The teacher panel — called the Centro de mando del Consejo in the interface — is the operational hub for running an MLflow Jedi classroom session. Unlike the student portal, which presents coding challenges and missions to teams, the teacher panel gives you a bird’s-eye view of every team’s progress without ever executing a single line of student code. From a single screen you can create sessions, share access codes, send announcements, unblock stuck teams, and export complete results when the activity ends.Documentation Index
Fetch the complete documentation index at: https://mintlify.com/elenacarino-max/Pildora4_ext_StarWars/llms.txt
Use this file to discover all available pages before exploring further.
Accessing the panel
Navigate to Profesorado in the Streamlit sidebar. You will be presented with a password prompt. Enter the teacher password to open the panel.The password is configured via the
MLFLOW_JEDI_TEACHER_PASSWORD environment variable and is never stored in the UI. The application ships with a default of 1234 for local development — always override this in a classroom or production deployment.Before the activity
Follow these steps before students enter the room:Start the application
Launch the app with Docker Compose (
docker compose up) or directly with streamlit run app.py. Confirm the Streamlit UI is reachable and MLflow tracking is running.Log into Profesorado
Open the sidebar, select Profesorado, and enter the teacher password. The panel will display the Centro de mando del Consejo header.
Create a session and obtain the JEDI code
Click Crear nueva sesión. The application generates a unique code such as
JEDI-4821 and selects six questions at random — one per core MLflow concept (Experiment, Run, Parameters, Metrics, Artifacts, Model). The session opens immediately for teams to join.Organize teams of three
Each group of three students takes the roles Navegante, Escriba, and Auditor. Together they choose one of the ten available team names from the dropdown in the student portal. Each team name may only be used once per session.
During the activity
Once teams are active, the panel provides continuous visibility and a set of targeted interventions: Live team tracking table — the table updates in real time and shows, for every team:| Column | What it represents |
|---|---|
| Equipo | Team name |
| Estado | Current phase (chambers, missions, comparator, defense, or completed) |
| Cámara | Chamber progress out of 6 (e.g. 3/6) |
| Misiones | Missions completed out of 3 (e.g. 1/3) |
| Intentos | Total code submission attempts across all challenges |
| Pistas | Number of hints used (each costs 5 points) |
| Puntos | Current score |
| Holocrón | Title of the assigned holocron challenge, or — if not yet assigned |
chamber_index=0 with a score of 0. Requires checking the confirmation box first.
Reassign a holocron — change which of the seven holocron defense challenges a team receives. Useful when a team’s assigned challenge does not fit the focus of today’s session.
Key message
The technical concept underpinning the entire activity is captured in one line from the teacher guide:Scikit-learn entrena. MLflow registra. El Consejo compara y decide.The app consistently shows narrative Star Wars terms alongside their professional MLflow equivalents — for example, Estrategia maps to Parameters, Informe to Metrics, and Holocrón to Artifact — so students absorb both vocabularies simultaneously.
Continue to
Sessions
Create JEDI-XXXX codes, control session status, manage multiple sessions, and send global hints.
Interventions & Export
Unlock chambers, reset teams, reassign holocrons, and export results as JSON or CSV.
Student code is never executed by the teacher panel. The teacher has no need to review or run any Python submitted by teams — all evaluation is handled automatically by the application’s validation layer.