Fix CalledProcessError

✅ Solution

The "CalledProcessError" in vllm often arises from issues during JIT compilation of custom CUDA kernels (e.g., FlashInfer), usually due to missing dependencies or build environment problems. Ensure that `nvcc`, `ninja`, and `g++` are correctly installed and accessible in your system's PATH. Verify proper CUDA driver installation and set `TORCH_CUDA_ARCH_LIST` to match your GPU architecture before retrying.