4 ms·SpeedLLM: An FPGA Co-Design of Large Language Model Inference Accelerator2 points by juanviera23 1y ago