The title uses “building” instead of “writing” because some experiences take time to prepare. For example, taking a course takes at least half a year, and longer if there are prerequisites. Competing in a Kaggle machine learning competition can take three to six months, and it’s not always possible to achieve results that are good enough to put on your resume. Deliberately building your resume in advance will give you the best chance of getting interviews when…

Accidentally stumbled upon Y Combinator’s YouTube channel and found their course for startup founders. Took the opportunity over the holiday to watch it all. Despite working in startups for years, I learned a lot, and some insights even challenged my previous beliefs.

The models nowadays are becoming larger, especially language models. Search engines need to provide real-time results, and models on mobile devices have limited memory and computational power, requiring model lightweighting. ONNX, developed by Microsoft, is a cross-platform machine learning framework that can convert models from various frameworks (such as PyTorch, TensorFlow, etc.) into the ONNX format and perform lightweighting. This article takes a PyTorch model as an example, using ONNX to…

I recently finished reading the Elon Musk biography written by Walter Isaacson. Each chapter is a short story that allows us to understand Musk’s personality and how he accomplished all of his achievements: the commercialization of electric vehicles, world-leading self-driving technology, reusable rockets, satellite communications, and more. These are daunting feats, yet he made them happen. Musk is not a perfect person, nor does he need to be. We can try to learn from his working methods…

Recently, FIRE (Financial Independence and Retire Early) has become quite popular. People probably don’t like going to work. The source of income for not working is the 4% Rule, but this is actually a simplified result with dnagers.

Many economic studies discuss the relationship between income and happiness. This is not a formal economics blog, so I will not introduce them. Therefore, the research part will not be too detailed or accurate. I mainly want to talk about my own thoughts. For details of the studies, please see my references.

This year, NVIDIA opened a new metaverse department in Taiwan and is recruiting for various positions, including front-end engineers, and research scientists. I went for an interview for the deep learning engineer position.

In June of last year (2022), I took out a loan to buy stocks. For a while, the unrealized gains and losses were negative, but recently it has been a year and the stock market has performed well, so I would like to share my experience.

I have been a software engineer for nine years. Ten years ago, I still had one year of student life to do things that office workers don’t have much time to do. I’ve taken some detours in the past. If I had myself as a career mentor ten years ago, I would definitely have a more successful career now.

There seem to be many misunderstandings about Leetcode (or algorithm problems), such as the need to memorize all Leetcode problems before going for an interview, or that algorithm problems are intelligence tests (so problem-solving ability cannot be improved through practice). Many people criticize algorithm interviews as being disconnected from work, but we have no control over how each company interviews candidates, and discussing this has no help in passing the interview. Here I want to share…