YOLO 11 in VOXL
-
Hi team, I make some tests in VOXL2 hardware with YOLO11 using the voxl-tflite-server code, actually I am using YOLOv8 working in the GPU, but for YOLO11 the GPU is not allowing to use the YOLO11 model with this error:
Feb 14 19:53:17 m0104 bash[6953]: oksi-tflite-nats: /local/mnt/workspace/lnxbuild/project/trees_in_use/free_tree_dir/checkout/build-qti-distro-ubuntu-fullstack-perf/tmp-glibc/work/qrb5165_rb5-oe-linux/adreno/1.0-r0/adreno200/llvm/lib/Target/Oxili/QGPUISelLowering.cpp:503: llvm::SDValue llvm::QGPUTargetLowering::LowerKernelFormalArguments(llvm::SDValue, llvm::CallingConv::ID, bool, const llvm::SmallVectorImplllvm::ISD::InputArg&, llvm::DebugLoc, llvm::SelectionDAG&, llvm::SmallVectorImplllvm::SDValue&) const: Assertion `false && "Exceeded maximum number of write image id's"' failed.
Feb 14 19:53:17 m0104 bash[6953]: ERROR: Failed to build program executable - Out of host memoryPass
Feb 14 19:53:17 m0104 bash[6953]: ERROR: Falling back to OpenGL
Feb 14 19:53:17 m0104 bash[6953]: ERROR: TfLiteGpuDelegate Init: OpenGL-based API disabled
Feb 14 19:53:17 m0104 bash[6953]: INFO: Created 0 GPU delegate kernels.
Feb 14 19:53:17 m0104 bash[6953]: ERROR: TfLiteGpuDelegate Prepare: delegate is not initialized
Feb 14 19:53:17 m0104 bash[6953]: ERROR: Node number 802 (TfLiteGpuDelegateV2) failed to prepare.
Feb 14 19:53:17 m0104 bash[6953]: ERROR: Restored original execution plan after delegate application failure.
Feb 14 19:53:17 m0104 bash[6953]: Failed to apply GPU delegate with errorcode: 2Do you have any thoughts on how to fix this?
Best regards.
Jesús Deloya.
Hello! It looks like you're interested in this conversation, but you don't have an account yet.
Getting fed up of having to scroll through the same posts each visit? When you register for an account, you'll always come back to exactly where you were before, and choose to be notified of new replies (either via email, or push notification). You'll also be able to save bookmarks and upvote posts to show your appreciation to other community members.
With your input, this post could be even better 💗
Register Login