Open CHYjeremy opened 1 year ago
I've been running on ubuntu 20.04. I wasn't able to make it work using the nvidia drivers and cuda that are distributed with ubuntu; I needed to uninstall these and install nvidia's versions of the driver, CUDA, and CUDNN which match the JAX version. However, after installing, CUDA wasn't on my PATH (IIRC I did get a similar error message about failing to load PTX as a result). I found that export PATH=/usr/local/cuda/bin:$PATH
before running the live demo made it work.
I have no idea if this is your issue, however. You might be better off posting this question on the JAX github.
have you solved your problem? if you solved, could you please show the resolution?
@xbowlove I did the following things and running the live demo on local laptop GPU worked for me.
created local virtual env using venv
git clone https://github.com/deepmind/tapnet.git
pip install --upgrade "jax[cuda12_pip]" -f https://storage.googleapis.com/jax-releases/jax_cuda_releases.html
For Referencejax jaxline
from given tapnet/requirements_inference.txt
. pip install -r requirements_inference.txt
After this followed the rest step mentioned in the REAME file to run the live demo.
thanks for your answer. but I occur the following error when 'from jaxline import platform' after finishing all the steps you offered.
Traceback (most recent call last):
File "/home/jishengyin/newpan/tapnet/./experiment.py", line 30, in
Hi everyone,
i find it really hard to get tapir to run on gpu, is there a standard procedure to do this?
the thing i do/try is: (after i create a new conda environment)
and the following error pops out jaxlib.xla_extension.XlaRuntimeError: INTERNAL: Failed to execute XLA Runtime executable: run time error: custom call 'xla.gpu.func.launch' failed: Failed to load PTX text as a module: CUDA_ERROR_INVALID_IMAGE: device kernel image is invalid; current tracing scope: fusion; current profiling annotation: XlaModule:#hlo_module=jit__threefry_seed,program_id=0#.
note that using only cpu version of this wouldn't hurt (simply pip install requirement_inference.txt)
could someone state your standard procedure for making it work? much thanks