Create Dataset
POST/api/2/auto_eval/dataset/create
Create a new Dataset. Use UploadDatasetFile to upload files to the dataset.
Request
- application/json
Body
Human-readable name for the dataset, if one exists.
Human-readable description of the dataset, if one exists.
PseudoLabel job fields.
Responses
- 200
Successful operation
- application/json
- Schema
- Example (from schema)
Schema
Array [
]
dataset
object
required
A Dataset in the most basic sense: metadata and ownership, but nothing tied to its data.
The ID of the dataset.
Human-readable name for the dataset, if one exists.
Human-readable description of the dataset, if one exists.
The ID of the user who owns the dataset.
columns
object[]
required
The ID of the dataset file.
Index of the column within the dataset file.
The literal name for the column.
labelState
object
The state of the latest labeling job for the dataset
The status of the latest general pseudo-labeling job for the dataset
aka user general instructions
if the labeling status is error, this field may contain an error message
{
"dataset": {
"id": "string",
"createdAt": "2024-07-29T15:51:28.071Z",
"updatedAt": "2024-07-29T15:51:28.071Z",
"name": "string",
"description": "string",
"ownerUserId": "string",
"numRows": 0,
"numCols": 0,
"initializationStatus": "string",
"initializationError": "string",
"columns": [
{
"id": "string",
"createdAt": "2024-07-29T15:51:28.071Z",
"updatedAt": "2024-07-29T15:51:28.071Z",
"index": 0,
"literalName": "string",
"dtype": "string"
}
],
"labelState": {
"labelingStatus": "string",
"promptTemplate": "string",
"error": "string"
}
}
}