Fine-tuning status

Bring your own fine-tuned model ID and we pass it straight through to Azure. Native passthrough is on the roadmap.

Today

LOCK accepts a fine-tuned model ID you provisioned in your Azure OpenAI resource. We forward it verbatim in the deployment manifest. No training, no validation — just passthrough.

"fine_tuned_model": "gpt-4o-mini-ft-abc123"

Roadmap

  • 1Native fine-tuning jobs — trigger training runs directly from LOCK, no Azure portal needed.
  • 2Dataset management — upload, validate, and version your JSONL files inside the dashboard.
  • 3Evaluation suite — side-by-side comparisons between base and fine-tuned models on your eval sets.

Have a fine-tuned model ready? Deploy it now →

LOCK — fine-tuning passthrough for Azure OpenAI