Self-directed AI training/tutoring project using TinyLlama on electrical/PLC and device protection concepts
Taught and trained a self-built seed Python code/model using an LLM base (TinyLlama/TinyLlama-V1.0) by providing interactive guidance on how electrical/PLC-related device behavior and protection rules should be handled. Used request/response style instruction to shape the model’s perspective so it could interpret and answer domain instructions in an expected way. Continued iterative development with potential plans to upgrade hardware for stronger processing. • Trained model via interactive “teaching” of electrical/automation perspectives and rule-based protection concepts. • Focused on domain understanding for devices (breaker/protection for home devices) and equal/opposite protection-from vs protection-to framing. • Worked with LLM-generated/assisted Python seed code and performed self-guided training. • Considered upgrading to more capable models/accessories if resources allow.