VerilogEval: Evaluating Large Language Models LLMs for Verilog Code Generation research.nvidia.com 2 points by teleforce 17 hours ago