Upload Minimal Dataset
If you don’t have production traces but can provide a small set of labeled examples for your task, you can prepare a minimal structured dataset. You only need a few dozen high-quality examples that capture the essence of your task.
Your data directory needs the following files:
| File | Format | Required | Description |
|---|---|---|---|
job_description.json |
JSON | Yes | Task description defining what the model should do |
train.jsonl |
JSONL | Yes | 20+ labeled examples, each a messages conversation |
test.jsonl |
JSONL | Yes | Held-out evaluation set |
config.yaml |
YAML | Yes | Training hyperparameters |
unstructured.jsonl |
JSONL | No | Domain-relevant text for synthetic data generation |
distil model upload-data <model-id> --data ./dataFor detailed formatting and structure requirements per task type, refer to:
- Question Answering data preparation →
- Classification data preparation →
- Tool Calling data preparation →
- Multi-Turn Tool Calling data preparation →
- Open Book QA (RAG) data preparation →
- Closed-Book QA data preparation →
- Creating the configuration file →
Migrating from the previous question/answer column layout? See Legacy data format →.