Artificial intelligence (AI) specialist Upstage (CEO Kim Seong-hun) announced on the 18th that it is going to offer a free course on developing a big language model (LLM) through the worldwide online education platform ‘DeepLearning.AI’.
Deep Learning AI is an academic platform created by Andrew Ng, a deep learning pioneer and certainly one of the ‘world’s top 4 AI scholars’, a professor at Stanford University within the US. You’ll be able to access Professor Ng’s specialized courses in addition to AI classes from big tech corporations similar to OpenAI, Google, Meta, and MS. Upstage is the primary domestic company to participate.
Particularly, it was reported that they’ll present an LLM pre-learning lecture planned with Professor Andrew Ng based on the know-how of developing their very own LLM ‘Solar’. Pre-learning is the technique of teaching AI models linguistic abilities based on massive text data. It’s also the core of LLM development with advanced natural language processing capabilities similar to sentence generation and context inference.
The lecture will likely be given by Upstage CEO Kim Seong-hun and CSO Park Eun-jung, who will personally cover the complete LLM pre-learning process from theoretical foundations to data set preparation, model training, and performance evaluation based on benchmark tests. The lecture will likely be conducted in English, and anyone with basic coding and machine learning knowledge can take the course totally free.
We plan to transfer practical know-how, similar to how one can reduce learning costs by making parameters lighter, specializing in the ‘DUS (Depth Expansion Scaling)’ methodology, which is Solra’s core technology, and how one can easily learn additional data based on pre-learning models released as open source.
Kim Sung-hoon, CEO of Upstage, said, “I’m delighted to be lecturing on the know-how of pre-LLM study with Professor Andrew Ng, a world authority on deep learning research.” He added, “Upstage will proceed to take the lead in enabling more people world wide to accumulate the newest AI knowledge based on the philosophy of ‘benefiting the world with AI.’”
Reporter Jang Se-min semim99@aitimes.com
