python3 -m trains_agent --config-file "~/trains.conf" daemon --queue default --log-level DEBUG --detached --gpus 1 > ~/trains-agent.startup.log 2>&1
I specified a torch @ https://download.pytorch.org/whl/cu100/torch-1.3.1%2Bcu100-cp36-cp36m-linux_x86_64.whl and it didn't detect the link, it tried to install latest version: 1.6.0
What you actually specified is torch the @ is kind of pip remark, pip will not actually parse it 🙂
use only the link https://download.pytorch.org/whl/cu100/torch-1.3.1%2Bcu100-cp36-cp36m-linux_x86_64.whl
(since you are using venv mode, if the cuda is not detected at startup time, it will not install the GPU version, as it has no CUDA support)
AgitatedDove14 one last question: how can I enforce a specific wheel to be installed?
That depends on what you have installed 🙂
This is also set in the command line.
--cpu-only or maybe without any --gpus flag at all
cuda 10.1, I guess this is because no wheel exists for torch==1.3.1 and cuda 11.0
Correct
how can I enforce a specific wheel to be installed?
You mean like specific CUDA wheel ?
you can simple put the http link to the wheel in the "installed packages", it should work
what do you see in the console when you start the trains-agent , it should detect the cuda version
Hi JitteryCoyote63
What do you have in the agent.cuda_version ?
(you can see it printed at the beginning of the log)
OK but nowhere I specified that, I just checked my trains.conf file
I have 11.0 installed but on another machine with 11.0 installed as well, trains downloads torch for cuda 10.1, I guess this is because no wheel exists for torch==1.3.1 and cuda 11.0
You can always force it with environment variable CUDA_VERSION=10.1
I don't know why it didn't detect it in first place
agent.cuda_version = 110 agent.cudnn_version = 0
agent.cuda_version = 0 agent.cudnn_version = 0