1. Designing an ML Pipeline for Lane Detection
Design an ML pipeline for lane detection, covering all steps from data collection to final application. Address details such as where to collect data, how to store it, how to prune/clean the data, where to get the loss for training, and how to use the model for inference.
2. Optimizing LLM Fine-tuning Efficiency
Complete a coding task in a shared colab notebook for an LLM interview question. The interviewer describes a method to improve LLM fine-tuning efficiency (no prior knowledge required), and based on understanding, the task is to complete the coding task and correctly perform inference. The scenario involves fine-tuning an LLM to learn specific information about a car from its manual to make accurate predictions.
