uvok on Nostr: Imagine trying to load a #TensorFlow model and all you get is a scary traceback and ...
Imagine trying to load a #TensorFlow model and all you get is a scary traceback and
'cuLaunchKernel...)' failed with 'CUDA_ERROR_INVALID_HANDLE'
Thanks for nothing, I guess. The search results don't help me.
(TF installed via mamba / conda-forge)
or rather, CUDA_ERROR_NO_BINARY_FOR_GPU ?
Published at
2024-10-03 13:56:06Event JSON
{
"id": "a665bc6aa41b15737815bcd859cd63e4c87317e9adc50fd73e9d01567dc68385",
"pubkey": "b15e52e9623da51cfe460020e7da27a086de784731ae755e0bb94e9f1c113cb2",
"created_at": 1727963766,
"kind": 1,
"tags": [
[
"t",
"tensorflow"
],
[
"proxy",
"https://furry.engineer/users/uvok/statuses/113243833403213516",
"activitypub"
]
],
"content": "Imagine trying to load a #TensorFlow model and all you get is a scary traceback and\n\n 'cuLaunchKernel...)' failed with 'CUDA_ERROR_INVALID_HANDLE'\n\n\nThanks for nothing, I guess. The search results don't help me.\n\n(TF installed via mamba / conda-forge)\n\nor rather, CUDA_ERROR_NO_BINARY_FOR_GPU ?",
"sig": "222cf2e5a3268fa1b9a15e64d1fd59e58d52578021c0b4f9690e961a75caffc65ecdceba036972853d8d74f7705bd0e6bf5cf669c0fd5111e497fd217f7d8a35"
}
