💫 Preliminary Schedule 24/25
Subject to changeSpeech2text
In the first block of the semester, we will focus on automatic speech recognition systems (ASR). In other words, systems that can map speech in form an audio signal to text.
Graph Neural Networks
In this block, we venture into graph neural networks and their application in football tactics.
- W43 Everything is Connected / Deepmind / 2023: Graph Neural Networks
- W44 GAN/ UofCambridge / 2018: Graph Attention Networks
- W45 Tactic AI / Deepmind / 2023: an AI assistant for football tactics
Training our own GPT
We have now read many interesting research papers. Now, it is time to get our hands dirty. We will make use of Andrej Karpathy’s open source implementation of GPT-2 to train our small language model. We will go through the whole setup from collecting and preparing the data to logging the training results. First week, we will spend on setting up the baseline model. The following week, we will try to add some tweaks to it to see if we can improve its performance.
- W46 nanoGPT baseline implementation
- W47 can we make our nanoGPT better?
Industry talks
- W48 Veo Technologies
- W49 TBA