Create Reinforcement Fine-tuning Job
Path parameters
The Account Id
Query parameters
ID of the reinforcement fine-tuning job, a random UUID will be generated if not specified.
Request body
The completed time for the reinforcement fine-tuning job.
The name of the dataset used for training.
The name of a separate dataset to use for evaluation.
Whether to auto-carve the dataset for eval.
JobState represents the state an asynchronous job can be in.
- JOB_STATE_PAUSED: Job is paused, typically due to account suspension or manual intervention.
The email address of the user who initiated this fine-tuning job.
The evaluator resource name to use for RLOR fine-tuning job.
The output dataset's aggregated stats for the evaluation job.
Data chunking for rollout, default size 200, enabled when dataset > 300. Valid range is 1-10,000.
The number of nodes to use for the fine-tuning job. If not specified, the default is 1.
The signed URL for the trainer logs file (stdout/stderr). Only populated if the account has trainer log reading enabled.
Accelerator seconds used by the job, keyed by accelerator type (e.g., "NVIDIA_H100_80GB"). Updated when job completes or is cancelled.
Maximum number of concurrent rollouts during the RFT job.
Maximum number of concurrent evaluations during the RFT job.
Response
A successful response.
The completed time for the reinforcement fine-tuning job.
The name of the dataset used for training.
The name of a separate dataset to use for evaluation.
Whether to auto-carve the dataset for eval.
JobState represents the state an asynchronous job can be in.
- JOB_STATE_PAUSED: Job is paused, typically due to account suspension or manual intervention.
The email address of the user who initiated this fine-tuning job.
The evaluator resource name to use for RLOR fine-tuning job.
The output dataset's aggregated stats for the evaluation job.
Data chunking for rollout, default size 200, enabled when dataset > 300. Valid range is 1-10,000.
The number of nodes to use for the fine-tuning job. If not specified, the default is 1.
The signed URL for the trainer logs file (stdout/stderr). Only populated if the account has trainer log reading enabled.
Accelerator seconds used by the job, keyed by accelerator type (e.g., "NVIDIA_H100_80GB"). Updated when job completes or is cancelled.
Maximum number of concurrent rollouts during the RFT job.
Maximum number of concurrent evaluations during the RFT job.