Skip to main content
Datasets are the foundation of every fine-tuning workflow on Pioneer. You create them through the synthetic data generation API or by uploading your own files, and Pioneer stores them at no cost. Use these endpoints to list your available datasets, inspect their versions, and delete datasets you no longer need.
Datasets are created via POST /generate or by uploading files directly. To start training, a dataset must be in the ready state.

​
List all datasets

GET /felix/datasets Returns all datasets associated with your account.
Response
​
boolean
true on success.
​
number
Total number of datasets returned.
​
object[]
Array of dataset objects.

​
Get dataset versions and details

GET /felix/datasets/:name Returns version history and metadata for a specific dataset. Path parameters
​
string
required
The dataset name exactly as it appears in your account.
Response
​
boolean
true on success.
​
number
Number of versions returned.
​
object[]
List of dataset versions, most recent first.

​
Delete a dataset

DELETE /felix/datasets/:name Permanently deletes a dataset and all its versions. This action cannot be undone.
Deleting a dataset does not affect training jobs that have already completed using it, but you will not be able to retrain or run new evaluations with the deleted dataset.
Path parameters
​
string
required
The name of the dataset to delete.
Response Returns 200 with an empty body on success.