R L
Get custom forward-backward operation
Retrieves the current status and result of a custom forward-backward operation.
Authorization
bearerAuth AuthorizationBearer <token>
In: header
Path Parameters
session_id*string
Training session ID
operation_id*string
Operation ID
Response Body
application/json
application/json
curl -X GET "https://example.com/rl/training-sessions/string/operations/custom-forward-backward/string"{ "id": "string", "status": "TRAINING_OPERATION_STATUS_UNSPECIFIED", "output": {}, "error": { "code": "TRAINING_OPERATION_ERROR_CODE_UNSPECIFIED", "message": "string" }}Custom forward-backward pass POST
Submits a forward-backward pass driven by externally computed gradients of the loss with respect to per-token log-probabilities.
Optimizer step POST
Submits an optimizer step operation that will asynchronously apply accumulated gradients to update model parameters. Does not make the updated parameters available for sampling; call `weights-sync` afterwards when you want subsequent samples to use the updated policy.