- Base-model discovery and dataset management
- Managed
configjob creation withPOST /v1/training/jobsfor GLiNER encoders - Interactive jobs for decoder training:
POST /v1/training/jobs/create-model, then forward-backward passes, optimizer steps, checkpoints, and sampling under/v1/training/jobs/{job_id} - Canonical lifecycle operations under
/v1/training/jobsfor bothconfigandinteractivejobs
Training lifecycle
Choose a model, create a job, poll it, and work with the resulting artifact.
Encoder Quickstart
Fine-tune a GLiNER encoder with managed training.
Decoder Quickstart
Train Nemotron or Qwen with interactive training.
Upload Dataset
Prepare JSONL, upload it, process it, and confirm a ready version.

