983987aab5
* add deep learning guide * seperate out oprimization, reference, and troubleshooting as standalone sections. * resolve lint errors * delete introduction to DL * correct syntax highlights and filename * remove out-of-date QAs * Renaming and cleanup * Spelling * Fixup TOC --------- Co-authored-by: Nara Prasetya <nara@streamhpc.com> Co-authored-by: Saad Rahim <44449863+saadrahim@users.noreply.github.com>
1.8 KiB
Troubleshooting
Q: What do I do if I get this error when trying to run PyTorch:
hipErrorNoBinaryForGPU: Unable to find code object for all current devices!
Ans: The error denotes that the installation of PyTorch and/or other dependencies or libraries do not support the current GPU.
Workaround:
To implement a workaround, follow these steps:
-
Confirm that the hardware supports the ROCm stack. Refer to the Hardware and Software Support document at https://docs.amd.com.
-
Determine the gfx target.
rocminfo | grep gfx -
Check if PyTorch is compiled with the correct gfx target.
TORCHDIR=$( dirname $( python3 -c 'import torch; print(torch.__file__)' ) ) roc-obj-ls -v $TORCHDIR/lib/libtorch_hip.so # check for gfx target
:::{note} Recompile PyTorch with the right gfx target if compiling from the source if the hardware is not supported. For wheels or Docker installation, contact ROCm support 1 . :::
Q: Why am I unable to access Docker or GPU in user accounts?
Ans: Ensure that the user is added to docker, video, and render Linux groups as described in the ROCm Installation Guide at https://docs.amd.com.
Q: Can I install PyTorch directly on bare metal?
Ans: Bare-metal installation of PyTorch is supported through wheels. Refer to Option 2: Install PyTorch Using Wheels Package in the section Installing PyTorch of this guide for more information.
Q: How do I profile PyTorch workloads?
Ans: Use the PyTorch Profiler to profile GPU kernels on ROCm.
-
AMD, "ROCm issues," [Online]. Available: https://github.com/RadeonOpenCompute/ROCm/issues ↩︎