inference 1

• Why I Built kaggle-vllm: Reproducible vLLM Inference on Dual NVIDIA T4 GPUs Sep 24, 2026