Pioneer training jobs: lifecycle, metrics, and weights

Fine-tuning in Pioneer adapts a base model to your specific task and domain using your labeled dataset. You submit a training job through the API, Pioneer handles the compute, and you get back a trained model you can call for inference or download. The whole process is asynchronous — you start the job, then poll until it finishes. Pioneer supports three post-training algorithms, selected with the training_algorithm parameter: supervised fine-tuning (sft, the default) and the reinforcement-learning methods GRPO (grpo) and DPO (dpo). See the LLM fine-tuning guide for when to use each and how to format datasets.

Training job lifecycle

Every training job moves through a fixed sequence of states:

requested

Your job has been accepted and is queued for execution. Pioneer is allocating compute.

running

Training is actively executing. You can stream logs to monitor progress.

complete

Training finished successfully. Metrics (F1, precision, recall) are available on the job record, and checkpoints are ready to download.

A job may also end in failed (an error occurred during training) or stopped (you manually cancelled it with POST /felix/training-jobs/:id/stop).

Key parameters

Parameter	Required	Description
`model_name`	Yes	A name for your trained model, used to identify it in your account.
`base_model`	Yes	The model ID to fine-tune. Use a value from `GET /base-models` or a checkpoint UUID from a previous job.
`datasets`	Yes	An array of dataset objects: `[{"name": "your-dataset-name"}]`.
`training_type`	No	`"lora"` (default, parameter-efficient) or `"full"` (all weights). Decoder LLM training is LoRA-only.
`training_algorithm`	No	`"sft"` (default), `"grpo"`, or `"dpo"`. RL algorithms require an `rl_config`. See LLM fine-tuning.
`nr_epochs`	No	Number of training epochs. Defaults to a reasonable value for the base model.
`learning_rate`	No	Learning rate. Omit to use the default for the chosen base model.

base_model is required. Omitting it returns a 422 validation error. The value must be a model ID from GET /base-models or a checkpoint UUID — not a free-form string.

Starting a training job

curl -X POST https://api.pioneer.ai/felix/training-jobs \
  -H "X-API-Key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model_name": "my-ner-model",
    "base_model": "fastino/gliner2-base-v1",
    "datasets": [{"name": "my-ner-dataset"}],
    "training_type": "lora",
    "nr_epochs": 5,
    "learning_rate": 5e-5
  }'

The response returns immediately with a job ID and initial status:

{
  "id": "job_abc123",
  "status": "requested"
}

Save the id — you’ll use it to poll status, retrieve metrics, and run inference against your trained model.

Polling status and reading metrics

Poll the job endpoint until status is complete or failed:

curl https://api.pioneer.ai/felix/training-jobs/job_abc123 \
  -H "X-API-Key: YOUR_API_KEY"

When training completes, the response includes evaluation metrics measured on your dataset:

{
  "id": "job_abc123",
  "status": "complete",
  "metrics": {
    "f1": 0.94,
    "precision": 0.96,
    "recall": 0.92
  }
}

To stream logs while the job is running:

curl https://api.pioneer.ai/felix/training-jobs/job_abc123/logs \
  -H "X-API-Key: YOUR_API_KEY"

Stopping a job

If you need to cancel a running job:

curl -X POST https://api.pioneer.ai/felix/training-jobs/job_abc123/stop \
  -H "X-API-Key: YOUR_API_KEY"

The job status changes to stopped. Partial checkpoints saved before the stop may still be available.

Checkpoints and downloading weights

Pioneer saves checkpoints during training. You can list them at any point after the job starts:

curl https://api.pioneer.ai/felix/training-jobs/job_abc123/checkpoints \
  -H "X-API-Key: YOUR_API_KEY"

To download the final trained weights:

curl https://api.pioneer.ai/felix/training-jobs/job_abc123/download \
  -H "X-API-Key: YOUR_API_KEY" \
  --output my-model-weights.zip

You can also use a checkpoint UUID as the base_model value in a new training job to continue training from that checkpoint.

Training endpoints summary

Method	Endpoint	Description
`POST`	`/felix/training-jobs`	Start a new training job
`GET`	`/felix/training-jobs`	List all training jobs
`GET`	`/felix/training-jobs/:id`	Get job status and metrics
`GET`	`/felix/training-jobs/:id/logs`	Stream training logs
`GET`	`/felix/training-jobs/:id/checkpoints`	List saved checkpoints
`GET`	`/felix/training-jobs/:id/download`	Download trained weights
`POST`	`/felix/training-jobs/:id/stop`	Stop a running job
`DELETE`	`/felix/training-jobs/:id`	Delete a training job record

Get Started

Integrations

Core Concepts

API Reference

Guides

Account

Pioneer training jobs: lifecycle, metrics, and weights

Training job lifecycle

Key parameters

Starting a training job

Polling status and reading metrics

Stopping a job

Checkpoints and downloading weights

Training endpoints summary

​Training job lifecycle

​Key parameters

​Starting a training job

​Polling status and reading metrics

​Stopping a job

​Checkpoints and downloading weights

​Training endpoints summary

Training job lifecycle

Key parameters

Starting a training job

Polling status and reading metrics

Stopping a job

Checkpoints and downloading weights

Training endpoints summary