Bioinfo Lab

Bioinformatics Laboratory at BiUH.

Core Values: Respect, Reflection, Communication, Commitment.

Main Page   |   Util & Src   |   Contact


Loading......

Password


Let’s make a movie ！！！


Basic Goal

  • Keywords: Python, LLM, API, Prompt, Movie

  • Learn to use Python with the APIs (Application Programming Interfaces) of Large Language Models (LLMs) to systematically generate coherent prompts (movie scripts) for movie clips.

  • Create a Python Movie-Scriptwriter program which can automatically generate coherent prompts (movie scripts) for movie clips.

  • Generate a 1-minute coherent mini-movie consisting of six 10-second movie clips.


Hint

  • There is an AI assistant on our server. The AI assistant is based on the API of Kimi.

  • The location of AI assistant on our linux server: “/home/toolkit/tools/ai/ai.py”. Use it, understand the code

  • Use Local Model (ComfyUI) or online Model Dou Bao (豆包) or Ji Meng (即梦) to generate the 5s or 10s movie clips from your movie scripts.

  • Use Jian Ying (剪映) to combine all movie clips.


Requirements

  • Your program must run on Linux.

  • Your code must be modified based on the “/home/toolkit/tools/ai/ai.py”.

  • Prompts for the movie’s six clips must be generated iteratively: each clip’s prompt is based on the previous one, implemented with Python loop syntax.

  • You should pack useful code block into functions.

  • User input is limited to a simple description under 20 English words.


Final Presentation

30 minutes presentation + 5 minutes QA

  • Play the 1-minute mini-movie demo

  • Live demo: Generate 6 coherent movie-scripts from input within 20 words.

  • Code: Explain the code implementation details of the Movie-Scriptwriter program in detail, with a focus on how to use functions and loop syntax.

  • Movie making: Show all the efforts you have done when making the “1-minute mini-movie demo”


Team Arrangement

  1. In total, we have 5 teams (5-7 members per team).

  2. We have multiple rounds of member selection.

  3. In each round, each Team Lead will introduce him/her-self.

  4. Each Team Lead will provide one offer to one or more members (you can reject the offer).

  5. For the final project, the maximum grade is 30 (final presentation). Team Lead can get 3 bonus points (10%).

  6. If more than 5 people want to be the Team Lead, let’s vote for it. If less than 5 people want to be the lead, I will randomly assign the remaining Team Leads.