create New Draft` You can add a Draft by clicking.
### 4. Generate data
Data is generated.
***
### 🔗 Related documents
* E-1. Creating and managing Criteria — How to configure Criteria, Subgroup, and Task
* E-2. Task Management — How to manage Task basic information, Schema, Metric, and Benchmark Set
* E-4. Adding and managing Metrics — How to create Metrics, map Keys, and set up Preview
* E-5. Run Eval — How to run an evaluation Run and check history
* E-6. Eval Result — How to check evaluation results and manage Errors and Retries
* E-7. Dashboard — How to visualize and analyze evaluation results