Run Qwen3.5 and Qwen3.6 dense models on AWS Trainium2 with vLLM-Neuron for online serving and offline batch inference. - View it on GitHub