The knowledge needed is going to vary on what type of team you're working with and the role. If you're only focused on software side of making an ML pipeline you may not need to know as much about ML models, if someone else is taking care of model selection, evaluation, etc,
Some models and libraries that are very common and I'd suggest to take a look at if you're newer to machine learning would be:
• The "traditional ml" models in
scikit learn
• Deep learning model in
PyTorch
• Use a LLM from
hugging face