Skip to main content
The Dashboard is the control center for your Promptify workspace. It shows:
  • Pipeline status – current fine‑tune job and progress
  • Assistant state – whether an assistant is configured and which model it’s using
  • Model details – base vs fine‑tuned model currently powering the app and API

​
Fine‑tuning job card

The main card on the Dashboard represents your current fine‑tuning job.
  • States: GENERATING_DATA, UPLOADING, QUEUED, RUNNING, SUCCEEDED, FAILED
  • Progress bar: percent completion of the current job
  • Logs: streaming log lines from dataset generation and the OpenAI fine‑tuning API
When no job is running you’ll see “No Active Jobs” with guidance to either:
  • Complete Onboarding to set up an assistant, or
  • Start a new fine‑tuning job once an assistant exists.

​
Starting a job

From the Fine‑Tuning Job section:
  1. Confirm you’ve created an assistant via Onboarding.
  2. Click Start Training.
  3. Promptify will:
    • Generate a JSONL dataset from your assistant spec
    • Upload it to OpenAI
    • Kick off a fine‑tune job
    • Poll the job via /api/fine-tune/status
Once a job completes, the Model stat card updates to show whether a custom fine‑tuned model is active. The Chat API will automatically start using this new model.