Skip to main content

University AI Course Lab Platform: From Teaching to Practice in One Stop

📌 Scenario Overview​

With the advancement of large language model (LLM) technology, AI courses in universities urgently need to move beyond "theoretical-only" teaching methods and integrate hands-on practice with theoretical learning. Transense Community provides a complete teaching lab platform for universities and AI education institutions, supporting model uploads, fine-tuning training, application deployment, and online inference—covering the entire AI project development process.

Instructors can flexibly organize students to participate in experiments and assign project tasks, while students can gain practical experience in model training, inference deployment, and AI application development, significantly enhancing the effectiveness of course practice.

🧭 Step-by-Step Guide​

1. Instructors Create Lab Groups and Invite Students​

  • Instructors create new lab groups on the platform, and students register accounts to join the groups. Organizational permissions ensure resource security and orderly group management.
    alt text
  • Instructors upload relevant models (e.g., open-source LLMs like LLaMA and Mistral) and teaching datasets (supporting formats like jsonl/txt) to prepare for student training.
  • Transense Community provides multiple model upload methods and dataset upload methods.

3. Students Complete AI Practice Tasks Online​

After joining a group, students can perform the following typical AI operations on the platform:

Model Fine-Tuning​

  • Select course models and datasets, configure parameters online, and launch fine-tuning tasks to experience the complete model fine-tuning process.
  • Multiple fine-tuning frameworks are available. Refer to the Fine-Tuning Framework Guide for scenario-specific options.

Create AI Application Demos​

Build interactive AI applications in the app space, such as Q&A assistants or text generation tools, for quick functional demonstrations.

Configure Model Inference APIs​

Deploy trained models as online inference APIs for front-end applications or other services, enabling end-to-end integration.

✨ Key Outcomes​

  • For Instructors: A complete AI lab teaching management platform with reusable models and datasets, reducing teaching and maintenance costs.
  • For Students: Master real-world LLM development workflows, improving practical skills in model training, application development, and deployment.
  • For Projects: Visible and callable teaching outcomes, facilitating demonstrations, grading, and iterative improvements.
  • For Universities: Private deployment of the lab platform allows institutions with computing resources to better serve teaching needs.