Specs
- POSTAnalyze dataset and create spec
- GETGet all specs for the user's company
- GETGet a specific spec, optionally with deletion impact info
- PUTUpdate spec and create new version
- DELDelete a spec and clean up associated runs and S3 files
- GETGet all versions for a specific spec
- POSTCreate a new spec version
- GETGet a specific spec version
- PUTUpdate a specific spec version
- DELDelete a specific spec version
DataFramer - Evaluation
- GETGet chat history for an evaluation
- POSTChat with evaluation results
- GETGet evaluations for a specific run
- POSTCreate evaluation for a run
- GETGet evaluation details (with optional status polling for in-progress evaluations)
- DELDelete an evaluation
- GETGet chat history for an evaluation
- POSTChat with evaluation results
- POSTCreate evaluation for a run
Datasets
Files
Runs
- GETGet all historical runs
- GETGet a specific run
- PUTUpdate a run (only certain fields)
- DELDelete a run and clean up associated S3 files
- GETGet evaluations for a specific run
- GETGet all generated files from a run
- GETGet content of a generated file for viewing
- GETGenerate presigned URL for downloading a generated file
- GETGet all human labels for a run
- POSTCreate a new human label for a run
- GETRetrieve generated samples with pagination
- POSTRetrieve specific samples by indices
- GETGet run status
Red Teaming
- GETList all red team runs for the user's company
- POSTCreate a new red team run and submit to data_gen backend
- GETRetrieve a red team run
- DELDelete a red team run
- GETGet red team run status and sync with data_gen backend
- GETList all red team specs for the user's company
- POSTCreate a new red team spec
- GETRetrieve a red team spec
- DELDelete a red team spec
- PATCHUpdate a red team spec
- POSTCreate a new red teaming task.
- GETGet the status of a red teaming task.
DataFramer - Runs
DataFramer - Specs
JavaScript
Copy
import Dataframer from 'dataframer';
const client = new Dataframer({
apiKey: 'My API Key',
});
const humanLabel = await client.dataframer.runs.humanLabels.create('run_id');
console.log(humanLabel);Copy
{}Runs
Create a new human label for a run
Create a new human label for a run
POST
/
api
/
dataframer
/
runs
/
{run_id}
/
human-labels
JavaScript
Copy
import Dataframer from 'dataframer';
const client = new Dataframer({
apiKey: 'My API Key',
});
const humanLabel = await client.dataframer.runs.humanLabels.create('run_id');
console.log(humanLabel);Copy
{}⌘I