self – AV

Last Friday was my last day at megacorp. I am thankful for the four years at my first job as a Software Engineer. Today I have an incredible opportunity to dedicate myself full time in advancing my personal projects and going deep into GenAI while living in SF. Right place, right time.

Though I already have some beginner experience such as building my own home server for working with LLM and Stable Diffusion. I want to refresh my ML foundation and hold myself accountable to a structured plan over the next 8 weeks of runway. So long as I achieve my laid out goals, I can continue building my projects. I have debated what is the best way of getting up to speed with the GenAI community fast. I want to be competent enough to contribute to the massive energy happening right now as I felt going to Ollama‘s meetup last Monday. Formally rebuild my foundation or just learn as I build (honestly leaning to the latter), but I will do both since it seems I already started at the end with high interest areas and now moving backwards to foundations to deepen my understanding.

PS: I am applying for jobs.

I asked ChatGPT to layout an 8 week plan for “how can i learn everything i can learn as a beginner about generative ai and become an expert at it”.

Week 1-2: Fundamentals of AI and Machine Learning
Objective: Understand the basics of AI, machine learning, and neural networks.
• Complete an introductory course on AI and ML (Coursera’s “Machine Learning” by Andrew Ng is highly recommended).
• Read chapters 1-3 of “Deep Learning” by Ian Goodfellow.
Week 3-4: Deep Learning Essentials
Objective: Get familiar with deep learning, focusing on CNNs and RNNs.
•Complete a deep learning course (fast.ai offers a practical, hands-on course).
• https://course.fast.ai/
• https://lnkd.in/ghhSsD9u
• Implement basic projects using TensorFlow or PyTorch (e.g., image classifier with CNN, text generation with RNN).
Week 5-6: Introduction to Generative AI
Objective: Understand and implement simple generative models.
• Study GANs and VAEs through tutorials and implement simple models.
• Explore tutorials on transformers and attempt to fine-tune a pre-trained model for a text generation task.
Week 7: Advanced Generative AI Models
Objective: Dive deeper into advanced generative models and applications.
• Read research papers or summaries on GPT and DALL·E.
• Experiment with more complex projects, such as using GANs for more sophisticated image generation or exploring text-to-image generation models.
Week 8: Real-world Applications and Exploration
Objective: Apply knowledge to a real-world problem or creative project.
• Identify a project that interests you, such as creating a chatbot, generating art, or synthesizing data.
• Begin building your project

Good plan?

Either way I’m excited for this new chapter. It’s been so much fun following the real time advancement and sharing of ideas.

I want to be a creator

It seems like there are many people creating their own custom LLM and AI art today, about 10 months after ChatGPT-3.0 surprised and changed the world. It felt like magic to literally the whole world and spur such great interest, innovation, investments, competition when the world was in a vulnerable state with high inflation, layoffs, and economic turmoil. This magic gave hope.

The tipping point where I felt comfortable enough to embark on my journey from observer to creator was being introduced to https://www.reddit.com/r/LocalLLaMA/ and seeing so many people there casually discussing and practicing LLM. Beautiful AI art has also popped up and taken a lot of space from traditional art on deviantart.com and IG.

So thank you to all the creators out there putting your work up, you’ve inspired me by making it seem doable and accessible. If so many people can do it, why can’t I?

This marks the start of my journey to learn and create LLMs and AI art.