SCOS 285 · GPU Inference Runtime Adapter

Status: CHECK

Runtime mode: CPU_FALLBACK_WITH_GPU_PROVIDER_PENDING

Selected backend: onnxruntime_cpu

GPU hardware ready: true

True GPU inference provider active: false

Local endpoint: http://127.0.0.1:5085

Provider Truth

Provider	State
torch	{"present":true,"cudaAvailable":false,"cudaDeviceCount":0,"cudaVersion":null,"version":"2.9.1+cpu","devices":[]}
cupy	{"present":false,"cudaAvailable":false,"version":null}
onnxruntime	{"present":true,"availableProviders":["AzureExecutionProvider","CPUExecutionProvider"],"version":"1.23.2"}
numpy	{"present":true,"version":"2.2.6"}
scipy	{"present":true,"version":"1.15.3"}